hypoxia experimental model

Experimental Hypoxia as a Model for Cardiac Regeneration in Mice

Series: Methods In Molecular Biology > Book: Cardiac Regeneration

Protocol | DOI: 10.1007/978-1-0716-0668-1_25

Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX, USA
Division of Cardiology, Department of Internal Molecular Biology, University of Texas Southwestern Medical Center, Dallas, TX, USA

Experimental hypoxia has been used for decades to examine the adaptive response to low-oxygen environments. Various models have been studied, including flies, worms, fish, rodents, and humans. Our lab has recently used this technology to examine the effect of environmental hypoxia on mammalian heart regeneration. In this chapter, we describe studies of systemic hypoxia in mice. We found that systemic hypoxia can blunt oxidative DNA damage and induce cardiomyocyte proliferation. While our primary interests are focused on cardiovascular research, these hypoxia protocols are applicable to any other organ system.

Figures ( 0 ) & Videos ( 0 )

Experimental specifications, other keywords.

Measurement of in vivo vldl and chylomicron secretion, toxoplasma gondii as a model of in vivo host-parasite interactions, a rotavirus-induced mouse model to study biliary atresia and neonatal cholestasis, assessing cellular hypoxic status in situ within the bone marrow microenvironment, method for generating a patient-derived xenograft model of cll, cancer susceptibility models in protease-deficient mice, oxazolone-induced colitis as a model of th2 immune responses in the intestinal mucosa, immunocytochemical assessment of blood–brain barrier structure, function, and damage, mast cell development and function in the zebrafish, mouse model of experimental dermal fibrosis: the bleomycin-induced dermal fibrosis.

Porrello ER, Mahmoud AI, Simpson E, Hill JA, Richardson JA, Olson EN, Sadek HA (2011) Transient regenerative potential of the neonatal mouse heart. Science 331:1078–1080
Porrello ER, Mahmoud AI, Simpson E, Johnson BA, Grinsfelder D, Canseco D, Mammen PP, Rothermel BA, Olson EN, Sadek HA (2013) Regulation of neonatal and adult mammalian heart regeneration by the miR-15 family. Proc Natl Acad Sci USA 110:187–192
Bergmann O, Bhardwaj RD, Bernard S, Zdunek S, Barnabe-Heider F, Walsh S, Zupicich J, Alkass K, Buchholz BA, Druid H, Jovinge S, Frisen J (2009) Evidence for cardiomyocyte renewal in humans. Science 324:98–102
Senyo SE, Steinhauser ML, Pizzimenti CL, Yang VK, Cai L, Wang M, Wu TD, Guerquin-Kern JL, Lechene CP, Lee RT (2013) Mammalian heart renewal by pre-existing cardiomyocytes. Nature 493:433–436
Dawes GS, Mott JC, Widdicombe JG (1954) The foetal circulation in the lamb. J Physiol 126:563–587
Poss KD, Wilson LG, Keating MT (2002) Heart regeneration in zebrafish. Science 298:2188–2190
Kikuchi K, Holdway JE, Werdich AA, Anderson RM, Fang Y, Egnaczyk GF, Evans T, Macrae CA, Stainier DY, Poss KD (2010) Primary contribution to zebrafish heart regeneration by gata4(+) cardiomyocytes. Nature 464:601–605
Jopling C, Sleep E, Raya M, Marti M, Raya A, Izpisua Belmonte JC (2010) Zebrafish heart regeneration occurs by cardiomyocyte dedifferentiation and proliferation. Nature 464:606–609
Rees BB, Sudradjat FA, Love JW (2001) Acclimation to hypoxia increases survival time of zebrafish, Danio rerio, during lethal hypoxia. J Exp Zool 289:266–272
Roesner A, Hankeln T, Burmester T (2006) Hypoxia induces a complex response of globin expression in zebrafish (Danio rerio). J Exp Biol 209:2129–2137
Puente BN, Kimura W, Muralidhar SA, Moon J, Amatruda JF, Phelps KL, Grinsfelder D, Rothermel BA, Chen R, Garcia JA, Santos CX, Thet S, Mori E, Kinter MT, Rindler PM, Zacchigna S, Mukherjee S, Chen DJ, Mahmoud AI, Giacca M, Rabinovitch PS, Aroumougame A, Shah AM, Szweda LI, Sadek HA (2014) The oxygen-rich postnatal environment induces cardiomyocyte cell-cycle arrest through DNA damage response. Cell 157:565–579
Nakada Y, Canseco DC, Thet S, Abdisalaam S, Asaithamby A, Santos CX, Shah AM, Zhang H, Faber JE, Kinter MT, Szweda LI, Xing C, Hu Z, Deberardinis RJ, Schiattarella G, Hill JA, Oz O, Lu Z, Zhang CC, Kimura W, Sadek HA (2017) Hypoxia induces heart regeneration in adult mice. Nature 541:222–227
Mohyeldin A, Garzon-Muvdi T, Quinones-Hinojosa A (2010) Oxygen in stem cell biology: a critical component of the stem cell niche. Cell Stem Cell 7:150–161
Cipolleschi MG, Dello Sbarba P, Olivotto M (1993) The role of hypoxia in the maintenance of hematopoietic stem cells. Blood 82:2031–2037
Kimura W, Xiao F, Canseco DC, Muralidhar S, Thet S, Zhang HM, Abderrahman Y, Chen R, Garcia JA, Shelton JM, Richardson JA, Ashour AM, Asaithamby A, Liang H, Xing C, Lu Z, Zhang CC, Sadek HA (2015) Hypoxia fate mapping identifies cycling cardiomyocytes in the adult heart. Nature 523:226–230

Alzheimer's disease & dementia
Arthritis & Rheumatism
Attention deficit disorders
Autism spectrum disorders
Biomedical technology
Diseases, Conditions, Syndromes
Endocrinology & Metabolism
Gastroenterology
Gerontology & Geriatrics
Health informatics
Inflammatory disorders
Medical economics
Medical research
Medications
Neuroscience
Obstetrics & gynaecology
Oncology & Cancer
Ophthalmology
Overweight & Obesity
Parkinson's & Movement disorders
Psychology & Psychiatry
Radiology & Imaging
Sleep disorders
Sports medicine & Kinesiology
Vaccination
Breast cancer
Cardiovascular disease
Chronic obstructive pulmonary disease
Colon cancer
Coronary artery disease
Heart attack
Heart disease
High blood pressure
Kidney disease
Lung cancer
Multiple sclerosis
Myocardial infarction
Ovarian cancer
Post traumatic stress disorder
Rheumatoid arthritis
Schizophrenia
Skin cancer
Type 2 diabetes
Full List »

share this!

August 2, 2024

This article has been reviewed according to Science X's editorial process and policies . Editors have highlighted the following attributes while ensuring the content's credibility:

fact-checked

trusted source

Novel approach to study hypoxia enables identification of a marker for ovarian cancers

by Myreille Larouche, University of Montreal

In a new study, the team led by Étienne Gagnon, Professor in the Department of Microbiology, Infectious Diseases and Immunology at the Université de Montréal and Director of IRIC's Cancer Immunobiology Research Unit, has developed a cell culture protocol that accurately reproduces the characteristic conditions of primary tumors.

The group also identified a novel form of the WT1 protein associated with poor long-term survival in ovarian cancer patients. Published in the journal Cancer Gene Therapy , the study was led by doctoral student Jordan Quenneville.

The LTHY method: To reproduce hypoxic conditions in the laboratory

Hypoxia, a reduced oxygen availability, characterizes the cellular environment of many solid tumors. This characteristic contributes to the resistance of tumor cells to chemotherapy, radiotherapy and immunotherapy. Hypoxia is thus associated with a poor prognosis for patients. However, existing methods for studying hypoxia in the laboratory do not reproduce the conditions observed during tumor development.

To overcome this problem, the Gagnon laboratory team has developed a new cell culture protocol, called LTHY (for "long-term hypoxia"). This method mimics the progressive development of severe hypoxia observed in vivo. The approach developed combines both duration and severity to mimic the appearance and progression of a tumor. This novel protocol is already setting new standards, and several other IRIC groups are beginning to use it in their respective research projects.

A new marker of aggressiveness and survival?

Cells subjected to the LTHY protocol spontaneously undergo an epithelial-mesenchymal transition (EMT), making them invasive and eventually leading to metastasis. At the start of this transition, the cultured cells produce a truncated form of the WT1 protein, known to promote EMT and the development of cancer.

The team also discovered that this new form is derived from an intronic portion of the WT1 gene, a region not normally used for protein production. The resulting product, in addition to being truncated, therefore contains an unconventional protein sequence, known as cryptic. The truncated form of WT1 remains functional and binds to several genes involved in EMT.

Although identified in several human cancer samples, this form of WT1 is particularly present in ovarian cancers, which are known to be hypoxic. Moreover, its presence in this type of cancer is associated with poor long-term survival for patients. Thus, the truncated form of WT1 could become a new marker for predicting aggressiveness and survival in ovarian cancer .

Explore further

Feedback to editors

Hospital pneumonia diagnoses are uncertain, revised more than half the time, study finds

EMS training on key skills improves heart attack survival

2 hours ago

Effective mental health care takes varying forms, says new study

New biomaterial regrows damaged cartilage in joints

3 hours ago

A new drug could turn back the clock on multiple sclerosis

4 hours ago

New method tracks how psychedelics affect neurons in minutes

Botanicals like turmeric, green tea are harming Americans' livers

5 hours ago

In law enforcement, a survey finds a link between head injuries and depression, PTSD

6 hours ago

Study finds omega-3 supplements reduce genetic risk of high total cholesterol, LDL and triglyceride levels

Honey added to yogurt supports probiotic cultures for digestive health

Molecular mechanisms regulating prostate cancer hypoxia identified

Sep 5, 2022

Team identifies protein key to cancer cells ability to spread

Nov 17, 2017

New research sheds light on why tumor cells become resistant to chemotherapy

Sep 30, 2020

Oxygen-starved tumor cells have survival advantage that promotes cancer spread

Nov 7, 2019

Smart hydrogel with oxygen-scavenging capability inhibits tumor growth and metastasis

Apr 4, 2023

Tumor resistance is promoted by anti-cancer protein

Sep 19, 2019

Recommended for you

Study finds noncoding RNAs dysregulated in several human cancers

10 hours ago

Proposed lung cancer screening guideline could eliminate racial disparities, allow for early detection

8 hours ago

Researchers develop promising therapy treatment that can kill glioblastoma cells in newly-discovered brain pathway

Aug 2, 2024

Researchers devise novel solution to preventing relapse after CAR T-cell therapy

Researchers highlight the genetic complexity of schizophrenia

Let us know if there is a problem with our content.

Use this form if you have come across a typo, inaccuracy or would like to send an edit request for the content on this page. For general inquiries, please use our contact form . For general feedback, use the public comments section below (please adhere to guidelines ).

Please select the most appropriate category to facilitate processing of your request

Thank you for taking time to provide your feedback to the editors.

Your feedback is important to us. However, we do not guarantee individual replies due to the high volume of messages.

E-mail the story

Your email address is used only to let the recipient know who sent the email. Neither your address nor the recipient's address will be used for any other purpose. The information you enter will appear in your e-mail message and is not retained by Medical Xpress in any form.

Newsletter sign up

Get weekly and/or daily updates delivered to your inbox. You can unsubscribe at any time and we'll never share your details to third parties.

More information Privacy policy

Donate and enjoy an ad-free experience

We keep our content available to everyone. Consider supporting Science X's mission by getting a premium account.

E-mail newsletter

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings
My Bibliography
Collections
Citation manager

Save citation to file

Email citation, add to collections.

Create a new collection
Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

Search in PubMed
Search in NLM Catalog
Add to Search

Experimental Hypoxia as a Model for Cardiac Regeneration in Mice

Affiliations.

1 Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX, USA.
2 Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX, USA. [email protected].
3 Division of Cardiology, Department of Internal Molecular Biology, University of Texas Southwestern Medical Center, Dallas, TX, USA. [email protected].
PMID: 32857385
DOI: 10.1007/978-1-0716-0668-1_25

Keywords: Cardiomyocyte; Heart regeneration; Hypoxia; Mice; Oxygen.

PubMed Disclaimer

Publication types

Search in MeSH

Related information

PubChem Compound (MeSH Keyword)

Grants and funding

R01 HL115275/HL/NHLBI NIH HHS/United States
UL1 TR001105/TR/NCATS NIH HHS/United States
CH/1999001/11735/BHF_/British Heart Foundation/United Kingdom

LinkOut - more resources

Full text sources, miscellaneous.

NCI CPTAC Assay Portal
Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Open access
Published: 29 July 2024

Pericytes recruited by CCL28 promote vascular normalization after anti-angiogenesis therapy through RA/RXRA/ANGPT1 pathway in lung adenocarcinoma

Ying Chen 1 , 2 , 3 ,
Zhiyong Zhang 1 , 2 , 3 ,
Fan Pan 3 , 6 ,
Pengfei Li 1 , 2 , 3 ,
Weiping Yao 3 , 6 ,
Yuxi Chen 1 , 2 , 3 ,
Lei Xiong 5 ,
Tingting Wang 1 , 2 , 3 ,
Yan Li 4 &
Guichun Huang 6 , 7

Journal of Experimental & Clinical Cancer Research volume 43 , Article number: 210 ( 2024 ) Cite this article

383 Accesses

Metrics details

It has been proposed that anti-angiogenesis therapy could induce tumor "vascular normalization" and further enhance the efficacy of chemotherapy, radiotherapy, target therapy, and immunotherapy for nearly twenty years. However, the detailed molecular mechanism of this phenomenon is still obscure.

Overexpression and knockout of CCL28 in human lung adenocarcinoma cell line A549 and murine lung adenocarcinoma cell line LLC, respectively, were utilized to establish mouse models. Single-cell sequencing was performed to analyze the proportion of different cell clusters and metabolic changes in the tumor microenvironment (TME). Immunofluorescence and multiplex immunohistochemistry were conducted in murine tumor tissues and clinical biopsy samples to assess the percentage of pericytes coverage. Primary pericytes were isolated from lung adenocarcinoma tumor tissues using magnetic-activated cell sorting (MACS). These pericytes were then treated with recombinant human CCL28 protein, followed by transwell migration assays and RNA sequencing analysis. Changes in the secretome and metabolome were examined, and verification of retinoic acid metabolism alterations in pericytes was conducted using quantitative real-time PCR, western blotting, and LC–MS technology. Chromatin immunoprecipitation followed by quantitative PCR (ChIP-qPCR) was employed to validate the transcriptional regulatory ability and affinity of RXRα to specific sites at the ANGPT1 promoter.

Our study showed that after undergoing anti-angiogenesis treatment, the tumor exhibited a state of ischemia and hypoxia, leading to an upregulation in the expression of CCL28 in hypoxic lung adenocarcinoma cells by the hypoxia-sensitive transcription factor CEBPB. Increased CCL28 could promote tumor vascular normalization through recruiting and metabolic reprogramming pericytes in the tumor microenvironment. Mechanistically, CCL28 modified the retinoic acid (RA) metabolism and increased ANGPT1 expression via RXRα in pericytes, thereby enhancing the stability of endothelial cells.

We reported the details of the molecular mechanisms of "vascular normalization" after anti-angiogenesis therapy for the first time. Our work might provide a prospective molecular marker for guiding the clinical arrangement of combination therapy between anti-angiogenesis treatment and other therapies.

Introduction

Lung cancer stands as the foremost cause of mortality among patients with malignant tumors, with lung adenocarcinoma being a predominant pathological subtype [ 1 , 2 ]. Angiogenesis is a crucial hallmark of lung adenocarcinoma, and significant strides have been made in applying anti-angiogenesis therapy for its treatment [ 3 ]. Extensive research data indicates that anti-angiogenic tumor therapy exhibits the potential to augment the effectiveness of diverse therapeutic modalities, including chemotherapy, radiotherapy, tyrosine kinase inhibitors (TKIs), and immunotherapy. Nevertheless, a subset of patients fails to derive benefits from the combined approach of anti-angiogenic therapy and other anti-tumor treatments. Unraveling the synergistic mechanisms underlying anti-angiogenic therapy holds substantial clinical implications for optimizing combination therapies.

Since the 2000s, there has been a fundamental shift in the understanding of anti-angiogenic therapy for tumors, transitioning from the initial concept of solely inhibiting angiogenesis to inducing vascular normalization in tumors [ 4 ]. Presently, there is a consensus that optimizing the effectiveness of additional anti-tumor treatments through anti-angiogenic therapy is imperative. This notion is central to anti-angiogenic drugs' capability to induce transient vascular normalization in tumor blood vessels [ 5 ]. Preliminary investigations undertaken by our research group propose that anti-angiogenic drugs, employing diverse mechanisms of action, can prompt vascular normalization in lung adenocarcinoma [ 6 , 7 , 8 , 9 ]. The balance between pro-angiogenesis and anti-angiogenesis factors was postulated as the key molecular base of vascular normalization [ 10 ]. However, uncertainties persist regarding the timing and the limited duration of vascular normalization post-treatment. Addressing this issue is critical, as the temporal aspects of vascular normalization and its short-lived nature need clarification [ 7 , 8 , 9 , 11 , 12 ]. The pressing clinical challenge involves establishing protocols to monitor and regulate vascular normalization in lung adenocarcinoma, offering valuable guidance for developing combined anti-angiogenic therapies with other anti-tumor treatment strategies. A comprehensive understanding of the molecular mechanisms underpinning anti-angiogenic therapy-induced vascular normalization in lung adenocarcinoma forms the foundational basis for resolving this intricate clinical dilemma.

Pericytes, one of the structural cells found in capillary vessels, envelop the basal membrane of endothelial cells and interact with vascular endothelial cells to ensure the stability and functionality of microvessels [ 13 , 14 ]. Commonly used markers for identifying pericytes include platelet-derived growth factor receptor β (PDGFRβ), chondroitin sulfate proteoglycan 4 (CSPG4, also known as NG2), and alpha-smooth muscle actin (α-SMA). Several reports suggest that platelet-derived growth factor B (PDGFB) can promote vascular normalization in colon cancer and malignant melanoma by recruiting pericytes [ 15 ]. However, PDGFB does not perform this function in pericytes in lung adenocarcinoma [ 16 ]. The mechanism of pericyte recruitment in lung adenocarcinoma following anti-angiogenesis therapy remains unclear.

Human CC motif chemokine ligands 28 (CCL28) can recruit various cell types, including lymphatic endothelial cells [ 13 ], T cells, and plasma cells, through its receptors CCR3 or CCR10. Apart from its chemotactic properties, CCL28 has been implicated in promoting tumor development in ovarian cancer [ 17 ], gastric cancer [ 18 ], liver cancer, and lung adenocarcinoma [ 19 ]. Under hypoxic conditions, tumor cells mobilize various cell types using different chemokines to induce angiogenesis in response to nutritional and oxygen requirements. Notably, hypoxia induces the expression of CCL28 in ovarian cancer, which, in turn, enhances angiogenesis by recruiting regulatory T cells (Tregs) [ 17 ]. In contrast, CCL28 has been identified as a negative regulator of tumor growth and bone invasion in oral squamous cell carcinoma [ 20 ]. These diverse findings suggest that CCL28 may play distinct roles in various tumors. While one study discovered that CCL28 increased the proliferation, migration, and secretion of IL-6 and HGF in oral fibroblasts [ 21 ], the functional role of CCL28 in pericytes in lung adenocarcinoma has not been elucidated. Furthermore, previous studies have reported that CCL28 was involved in angiogenesis in various diseases, including tumors, skin wound healing [ 22 ], and rheumatoid arthritis [ 23 ]. However, little is known about the mechanism of CCL28 in tumor vascular normalization.

Our previous report highlighted that CCL28 can moderately enhance the angiogenesis in lung adenocarcinoma [ 19 ]. Interestingly, we found a normalized vasculature in CCL28 highly expressed tumors. The current study delves into the previously undisclosed connection between pericytes and CCL28 in lung adenocarcinoma. Both in vivo and in vitro experiments collectively illustrate that CCL28 derived from tumor cells is pivotal in advancing vascular normalization by mobilizing and reprogramming pericytes.

Materials and methods

Cancer cell lines.

The human lung adenocarcinoma cell line (A549, SPC-A1, and H1975) and mouse Lewis lung cancer cell line (LLC) were purchased from the Shanghai Institute of Biochemistry and Cell Biology (SIBCB) and maintained in our lab. Lung cancer cells were cultured in RPMI-1640, or DMEM medium (Gibco, LifeTech, USA) supplemented with 10% fetal bovine serum (FBS), 1% antibiotics (100U/ml penicillin and 100 μg/ml streptomycin) at 37℃ in a humidified 5% CO 2 atmosphere.

Isolation and identification of pericytes

Pericytes were isolated and cultured as we previously described [ 24 ]. Fresh lung cancer samples were collected from patients in Jinling Hospital (Nanjing, China). Tissue samples were cut into small blocks of approximately 1 ~ 2 mm in diameter, digested with trypsin and 0.5% collagenase, and filtered through the cell strainer to obtain single-cell suspension. PDGFRβ + cells were isolated by PE-PDGFRβ antibody and anti-PE magnetic beads (Miltenyi Biotec,130–123-772). The separated cells were cultured in F12K medium (Gibco, LifeTech, USA) containing 10% fetal bovine serum (Gibco, Life Tech, USA) with 100 U/mL penicillin and 100 μg/mL streptomycin. After two or three passages, pericytes were identified by morphology and immunofluorescence staining for α-Smooth Muscle Actin (α-SMA, CST, #19,245), platelet-derived growth factor receptor β (PDGFRβ, Abcam, ab69506) and chondroitin sulfate proteoglycan 4 (NG2, Abcam, ab129051). Patients with incomplete data were excluded to evaluate the clinical effect. Written informed consent was obtained from all subjects before collecting the samples. All the methods followed the institutional guidelines and were approved by the Ethical Review Committee of Jinling Hospital, Nanjing, China(2022DZGZR-QH-005).

Hypoxic culture model and RNA sequence assay

As previously reported, the hypoxic cell culture model was established with hypoxic chambers in our lab [ 19 ]. Briefly, lung adenocarcinoma cells were cultured under two different oxygen concentrations, 1% and 20%, respectively. The model was testified by the expression changes of HIF-1α and its regulated genes, such as GLUT1 and VEGFA (Fig. 1 C). Pericytes were cultured under different stimulation. After culturing for 24 h, the cells were collected, and the total amount of RNA or protein was extracted for western blot or quantitative PCR. RNA sequence assay was applied to detect the gene expression differences of pericytes cultured with or without the stimulation of CCL28 (MCE, HY-P7250) under hypoxia.

CCL28 expression is upregulated after anti-angiogenesis therapy by hypoxia-sensitive transcription factor CEBPB in lung adenocarcinoma

RNA extraction and quantitative PCR

Total RNA was extracted from different cell lines using Trizol reagent (Invitrogen, USA). Subsequently, reverse transcription and quantitative PCR were performed using SYBR Green with an ABI StepOne Plus System (Applied Biosystems, Life Tech, USA). Relative gene expression was calculated by the ΔΔCt method based on glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or β-actin levels. All reactions were run in triplicate.

Luciferase reporter assay

Wild-type (pGL3 wt, 5’-TGATTATGCAATGG-3') and mutant (pGL3 mut, 5’-ACTAATACGTTACC-3') promoters of the CCL28 gene were constructed into pGL3 firefly luciferase reporter plasmid vector purchased from Nanjing Realgene Bio-Technology Company (Nanjing China). The pRL-TK vector expressing renilla luciferase was used as an internal control. The reporter plasmid was co-transfected with pRL-TK vector and CEBPB expressing pcDNA3.1 vector or control vector into A549 cell. Dual-Luciferase Reporter Assay System (Promega, USA) was used to detect the luciferase activity after 48-h incubation.

Chromatin Immunoprecipitation (ChIP)

Briefly, pericytes treated with or without recombinant human CCL28 or CCR3 (R&D Systems, MAB155-100) neutralizing antibodies were collected and fixed by adding a cross-linking agent, formaldehyde, to stabilize the interactions between chromatin proteins and DNA. Subsequently, chromatin immunoprecipitation was conducted to enrich DNA fragments bound by chromatin proteins. After cross-link reversal and DNA purification, real-time quantitative PCR (qPCR) technology is employed to measure enriched DNA, determining the relative abundance at specific gene loci.

Liquid Chromatography-Mass Spectrometry (LC–MS)

Liquid chromatography-mass spectrometry was applied to measure retinoic acid (RA, HY-14649) in pericytes. Briefly, after treatment with or without CCL28, pericytes were subjected to enzymatic digestion, washed three times with PBS, flash-frozen in liquid nitrogen for one minute, and then stored at -80 °C. When performing the detection, begin by thawing the sample at 4 °C. Next, add 0.5 mL of methanol solution, followed by 10 min of sonication and 30 min of shaking for extraction. Subsequently, place the centrifuge tube in a low-temperature centrifuge and centrifuge for 10 min at 4 °C and 12,000 rpm. Collect 800μL of the supernatant, evaporate it, and then add 100μL of methanol solution. Finally, retain 80 μL of the supernatant for subsequent liquid chromatography-mass spectrometry analysis. A standard curve is generated to calculate the sample's content using a retinoic acid standard (yuanye Bio-Technology).

Single-cell RNA sequencing

Lewis lung cancer cells overexpressing CCL28 were inoculated in C57BL/6 mice subcutaneously (1 × 10 6 cells per mouse). When the tumor grew to about 100 mm 3 , fresh tumor samples were collected and dissociated into single cells. The tissue was washed with PBS, then chopped and incubated in an enzyme digestion solution for 40 min at 37 °C. After digestion, the solution was filtered through a 40 μm filter and centrifuged. Red blood cells were removed using a red blood cell lysis solution, followed by cell counting. The cell suspension was filtered using FACS tubes, centrifuged again, and washed twice. Finally, after microscopic examination and cell counting, the single-cell suspension was used for scRNA sequencing.

The scRNA sequencing analysis was carried out according to a single cell analysis workflow with BD Rhapsody™ Systems (BD Biosciences, USA), and RNA sequencing and data analysis were completed on the platform of NovaSeq 6000 system (Illumina, USA). The entired scRNA sequencing workflow was provided in the Supplementary document 1.

Animal model

Six-week-old female BALB/c nude mice and C57BL/6 female mice were used for tumor model establishment. LLC (1 × 10 6 per mouse) with or without overexpressing CCL28 were subcutaneously injected in C57BL/6 mice ( n = 6, each group). Mice inoculated with CCL28 knock-out LLC cells and wild-type LLC were randomly divided into two groups, respectively, and treated with or without RA by gavage daily (n = 6, each group). Mice were observed, and the tumors' length and width were measured daily. When tumors grew to a specific size, subcutaneous tumors were collected and photographed. The A549 (CCL28 overexpression or CCL28 knock-out) tumor model followed the same procedure. All animal experiments were carried out following the institutional guidelines and approved by the Ethical Review Committee of Comparative Medicine, Jinling Hospital, Nanjing, China (2022DZGKDWLS-0091).

Western blot and ELISA

Proteins were extracted from the cultured cells by lysis buffer, separated by SDS-PAGE, and transferred to PVDF membranes (Millipore, USA). The filters were blocked in Tris-buffered saline containing 0.2% Tween plus 5% non-fat milk and incubated with primary antibodies overnight at 4 °C. Secondary antibodies were used for visualization through chemiluminescence (ECL, Amersham Pharmacia Biotech, UK). Primary antibodies against RXRα (CST,1:1000), RARα (CST,1:1000), RDH13 (Proteintech,1:1000), DHRS11 (Proteintech,1:1000), CCR10 (Abmart,1:1000), β-actin (Servicebio, 1:1000) and neutralizing antibody against CCR3 (R&D Systems, USA) were applied in the present study. Anti-rabbit IgG, HRP-linked Antibody (CST, 1:1000) and anti-mouse IgG, HRP-linked Antibody (CST, 1:1000) were utilized as the secondary antibodies. CCL28 in serum was detected by an ELISA kit (Abcam, USA) according to the manufacturer's instructions.

Immunofluorescence and multiplex Immunohistochemistry (mIHC)

Tumor samples were collected and fixed in 10% formalin before processing and paraffin embedding. Immunofluorescence was conducted on 5 µm sections. For culture cell staining, lung adenocarcinoma-associated pericytes were washed with 1 × PBS and fixed with acetone for 10 min on ice. Tumor microvascular endothelial cells and cancer pericytes were stained by CD31 antibody (Abcam, ab281583) and rat anti-human/mouse monoclonal NG2 antibody (Abcam, ab129051). At the same time, tumor cells were stained by mouse anti-human pan-cytokeratin (Pan-ck) monoclonal antibody (Abcam, ab215838). Goat anti-mouse IgG antibody (labeled with Alex Fluor 488), goat anti-rabbit IgG antibody (labeled with Alex Fluor 555 or Alex Fluor 488), and goat anti-rat IgG antibody (labeled with Alex Fluor 555 or Alex Fluor 488) (Abcam) were applied as secondary antibody in immunofluorescence staining analysis.

For multiplex immunohistochemistry, the slides were incubated with the primary antibodies (anti-CD31, anti-NG2, anti-Pan-CK antibody, and anti-CCL28 (Proteintech, 18,214–1-AP), respectively) and horseradish peroxidase-conjugated secondary antibody, and tyramine signal amplification (TSA) was performed following the pre-optimized antibody concentration and the order of staining. Antibody stripping and antigen retrieval were performed after each round of TSA. DAPI (Sigma-Aldrich, USA) was used for nuclei staining. A whole slide scan of the multiplex tissue sections produced multispectral fluorescent images visualized in SlideViewer software. A specialized pathologist chose representative regions of interest (ROI), and multiple fields of view were acquired at 20 × power for further analysis. The mean fluorescence intensity of CCL28 in each ROI was calculated by the ImageJ program ( https://imagej.nih.gov/ij/ ).

CRISPR-cas9 knock out

Briefly, three designed sgRNAs were synthesized and co-transfected with Cas9 nuclease/sgRNA in A549 lung adenocarcinoma cells. After 72 h, the cells were seeded in a 96-well plate for single-cell cloning by limiting dilution analysis (LDA). After 7 to 15 days, about ten single-cell clones were selected for PCR and ELISA. Mutant sites were further confirmed by DNA sequencing.

Knockdown of RDH13 and DHRS11

For RDH13 and DHRS11 gene silence, pericytes were infected with negative control siRNAs, RDH13-targeting, or DHRS11-targeting siRNAs (genepharma). Briefly, pericytes were seeded in 24-well plates, and the cells were transfected with siRNAs using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions.

Bioinformatic analysis

Transcription factor binding sites prediction in promotor of CCL28 gene was conducted in PROMO 3.0, virtual laboratory TFSEARCH (ver.1.3) and JASPAR ( https://alggen.lsi.upc.edu/rerecer/menu_recerca.html , https://diyhpl.us/~bryan/irc/protocol-online/protocol-cache/TFSEARCH.html , https://jaspar.elixir.no/ ). RNA sequencing dates of lung adenocarcinoma tumor samples or lung adenocarcinoma cell lines were extracted on the cBioPortal platform ( https://www.cbioportal.org/ ). The relationship between expression levels of CEBPB and CCL28, RXRα, and ANNGPT1 were calculated online, as well as the predictive value of CEBPB for the survival of lung adenocarcinoma patients.

Statistical analyses

Data were presented as mean ± SEM. Student's unpaired two-tailed tests were used for comparisons between two groups. Pearson or Spearman correlation was applied to analyze the relationship between the expression scores of two genes. Statistical analyses were performed on GraphPad Prism 8.0. Multiple comparisons were analyzed by one-way ANOVA using the LSD test. Statistical significance was confirmed when p < 0.05.

CCL28 expression is up-regulated after anti-angiogenesis therapy by hypoxia-sensitive transcription factor CEBPB in lung adenocarcinoma

The GEO database ( https://www.ncbi.nlm.nih.gov/gds/ ) was retrieved to elucidate the impact of anti-angiogenesis therapy on the gene expression of cancer patients. Molecular expression profile data (GSE61676) of lung adenocarcinoma patients undergoing anti-angiogenesis treatment was reanalyzed. Twenty-three stage IV lung adenocarcinoma patients, treated with a combination of anti-VEGF monoclonal antibody (Bevacizumab) and TKI, were selected for analysis. The changes in mRNA profile (Affymetrix Human Exon 1.0 ST Arrays) were examined before and 24 h after Bevacizumab treatment. The findings revealed a significant increase in the hypoxia index (GLUT1, one of the primary target genes regulated by hypoxia-inducing factor HIF-1) and up-regulation of CCL28 expression in lung adenocarcinoma patients 24 h after VEGF monoclonal antibody treatment compared to baseline levels (Fig. 1 A, 1B). After treatment with bevacizumab, CCL28 expression showed a log2 fold change of 1.258 compared to baseline, with a corresponding p-value of 0.0152, as depicted in Fig. 1 A. After anti-angiogenesis therapy, tumor cells often experience a state of ischemia and hypoxia. To simulate this hypoxic state of tumor cells, nine cell lines of 4 different tumor types were culture in hypoxia and normoxia conditions, including two hepatoma cell lines (HEPG2 and SMMC), three lung adenocarcinoma cell lines (A549, SPC-A1, and H1975), two breast carcinoma cell lines (MCF7 and MDA-MB-231), and two colorectal cancer cell lines (HCT116 and SW480). The expression of GLUT1 and VEGFA, which was driven by the hypoxia-induced factor (HIF), was detected by qRT-PCR. VEGFA and GLUT1 were up-regulated in all nine cell lines in the hypoxic chamber, indicating that the hypoxic microenvironment was established (Fig. 1 C). We previously reported that high expression of CCL28 was induced in hypoxic lung adenocarcinoma cell lines, A549 and SPC-A1. To further confirm our study, we examined the expression of CCL28 under hypoxia conditions in the nine cell lines. CCL28 was up-regulated in all three lung adenocarcinoma cell lines, whereas others were not (Fig. 1 D). Then, we utilized PROMO 3.0 to predict the transcription factors binding sites to the CCL28 promoter region and intersected them with previously reported hypoxia-related transcription factors, revealing six transcription factors involved (Fig. 1 E). CEBPB and Sp1 can potentially regulate CCL28, while only CEBPB was highly expressed in hypoxic tumor cells (as listed in Supplementary Table 1). We then used a public integrative database (cBioPortal) to analyze the correlation of the expression of CEBPB with that of CCL28 in 44 lung adenocarcinoma cell lines (as listed in Supplementary Table 2). As expected, CEBPB expression strongly correlated with CCL28 (Fig. 1 F). The CEBPB binding motif was shown in Fig. 1 G. Thus, we speculated that hypoxia-induced high expression of CCL28 was mediated by CEBPB. To verify our prediction, we generated luciferase reporter plasmids harboring either wild-type (WT) or mutated (MUT) CEBPB binding sequencing within the promotor element of the CCL28 gene. Luciferase reporter analysis indicated that CEBPB could directly regulate the expression of the CCL28 gene in A549 cells (Fig. 1 H). In addition, disease-free survival and overall survival of lung adenocarcinoma patients with high CEBPB expression were significantly decreased ( p = 0.0065 and p = 0.032, respectively) (Fig. 1 I).

Tumor-derived CCL28 promotes vascular normalization and pericyte recruitment in the tumor microenvironment

To investigate the effects of CCL28 on tumor growth and vascular normalization, we established CCL28 overexpression and knock-out lung adenocarcinoma cells by lentivirus vectors and Cas9 nuclease/sgRNA (Supplementary Fig. 1), respectively. As we previously reported, in vivo studies indicated that CCL28 could promote tumor growth in A549 human lung adenocarcinoma (Fig. 2 A). Microvessels and pericytes were subsequently assessed by immunofluorescence staining using antibodies against CD31 (red) and NG2 (green) in tumor tissue in different groups. The percentage of pericyte coverage and microvessel density (MVD) in A549-CCL28 tumor tissue significantly increased compared to A549-NC, and this effect was reversed after CCL28 was knocked out (Fig. 2 B). These results suggested that CCL28 could promote vascular normalization and mobilize pericytes.

Tumor-derived CCL28 recruits pericytes to promote vascular normalization in the tumor microenvironment

Further, we isolated primary pericytes from lung cancer tissues to explore how CCL28 recruits pericytes in vitro. The characteristics of pericytes were identified by immunofluorescence staining α-SMA, PDGFRβ, and NG2 (Fig. 2 C). Cell viability remained unaffected by different concentrations of recombinant CCL28 in pericytes and A549 (Supplementary Fig. 2A,2B), but enhanced cell migration of A549 (Supplementary Fig. 2C). The transwell migration assay was performed to confirm the effect of CCL28 on the migration of pericytes by using different concentrations of human recombinant CCL28 protein (Fig. 2 D, left panel). The number of cells migrated to the lower chamber was calculated (Fig. 2 D, right panel), and we found that the optimal concentration of CCL28 for enhancing the migration efficiency of pericytes is 250 ng/ml. Thus, we selected a concentration of 250 ng/ml to conduct subsequent migration experiments. To further verify the recruitment effect of CCL28 on pericytes, cells were inoculated around matrigel mixed with or without CCL28 recombinant protein. The graphs visually illustrate cell migration distance and quantity (Fig. 2 E, left panel). CCL28 significantly increased the migration distance of pericytes compared to the control group (Fig. 2 E, right panel). Subsequently, we assessed the expression of two CCL28 receptors, CCR3 and CCR10, in pericytes with or without CCL28. Western blot experiments revealed that the expression of CCR3 was significantly higher than that of CCR10 in pericytes, suggesting that CCR3 likely plays a predominant role in mediating CCL28 signal transduction within pericytes (Fig. 2 F). Blocking CCR3 with neutralizing antibodies diminished the chemotactic effect of CCL28 on pericytes (Fig. 2 G). Further, we validated the relationship between CCL28 and pericytes coverage in biopsy tissue using multiplex immunofluorescence. CCL28 was mainly expressed in lung adenocarcinoma cells due to the fluorescence overlapping between CCL28 and PAN-CK (Fig. 2 H). We found a significant positive correlation between CCL28 expression levels and the percent of pericytes coverage (Fig. 2 I). The above findings indicate that CCL28 could recruit pericytes through the receptor CCR3, thereby promoting vascular normalization. Because the action of CCL28 leads to vascular normalization, making the vascular network healthier and more regular, it may also increase the density and functionality of tumor blood vessels, providing more nutrients and oxygen to promote tumor growth.

Tumor-derived CCL28 promotes expression of angiopoietin-1 via CCR3 in pericytes

Pericytes interact with vascular endothelial cells to promote the maturation of neovasculature [ 25 ]. However, the molecular mechanisms underlying pericytes-induced vascular normalization in lung adenocarcinoma after anti-angiogenesis therapy remain unclear. To clarify the exact role and potential mechanism of CCL28 in vascular normalization, we mimicked the hypoxic microenvironment in lung adenocarcinoma, cultured vascular pericytes, and conducted secretome analysis after CCL28 treatment. We detected augmented production of angiopoietin-1 in pericytes stimulated by CCL28 (Fig. 3 A). Compared to the control group, the expression of ANGPT1 changed by 5.415-fold under CCL28 stimulation, with a p-value of 0.033. RNA-seq results were confirmed by quantitative PCR in different concentrations of recombinant CCL28 protein under hypoxia conditions (Fig. 3 A right panel). Angiopoietin-1, a protein typically associated with angiogenesis and vascular normalization, maintains vascular stability and integrity by interacting with endothelial cells. Consistent with prior studies [ 24 , 26 ], western blot experiments showed that recombinant human angiopoietin-1 could up-regulate endothelial nitric oxide synthase (eNOS) expression in a dose-dependent manner in human umbilical vein endothelial cell (HUVEC) (Fig. 3 B). Likewise, recombinant human angiopoietin-1promoted cell survival by activating the PI3K-AKT signaling pathway in a Tie2-dependent manner in endothelial cells (Fig. 3 C). To further explore the regulatory mechanism of ANGPT1 expression, we identified a significant increase in the transcription factor RXRa and RARα after CCL28 stimulation in transcriptome data, verified by qPCR (Fig. 3 D and E ). However, protein level detected by western blot, CCL28 up-regulated RXRα and CCR3 but not RARα (Fig. 3 F). Also, the effect of CCL28 on RXRα up-regulation was reversed by adding a CCR3 neutralizing antibody (Fig. 3 G). Next, we confirmed the relation between transcription factor RXRα and expression of ANGPT1. Correlation analysis showed that RXRα expression positively correlates with ANGPT1 expression (Fig. 3 H). The binding motif of the RARα and RXRα complex (Fig. 3 I) and the transcription factor binding sites in the promotor area of the CCL28 gene (Fig. 3 J) were shown. CHIP-qPCR was performed to investigate the binding of transcription factor RXRα to the promoter region of the target gene ANGPT1 . Compared to the NC group, CCL28 enhanced the degree of enrichment, and this effect could be dampened after the CCR3 neutralization, suggesting a direct interaction between RXRα and the ANGPT1 promoter. The electrophoresis image displayed the specificity of the amplified DNA fragments (Fig. 3 K).

Tumor-derived CCL28 promotes the expression of angiopoietin-1 via CCR3 in pericytes

Furthermore, the mRNA level of ANGPT1 was increased by the addition of CCL28 but attenuated by the blockade of CCR3 (Fig. 3 L). RXRα is a subtype of nuclear receptor closely associated with vitamin A metabolism and retinoic acid signaling. These results led us to speculate whether retinoic acid metabolism was also involved in regulating vascular normalization by CCL28. Thus, we measured the retinoic acid content in pericytes using LC–MS technology. Interestingly, the retinoic acid production was notably increased (almost 2.5-fold change) compared to the control group in pericytes stimulated by CCL28 (Fig. 3 M). Additionally, the exogenous addition of retinoic acid in pericytes could promote the expression of ANGPT1 in a dose-dependent manner (Fig. 3 N). These results indicate that CCL28 can activate the nuclear transcription factor RXRα through the CCR3 receptor, promoting the expression of ANGPT1 in pericytes. In interaction with endothelial cells, pericytes-deriore, the mRNA level of ANGPT1 was increased by the addition of CCL28 but attenuated by the blockade ofon.

CCL28 activates retinoic acid signaling in pericytes through CCR3

We further analyzed the sequencing data to understand better the mechanism responsible for retinoic acid accumulation in pericytes after CCL28 stimulation. Enrichment analysis of gene pathways showed that RA signaling pathways were activated after CCL28 stimulation (Fig. 4 A). Two key enzymes involved in retinoic acid metabolism have changed, including RDH13 and DHRS11 (Fig.4B). RDH13 plays a role in converting retinol to retinaldehyde, a critical rate-limiting enzyme step in the retinoid metabolic pathway. However, DHRS11 reduces retinoic acid metabolites to related forms of retinol. In this way, the balance between RDH13 and DHRS11 can affect the level of retinoic acid in cells. The metabolic balance model diagram of retinoic acid shows the transformation process of the three substances (Fig. 4 C). In this metabolic process, RDH13 was up-regulated, but DHRS11 was decreased in pericytes after treatment of CCL28 (Fig. 4 B). These results were confirmed by qPCR (Fig. 4 D) and western blot (Fig. 4 E). The correlation analysis exhibited a positive relationship between RDH13 and CCL28 in lung adenocarcinoma (Fig. 4 F). Expression changes of RDH13 and DHRS11 in pericytes stimulated by CCL28 could be reversed by blocking CCR3 (Fig. 4 G). However, treating pericytes directly with retinoic acid did not result in changes in key molecules involved in RA metabolism (Supplementary Fig. 3 A). To further investigate whether the expression of two key enzymes in the retinoic acid synthesis process is correlated with the expression of ANGPT1, we knocked down the expression of RDH13 and DHRS11 using siRNA and then examined the expression of ANGPT1. After knockdown of RDH13 (Fig. 4 H), RXRα and ANGPT1 were significantly reduced (Fig. 4 I, J). However, DHRS11 did not exhibit this effect (Supplementary Fig. 3B, 3C). To further elucidate the relationship between CCL28 and its downstream molecules, immunofluorescence staining of key molecules—RDH13, DHRS11, and ANGPT1—was performed on tumor biopsy tissues (Fig. 4 K, left panel), followed by correlation analysis. In these samples, we observed a positive correlation between CCL28 expression and the levels of ANGPT1 and RDH13, while a negative correlation was found with DHRS11 (Fig. 4 K, right panel).All these results illuminated that CCL28 promoted the retinoic acid synthesis process by disturbing the balance between RDHs and DHRS, but the changes are temporary and dynamic.

Retinoic acid signaling is activated by CCL28 in pericytes through CCR3 A , Volcano plot of changes in metabolic pathways after CCL28 stimulation. B Volcano plot of the enrichment of gene expression after CCL28 stimulation. C Diagram of the metabolic conversion process in the retinoic acid metabolic signaling pathway. D and E Expression of RDH13 and DHRS11 detected by qPCR and western blot with or without exogenous supplement of CCL28. F Correlation of expression of CCL28 with RDH13 in lung adenocarcinoma. G The protein level of DHRS11 and RDH13 stimulated with or without CCL28 and CCR3 neutralizing antibody in pericytes (left) and gray value was calculated(right). H Knockdown efficiency of RDH13 was confirmed by qPCR. I and J Relative expression of RXRα and ANGPT1 after knockdown of RDH13 with or without stimulation of CCL28. K Representative immunofluorescence images of PAN-CK, NG2, CCL28 with DHRS11 or RDH13 or Angiopoietin-1 on biopsy tissues from lung cancer patients (left panel). Scale bar = 100 μm. The correlation between the expression of CCL28 and the levels of DHRS11, RDH13, and angiopoietin-1 (right panel). Data with error bars are shown as mean ± SEM. Each symbol represents data from a replicate. Each panel is a representative experiment of at least three independent biological replicates. *, **, *** represent p < 0.05, p < 0.01 and p < 0.001, respectively. Abbreviation: MFI, Mean fluorescence intensity

Both CCL28 and retinoic acid could promote vascular normalization in vivo

Further, we investigated the role of CCL28 in tumor growth and vascular normalization in immunocompetent mice. We overexpressed CCL28 in LLC cells, which was validated by ELISA. Compared to the control group, CCL28 overexpression showed a significant increase in CCL28 concentration in both the cell supernatants and serum in mice implanted with tumors (Supplementary Fig. 4A, 4B). As expected, in vivo studies indicated that CCL28 could promote tumor growth in Lewis lung adenocarcinoma (LLC) (Fig. 5 A). Compared to the control group, NG2 + pericytes accumulation and angiogenesis enhanced in the LLC-CCL28 group (Fig. 5 B).

To more clearly explore the function of CCL28 on stromal cells and immune cells in the tumor microenvironment, we performed single-cell sequencing analysis on tumor tissues of LLC-NC and LLC-CCL28. Single-cell sequencing analysis provided a panoramic study of the tumor microenvironment, and the cell population in the tumor microenvironment underwent significant alterations. In total, 6380 cells were obtained after quality filtering in the 2 conditions. Specifically, there were 3829 cells in the CCL28 overexpression group, and 2551 cells in the control group. The t-SNE diagrams showed that these cells were divided into 19 cell clusters with a total of 10 cell types (Fig. 5 C and Supplementary Fig. 4D, 4E). Of all the cell types, three types showed the most significant changes in cell proportions: fibroblasts, pericytes, and macrophages, with fold changes of 2.39, 1.95, and 1.45, respectively. The percentage of pericytes was 1.09% in LLC-NC and rose to the rate of 2.14% in LLC-CCL28 (Fig. 5 D). In the CCL28 overexpressing group, we identified 82 pericytes, whereas the control group had 28 pericytes. This significant difference suggested that CCL28 may play a pivotal role in influencing pericyte populations within the tumor microenvironment. Moreover, a previous study has reported that CCL28 could recruit Tregs and promote angiogenesis in ovarian carcinoma [ 17 ]. However, the percentage of T cells was reduced in the CCL28 up-regulated LLC, indicating that CCL28 might play different roles in different cancer types. Moreover, the number of DC cells remains unchanged, but their proportion decreased (Fig. 5 D).

Subsequently, we focused on the analysis of pericytes. Cluster 14, which was CSPG4 + and ACTA2 + , was recognized as pericytes. The marker genes t-SNE plot for Cluster 14 was presented in Supplementary Fig. 5. Interestingly, metabolic pathway enrichment analysis revealed that the retinol metabolism was activated in 13 clusters, including DC, macrophages, fibroblasts, and pericytes, suggesting that the synthesis and metabolism of retinoic acid play essential physiological roles in tumor microenvironment (Fig. 5 E). In the cellular communication network, pericytes interact closely with various cell types. The heat map indicated that pericytes exhibited the strongest interaction with themselves, followed by cluster 3 and cluster 6, representing tumor and endothelial cells. Cluster 3 was characterized as Ki-67-positive and Top2a-positive, and cluster 6 was positive for Pecam (Fig. 5 F). Our results in Fig. 3 M also suggested CCL28 can promote retinoic acid accumulation in pericytes in vitro. We further investigated the effects of retinoic acid by daily gavage on tumor growth and vascular normalization in vivo. Knocking out CCL28 (LLC-CCL28-KO) or supplementing retinoic acid (LLC-NC + RA) could suppress tumor growth in LLC models (Fig. 5 G and Supplementary Fig. 4C). However, the suppression effect is more pronounced after knocking out CCL28 than the control group (LLC-NC), and CCL28 knockout with supplementing RA could further synergistically inhibit tumor growth (Fig. 5 G). Multiplex immunofluorescence was used to detect the normalization of tumor blood vessels. We found that compared to the control group, knocking out CCL28 resulted in a significant reduction in the proportion of NG2 + cells and decreased coverage of pericytes. As expected, supplementation of retinoic acid was able to promote a certain degree of restoration of pericytes coverage both in wild-type and CCL28 knockout tumors (Fig. 5 H). Then, we quantified the vascular density in different groups and found that, compared to the control group, CCL28 knockout resulted in significantly decreased vascular density, while retinoic acid had only a slight effect (Fig. 5 H). Additionally, we evaluated the effect of retinoic acid on tumor cell viability using a CCK8 assay (Supplementary Fig. 4F). The results indicated that retinoic acid did not significantly affect tumor cell viability at pharmacological concentrations (1–2 μg/ml, equivalent to 3.33–6.66 μM). However, cytotoxic effects were observed at higher concentrations, which exceed pharmacologically relevant levels in the body necessary to support essential cellular functions. To confirm the influence on tumor hypoxia, we immunofluorescently labeled mouse tumor tissues using CA9 (Carbonic Anhydrase 9), which is considered an indicator of hypoxia. Our findings indicated that compared to the control group, tumor hypoxia significantly increased in tumors with CCL28 knockout. However, retinoic acid does not appear to have a significant effect on hypoxia (Supplementary Fig. 4G). These results indicated that RA inhibited tumor growth and enhanced vascular normalization, rather than angiogenesis.

The above data indicate that CCL28-regulated retinoic acid plays a crucial role in vascular normalization, suggesting its potential as a therapeutic agent in anti-angiogenic tumor treatment. To investigate whether the combination of retinoic acid and bevacizumab has a synergistic effect, we established an A549 mouse model and administered intravenous injections of bevacizumab, coupled with oral gavage of retinoic acid. We observed that retinoic acid and bevacizumab could attenuate tumor growth in mice, respectively. Moreover, simultaneous administration of retinoic acid and bevacizumab had a synergistic effect on tumor growth (Fig. 5 I). These data indicate that retinoic acid can be used in combination with bevacizumab for tumor treatment, providing a new direction for clinical combination therapy.

CCL28 is involved in bevacizumab-mediated vascular normalization

Moreover, we investigated the synergistic effects of CCL28 knock-out and VEGF blocking on the tumor growth and vascular normalization of lung adenocarcinoma. Subcutaneously implanted lung adenocarcinoma cells with CCL28 knock-out grew much slower than wild-type tumor cells, while combination of VEGF blocker could stop the growth of the tumors (Fig. 6 A). In addition, significant promotion of vascular normalization is observed after bevacizumab treatment. However, this effect disappeared after knocking out CCL28, regardless of whether bevacizumab was added (Fig. 6 B). Knocking out CCL28 could inhibit tumor angiogenesis, and the effect was more pronounced when bevacizumab was combined (Fig. 6 B). These results indicated that CCL28 could participate in bevacizumab-mediated vascular normalization, and CCL28 might be a potential target for anti-angiogenesis therapy in lung adenocarcinoma.

Further, we evaluated the expression levels of CCL28 within the tumor tissue (Fig. 6 C, upper panel). The mean fluorescence intensity of CCL28 was significantly upregulated when bevacizumab was used compared to the control group (Fig. 6 C, lower panel), consistent with the results we found in the clinical samples. Also, the expression level of CCL28 in lung adenocarcinoma correlates with therapeutic efficacy. In lung adenocarcinoma patients, based on their treatment outcomes with bevacizumab, they were categorized into two groups: poor responders and good responders. PAN-CK, NG2, CD31, and CCL28 were stained in tumor tissues to analyze their expression levels with the therapeutic efficacy of bevacizumab treatment. Patients who responded well to bevacizumab showed significantly increased CCL28 expression, and high expression of CCL28 in tumor cells,, was associated with enhanced vascular maturity and favorable treatment outcomes (Fig. 6 D, 6E). The conclusion was evidenced by a notable decrease in malignant pericardial and pleural effusion and a significant reduction in metastatic lymph nodes after bevacizumab-based treatment (Fig. 6 F). Then, we analyzed the expression levels of DHRS11 and RDH13 in two groups of patients. Consistently, the therapeutic efficacy of bevacizumab appeared to be positively correlated with RDH13 expression and negatively correlated with DHRS11 expression (Fig. 6 G, H). These results suggest that the favorable response to anti-angiogenic therapy in patients may be attributed to the activation of CCL28, thereby deepening our understanding of treatment response variations.

Anti-angiogenesis therapy stands as a pivotal approach for metastatic lung adenocarcinoma [ 3 , 27 ], targeting the well-established VEGF/VEGFR pathway with numerous drugs developed over the last four decades [ 28 , 29 ]. However, its clinical impact has proven more complicated than initially anticipated. Recognizing the vasculature normalization effects of anti-angiogenic drugs has led to their integration into combination regimens with chemotherapy, radiotherapy, immunotherapy, and targeted therapy [ 4 ]. For example, bevacizumab, through the modulation of angiogenesis and improvement of the tumor microenvironment, lowers tumor vascular density and edema, thereby enhancing the efficiency of oxygen supply and making other treatment modalities more effective. Consequently, anti-angiogenesis drugs are employed as modulators for tumor vasculature and the microenvironment within combination regimens.

Anti-angiogenic therapy inhibits abnormal blood vessel formation, fostering vascular normalization through mechanisms like reducing vessel density, remodeling the extracellular matrix, regulating inflammation, balancing growth factors, and modulating the immune system. The tipped balance between pro-angiogenesis and anti-angiogenesis factors contributes to maintaining tumor vascular growth [ 30 ]. Anti-angiogenic therapy disrupts this tipped balance by inhibiting VEGF, enhancing the relative action of angiopoietin-1(ANGPT1), which aids in regulating and inducing vascular normalization. This rebalancing contributes to vascular normalization, ensuring newly formed blood vessels exhibit a more organized structure and function more normally. The action of angiopoietin-1 helps consolidate and stabilize the newly formed vessels, making them a more effective transportation system.

However, it is also widely accepted that agents targeting VEGF/VEGFR may ultimately elevate hypoxia levels within tumors [ 4 ]. This hypoxic microenvironment triggers alternative pro-angiogenesis molecular pathways in tumor or stromal cells within the tumor microenvironment, including FGF-2, HGF, DLL4/Notch, CCL28, and others [ 19 , 21 , 22 , 31 ]. Furthermore, hypoxia prompts a cascade of biological responses, leading tumor cells to adapt their metabolic pathways to low-oxygen conditions. The dysregulated metabolism observed in rapidly proliferating tumor cells is a hallmark of malignancy [ 32 ], contributing to the activation of various metabolism-related genes, including several hypoxia-related transcription factors like hypoxia-inducible factor (HIF), nuclear factor kappa-B, CREB, AP-1, p53, Sp1/3, Egr-1, and CEBPB [ 33 ].

In this study, we identified CCL28 as another crucial molecular target for vascular normalization in lung adenocarcinoma. It was observed that CCL28 could be up-regulated under hypoxic conditions. However, the molecular mechanism underlying hypoxia-induced transcriptional activation of the CCL28 gene remains unclear. Our investigation revealed that the transcription factors CEBEPB could regulate the expression of CCL28. CCL28, belonging to the subfamily of small cytokine CC chemokines, binds to chemokine receptors CCR3 and CCR10 [ 19 ]. It has been reported that CCL28 exhibits chemotactic activity for various immune cells and plays a role in the physiology of extracutaneous epithelial tissues [ 34 ]. Several studies have delved into the functions of CCL28 in the tumor microenvironment [ 17 , 18 , 20 ]. Therefore, we hypothesized that CCL28 might play a pivotal role in modulating the tumor microenvironment in lung adenocarcinoma. Here, we noted an augmentation in pericyte accumulation within the tumor microenvironment associated with the overexpression of CCL28. However, the mechanisms of pericyte recruitment in lung adenocarcinoma remain unknown. Thus, we examined the expression of CCR3/CCR10 on various tumor stromal cells, revealing widespread expression of CCR3 on endothelial cells, cancer-associated fibroblasts, and pericytes in lung adenocarcinoma. Two chemotactic experiments substantiated the effects of CCL28 on recruiting pericytes through the receptor CCR3.

Abnormal differentiation of stromal cells is an essential characteristic of malignant tumors and is correlated with abnormal metabolism. Many tumor stromal cells were differentiated from anti-tumor type to pro-tumor type, such as tumor-associated macrophage, cancer-associated fibroblasts, tumor-associated neutrophils, etc. [ 35 , 36 , 37 ]. Reprogramming the metabolism and modulating tumor stromal cell differentiation is a promising cancer treatment strategy. Interestingly, the present study found that retinoic acid metabolism was activated in a series of stromal cells in lung cancer.

Importantly, chemotactic factors play a crucial role in cell migration, influencing not only the cell's movement [ 38 ] but also closely interacting with cellular metabolism. They regulate energy production, metabolic pathways, and antioxidant responses, ensuring that cells have sufficient energy and resources during migration. Chemotactic factors influence cellular metabolism by regulating intracellular signaling pathways, such as PI3K/AKT, MAPK, AMPK, and mTOR [ 39 , 40 ]. The activation or inhibition of these signaling pathways can modulate the activity of intracellular metabolic enzymes, affecting processes such as glucose metabolism, lipid synthesis, and amino acid utilization. In the present study, we found chemokine CCL28 could influence the cellular levels of retinoic acid by modulating its synthesis.

Retinoic acid (RA), also known as vitamin A acid, and its related analogs participate in regulating the gene networks involved in cell growth, differentiation, homeostasis, and apoptosis. RA is an active metabolite of retinol. Retinol (Vitamin A) is a fat-soluble essential micronutrient that plays a crucial role in embryonic development, organ formation, immune system function, and vision [ 26 ]. Retinol can be metabolized into retinal by the retinol dehydrogenases (RDHs), and retinal can also be metabolized into retinol by the dehydrogenase/reductase SDR family (DHRS) [ 41 ]. Retinal can be further irreversibly metabolized to RA by the retinaldehyde dehydrogenases (ALDHs). The synthesis of RA depends on the balance between RDHs and DHRS. The present study found that CCL28 could disturb the balance between RDH13 and DHRS11. After being treated with CCL28, RDH13 in pericytes was significantly upregulated in a dose-dependent manner. However, the effect curve of CCL28 on DHRS11 expression in pericytes is not linear and might be bell-shaped, like the effects of VEGFA on endothelial cells. In addition, there might be an alternative pathway to maintain the balance of DHRS11 and RDH13 in pericytes. RA binds to RARα, promoting the formation of a heterodimer with retinoid X receptor alpha (RARα/RXRα) in the cell nucleus. This heterodimer binds to retinoic acid response elements (RAREs) in the promoter region of target genes [ 42 ], including ANGPT1 .

The angiopoietin family, including ANGPT1, ANGPT2, ANGPT3, and ANGPT4, is crucial in vascular development and normalization. They regulate endothelial cells' survival, proliferation, and migration by interacting with Tie receptors, thus vital to vascular normalization. ANGPT1 can activate the TIE2 receptor on endothelial cells, maintaining endothelial cell stability, enhancing tight connections between endothelial cells, and reducing microvascular permeability through a series of signaling pathways [ 26 ]. These signaling pathways include tyrosine kinase-related protein DOKR (also known as DOK2), endothelial nitric oxide synthase (eNOS), SH2 domain-containing phosphatase (SHP2), growth factor receptor-binding protein 2 (GRB2), and PI3K-Akt [ 24 ]. Various regulatory mechanisms influence the expression of the ANGPT1 . These include the hypoxia-inducible factor-1α (HIF-1α) signaling pathway under low oxygen conditions, interaction with vascular endothelial growth factor (VEGF), involvement of anti-inflammatory factors and growth factors, cell–cell interactions, as well as the nuclear factor-κB (NF-κB) signaling pathway and hormonal regulation. In the present study, we identified another regulatory mechanism controlling the expression of ANGPT1.

Vitamin A and its metabolites, particularly retinoic acid, exert regulatory effects on the vascular system through various pathways [ 43 ]. These include maintaining endothelial cell function, regulating angiogenesis, inhibiting vascular smooth muscle cell proliferation, suppressing inflammatory responses, and promoting cell differentiation. This comprehensive regulatory process contributes to maintaining blood vessels' stable structure and function, playing a crucial role in embryonic development, tissue repair, angiogenesis, and vascular health during inflammation and diseases. Interestingly, vitamin A and its metabolites have been used as a differentiation modulator of malignant cells for cancer treatment. The present study proposed a new strategy to modulate the differentiation of cancer stromal by vitamin A and its metabolites.

Conclusions

In conclusion, we elucidated that a specific chemokine CCL28 induced after anti-angiogenesis therapy can alter tumor stromal cell metabolism and reshape the tumor microenvironment. In summary, we identified a mechanism through which CCL28 promotes vascular normalization via a CCR3-pericytes-RA-RXRα-ANGPT1-dependent pathway (Fig. 7 ). With an in-depth exploration of the interplay between chemokine and ANGPT1, our work may provide valuable insights into the regulatory network of vascular normalization, offering new avenues for modulating tumor microenvironment and overcoming resistance to anti-angiogenesis therapy in lung adenocarcinoma treatment.

A schematic diagram of tumor microenvironment modulation effects of CCL28

Availability of data and materials

The data that support the findings of this study are available from the corresponding author upon reasonable request.Competing interests.

The authors declare that they have no competing interests.

Abbreviations

Platelet-derived growth factor receptor β

Chondroitin sulfate proteoglycan 4

Human CC motif chemokine ligands 28

Human umbilical vein endothelial cell

Dehydrogenase/reductase 11

Retinol dehydrogenase 13

Bevacizumab

Zhang Y, Vaccarella S, Morgan E, Li M, Etxeberria J, Chokunonga E, et al. Global variations in lung cancer incidence by histological subtype in 2020: a population-based study. Lancet Oncol. 2023;24:1206–18.

Article PubMed Google Scholar

Xia C, Dong X, Li H, Cao M, Sun D, He S, et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin Med J (Engl). 2022;135:584–90.

Reck M, Popat S, Grohé C, Corral J, Novello S, Gottfried M, et al. Anti-angiogenic agents for NSCLC following first-line immunotherapy: rationale, recent updates, and future perspectives. Lung Cancer. 2023;179:107173.

Article PubMed CAS Google Scholar

Augustin HG, Koh GY. Antiangiogenesis: vessel regression, vessel normalization, or both? Cancer Res. 2022;82:15–7.

Zhang N, Yin R, Zhou P, Liu X, Fan P, Qian L, et al. DLL1 orchestrates CD8(+) T cells to induce long-term vascular normalization and tumor regression. Proc Natl Acad Sci U S A. 2021;118(22):e2020057118.

Huang G, Chen L. Recombinant human endostatin improves anti-tumor efficacy of paclitaxel by normalizing tumor vasculature in Lewis lung carcinoma. J Cancer Res Clin Oncol. 2010;136:1201–11.

Shi S, Chen L, Huang G. Antiangiogenic therapy improves the antitumor effect of adoptive cell immunotherapy by normalizing tumor vasculature. Med Oncol. 2013;30:698.

Shi S, Tao L, Song H, Chen L, Huang G. Synergistic antitumor effect of combining metronomic chemotherapy with adoptive cell immunotherapy in nude mice. APMIS. 2014;122:380–91.

Tao L, Huang G, Shi S, Chen L. Bevacizumab improves the antitumor efficacy of adoptive cytokine-induced killer cells therapy in non-small cell lung cancer models. Med Oncol. 2014;31:777.

Jain RK. Normalization of tumor vasculature: an emerging concept in antiangiogenic therapy. Science. 2005;307:58–62.

Shi S, Wang R, Chen Y, Song H, Chen L, Huang G. Combining antiangiogenic therapy with adoptive cell immunotherapy exerts better antitumor effects in non-small cell lung cancer models. PLoS One. 2013;8:e65757.

Article PubMed PubMed Central CAS Google Scholar

Ramjiawan RR, Griffioen AW, Duda DG. Anti-angiogenesis for cancer revisited: is there a role for combinations with immunotherapy? Angiogenesis. 2017;20:185–204.

Article PubMed PubMed Central Google Scholar

Hartmann DA, Berthiaume AA, Grant RI, Harrill SA, Koski T, Tieu T, et al. Brain capillary pericytes exert a substantial but slow influence on blood flow. Nat Neurosci. 2021;24:633–45.

Figueiredo AM, Villacampa P, Dieguez-Hurtado R, Jose Lozano J, Kobialka P, Cortazar AR, et al. Phosphoinositide 3-kinase-regulated pericyte maturation governs vascular remodeling. Circulation. 2020;142:688–704.

Zhang Y, Cedervall J, Hamidi A, Herre M, Viitaniemi K, D’Amico G, et al. Platelet-specific PDGFB ablation impairs tumor vessel integrity and promotes metastasis. Cancer Res. 2020;80:3345–58.

Hosaka K, Yang Y, Seki T, Nakamura M, Andersson P, Rouhi P, et al. Tumour PDGF-BB expression levels determine dual effects of anti-PDGF drugs on vascular remodelling and metastasis. Nat Commun. 2013;4:2129.

Facciabene A, Peng X, Hagemann IS, Balint K, Barchetti A, Wang LP, et al. Tumour hypoxia promotes tolerance and angiogenesis via CCL28 and T(reg) cells. Nature. 2011;475:226–30.

Ji L, Qian W, Gui L, Ji Z, Yin P, Lin GN, et al. Blockade of beta-catenin-induced CCL28 suppresses gastric cancer progression via inhibition of treg cell infiltration. Cancer Res. 2020;80:2004–16.

Huang G, Tao L, Shen S, Chen L. Hypoxia induced CCL28 promotes angiogenesis in lung adenocarcinoma by targeting CCR3 on endothelial cells. Sci Rep. 2016;6:27152.

Park J, Zhang X, Lee SK, Song N-Y, Son SH, Kim KR, et al. CCL28-induced RARβ expression inhibits oral squamous cell carcinoma bone invasion. J Clin Investig. 2019;129:5381–99.

Incio J, Ligibel JA, McManus DT, Suboj P, Jung K, Kawaguchi K, et al. Obesity promotes resistance to anti-VEGF therapy in breast cancer by up-regulating IL-6 and potentially FGF-2. Sci Transl Med. 2018;10:10.

Article Google Scholar

Cascone T, Xu L, Lin HY, Liu W, Tran HT, Liu Y, et al. The HGF/c-MET pathway is a driver and biomarker of VEGFR-inhibitor resistance and vascular remodeling in non-small cell lung cancer. Clin Cancer Res. 2017;23:5489–501.

Chen Z, Kim SJ, Essani AB, Volin MV, Vila OM, Swedler W, et al. Characterising the expression and function of CCL28 and its corresponding receptor, CCR10, in RA pathogenesis. Ann Rheum Dis. 2015;74:1898–906.

Huang H, Bhat A, Woodnutt G, Lappe R. Targeting the ANGPT-TIE2 pathway in malignancy. Nat Rev Cancer. 2010;10:575–85.

Teichert M, Milde L, Holm A, Stanicek L, Gengenbacher N, Savant S, et al. Pericyte-expressed Tie2 controls angiogenesis and vessel maturation. Nat Commun. 2017;8:16106.

Khan KA, Wu FT, Cruz-Munoz W, Kerbel RS. Ang2 inhibitors and Tie2 activators: potential therapeutics in perioperative treatment of early stage cancer. EMBO Mol Med. 2021;13:e08253.

Wang M, Herbst RS, Boshoff C. Toward personalized treatment approaches for non-small-cell lung cancer. Nat Med. 2021;27:1345–56.

Ellis LM, Hicklin DJ. VEGF-targeted therapy: mechanisms of anti-tumour activity. Nat Rev Cancer. 2008;8:579–91.

Geindreau M, Ghiringhelli F, Bruchard M. Vascular endothelial growth factor, a key modulator of the anti-tumor immune response. Int J Mol Sci. 2021;22:22.

Yang T, Xiao H, Liu X, Wang Z, Zhang Q, Wei N, et al. Vascular normalization: a new window opened for cancer therapies. Front Oncol. 2021;11:719836.

Li J-L, Sainson RCA, Oon CE, Turley H, Leek R, Sheldon H, et al. DLL4-Notch signaling mediates tumor resistance to anti-VEGF therapy in vivo. Can Res. 2011;71:6073–83.

Article CAS Google Scholar

Hanahan D. Hallmarks of cancer: new dimensions. Cancer Discov. 2022;12:31–46.

Cummins EP, Taylor CT. Hypoxia-responsive transcription factors. Pflugers Arch. 2005;450:363–71.

Mohan T, Deng L, Wang B-Z. CCL28 chemokine: an anchoring point bridging innate and adaptive immunity. Int Immunopharmacol. 2017;51:165–70.

Zhang X, Ji L, Li MO. Control of tumor-associated macrophage responses by nutrient acquisition and metabolism. Immunity. 2023;56:14–31.

Zhang H, Yue X, Chen Z, Liu C, Wu W, Zhang N, et al. Define cancer-associated fibroblasts (CAFs) in the tumor microenvironment: new opportunities in cancer immunotherapy and advances in clinical trials. Mol Cancer. 2023;22:159.

Que H, Fu Q, Lan T, Tian X, Wei X. Tumor-associated neutrophils and neutrophil-targeted cancer therapies. Biochim Biophys Acta Rev Cancer. 2022;1877:188762.

Stoll G, Pol J, Soumelis V, Zitvogel L, Kroemer G. Impact of chemotactic factors and receptors on the cancer immune infiltrate: a bioinformatics study revealing homogeneity and heterogeneity among patient cohorts. Oncoimmunology. 2018;7:e1484980.

Xu H, Lin S, Zhou Z, Li D, Zhang X, Yu M, et al. New genetic and epigenetic insights into the chemokine system: the latest discoveries aiding progression toward precision medicine. Cell Mol Immunol. 2023;20:739–76.

Lazennec G, Rajarathnam K, Richmond A. CXCR2 chemokine receptor - a master regulator in cancer and physiology. Trends Mol Med. 2024;30:37–55.

Adams MK, Belyaeva OV, Wu L, Kedishvili NY. The retinaldehyde reductase activity of DHRS3 is reciprocally activated by retinol dehydrogenase 10 to control retinoid homeostasis. J Biol Chem. 2014;289:14868–80.

Ghyselinck NB, Duester G. Retinoic acid signaling pathways. Development. 2019;146:146.

DiKun KM, Gudas LJ. Vitamin A and retinoid signaling in the kidneys. Pharmacol Ther. 2023;248:108481.

Download references

Acknowledgements

Not applicable.

This work was supported by the National Natural Science Foundation of China (Grant No. 82273326), Jiangsu Provincial Youth Medical Key Talents Project (Grant No. QNRC2016887), the Bethune Charitable Foundation (Grant No. KY202301-17) and Postgraduate Research & Practice Innovation Program of Jiangsu Province (Grant No. KYCX22_0176).

Author information

Authors and affiliations.

The State Key Laboratory of Pharmaceutical Biotechnology, Chemistry and Biomedicine Innovation Center (ChemBIC), Division of Immunology, Medical School, Nanjing University, Nanjing, 210093, China

Ying Chen, Zhiyong Zhang, Pengfei Li, Yuxi Chen & Tingting Wang

Jiangsu Key Laboratory of Molecular Medicine, Division of Immunology, Medical School, Nanjing University, Nanjing, 210093, China

Medical Schoolof, Nanjing University, Nanjing, Jiangsu, 210093, China

Ying Chen, Zhiyong Zhang, Fan Pan, Pengfei Li, Weiping Yao, Yuxi Chen & Tingting Wang

Department of Respiratory Critical Care Medicine, Affiliated Hospital of Medical School, Nanjing Drum Tower Hospital, Nanjing University, Nanjing, Jiangsu, 210008, China

Department of Cardio-Thoracic Surgery, Affiliated Hospital of Medical School, Jinling Hospital, Nanjing University, Nanjing, Jiangsu, 210008, China

Department of Medical Oncology, Affiliated Hospital of Medical School, Jinling Hospital, Nanjing University, Nanjing, Jiangsu, 210008, China

Fan Pan, Weiping Yao & Guichun Huang

Department of Oncology, Medical School, Zhongda Hospital, Southeast University, Nanjing, 210009, China

Guichun Huang

You can also search for this author in PubMed Google Scholar

Contributions

TTW, YL and GCH: Provided funding support, conceptualization, methodological guidance, and manuscript revision for this study.. YC: conceived of the study, performed the experiments and wrote the paper. WPY: participated in animal experiments. ZYZ and FP: performed the statistical analysis. PFL, YXC and LX: Participated in experiment design and sample collection. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Tingting Wang , Yan Li or Guichun Huang .

Ethics declarations

Ethics approval and consent to participate.

All animal experiments were carried out following the institutional guidelines and approved by the Ethical Review Committee of Comparative Medicine, Jinling Hospital, Nanjing, China. Written informed consent was obtained from all patients before collecting the samples. All the methods were carried out following the institutional guidelines and approved by the Ethical Review Committee of Jinling Hospital, Nanjing, China.

Consent for publication

All authors have consented for publication.

Competing interests

Additional information, publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., supplementary material 2., supplementary material 3., supplementary material 4., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Chen, Y., Zhang, Z., Pan, F. et al. Pericytes recruited by CCL28 promote vascular normalization after anti-angiogenesis therapy through RA/RXRA/ANGPT1 pathway in lung adenocarcinoma. J Exp Clin Cancer Res 43 , 210 (2024). https://doi.org/10.1186/s13046-024-03135-3

Download citation

Received : 21 March 2024

Accepted : 22 July 2024

Published : 29 July 2024

DOI : https://doi.org/10.1186/s13046-024-03135-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Anti-angiogenesis therapy
Vascular normalization
Lung adenocarcinoma
Retinoic acid

Journal of Experimental & Clinical Cancer Research

ISSN: 1756-9966

Submission enquiries: Access here and click Contact Us
General enquiries: [email protected]

Open access
Published: 31 July 2024

Genome assembly of redclaw crayfish ( Cherax quadricarinatus ) provides insights into its immune adaptation and hypoxia tolerance

Ziwei Liu 1 na1 ,
Jianbo Zheng 2 na1 ,
Haoyang Li 3 , 4 na1 ,
Ke Fang 1 ,
Sheng Wang 3 ,
Jian He 1 , 4 ,
Dandan Zhou 1 ,
Shaoping Weng 3 , 4 ,
Meili Chi 2 ,
Zhimin Gu 2 , 5 ,
Jianguo He 1 , 3 , 4 ,
Fei Li 2 &
Muhua Wang 1 , 3 , 4

BMC Genomics volume 25 , Article number: 746 ( 2024 ) Cite this article

238 Accesses

Metrics details

The introduction of non-native species is a primary driver of biodiversity loss in freshwater ecosystems. The redclaw crayfish ( Cherax quadricarinatus ) is a freshwater species that exhibits tolerance to hypoxic stresses, fluctuating temperatures, high ammonia concentration. These hardy physiological characteristics make C. quadricarinatus a popular aquaculture species and a potential invasive species that can negatively impact tropical and subtropical ecosystems. Investigating the genomic basis of environmental tolerances and immune adaptation in C. quadricarinatus will facilitate the development of management strategies of this potential invasive species.

We constructed a chromosome-level genome of C. quadricarinatus by integrating Nanopore and PacBio techniques. Comparative genomic analysis suggested that transposable elements and tandem repeats drove genome size evolution in decapod crustaceans. The expansion of nine immune-related gene families contributed to the disease resistance of C. quadricarinatus . Three hypoxia-related genes ( KDM3A , KDM5A , HMOX2 ) were identified as being subjected to positive selection in C. quadricarinatus . Additionally, in vivo analysis revealed that upregulating KDM5A was crucial for hypoxic response in C. quadricarinatus . Knockdown of KDM5A impaired hypoxia tolerance in this species.

Conclusions

Our results provide the genomic basis for hypoxic tolerance and immune adaptation in C. quadricarinatus , facilitating the management of this potential invasive species. Additionally, in vivo analysis in C. quadricarinatus suggests that the role of KDM5A in the hypoxic response of animals is complex.

Peer Review reports

Freshwater species face a consistently higher risk of extinction compared to their terrestrial and marine counterparts [ 1 ]. The introduction of non-native species affects ecosystem functioning and is a major driver of biodiversity loss in freshwater ecosystems [ 2 ]. Crayfish (Decapoda: Astacidea) are a diverse taxonomic group of freshwater crustaceans that play a key role in freshwater ecosystems. Crayfish have been consumed as food by human since prehistoric times, and have recreational, cultural, and scientific value [ 3 , 4 ]. Therefore, they have been transported by human across the globe. They are relatively large-bodied, long-lived, densely populated, and feed at multiple trophic levels, thus affecting entire trophic webs [ 5 ]. When introduced outside their native range, crayfish species are likely to establish self-reproducing populations, spread from the point of introduction and become invasive. However, our understanding of the mechanisms driving crayfish invasions remains limited, hindering the development of effective preventative strategies.

Redclaw crayfish ( Cherax quadricarinatus ), which is native to northern Australia and southern New Guinea, is the second most cultured and caught crayfish species [ 6 , 7 ]. Interest in redclaw crayfish for aquaculture has resulted in worldwide translocation of this species. The redclaw crayfish is resilient to hypoxic stress, tolerates a broad pH range, adapts to fluctuating temperatures, and withstands high levels of ammonia concentration [ 8 ]. The hardy physiological characteristics of the redclaw crayfish allow it to establish self-sustaining populations in the wild and negatively impact tropical and subtropical ecosystems [ 9 , 10 ]. Therefore, C. quadricarinatus is considered as a potential invasive species. Additionally, C. quadricarinatus exhibits resistance to certain pathogens. It is reported that C. quadricarinatus is resistant to the acute hepatopancreatic necrosis disease (AHPND), which is caused by specific Vibro spp infections [ 11 ]. Infections by the bacilliform virus have been found in both wild and cultured C. quadricarinatus , but these do not cause disease or mortalities [ 12 ]. Experimental infection of C. quadricarinatus with Macrobrachium rosenbergii nodavirus showed that C. quadricarinatus has low susceptibility to the virus [ 13 ]. The genome sequence of C. quadricarinatus has been reported in previous studies [ 14 , 15 ], while the genomic basis of its adaptation remains largely unknown. Investigating the genomic basis of its environmental tolerances and disease resistance will facilitate the development of management strategies for C. quadricarinatus .

It has been demonstrated that the changes of gene expression related to hypoxia is largely mediated by the activation of hypoxia-inducible factors (HIFs) [ 16 ]. Under normoxia, HIFα subunits are polyubiquitylated by the von Hippel–Lindau tumor suppressor protein (pVHL) complex and targeted for proteasomal degradation. Hypoxia inhibits the proteasomal degradation of HIFα, allowing it to bind with HIF1β and transcriptionally activate genes that promote adaptation to inadequate oxygen [ 17 ]. 2-oxoglutarate-dependent dioxygenases (2-OGDDs) are a family of over 60 enzymes that rely on oxygen for their activities [ 18 ]. 2-OGDDs with low oxygen affinity are major regulators of HIFs, thereby contributing to hypoxia-related transcriptional regulation [ 19 ]. Recent studies found that Jumonji C (JmJc) domain-containing histone lysine demethylases (KDM3A, KDM5A, KDM6A, KDM6B), members of the 2-OGDD family, act as direct sensors of hypoxia. The activities of these JmJc demethylases were altered under hypoxic conditions in human and mouse, leading to an increase to histone methylation and triggering the hypoxia-induced transcriptional changes [ 20 , 21 , 22 ]. However, the roles of KDMs in hypoxia-induced responses in invertebrates remain largely elusive.

Here, we assembled a chromosome-level genome of C. quadricarinatus by integrating Nanopore and PacBio techniques. Factors contributing to the evolution of genome sizes in decapod crustaceans were determined through comparative genomic analysis. The robust disease resistance of C. quadricarinatus was found to attributed to the expansion of immune-related gene families. Three hypoxia-related genes ( KDM3A , KDM5A , HMOX2 ) were identified as being subjected to positive selection in C. quadricarinatus . Furthermore, we investigated the role of KDM5A in the hypoxic response of C. quadricarinatus . Our results provided insights into the genomic basis of disease resistance and hypoxia tolerance in C. quadricarinatus .

Chromosome-level genome assembly of C. quadricarinatus

The genome of C. quadricarinatus were sequenced using a combination of Nanopore, PacBio, and Illumina shotgun sequencing. A total of 358 Gb of Nanopore reads and 159 Gb of Illumina reads were generated (Supplementary Table 1 and 2). In addition, 280 Gb of PacBio HiFi reads were generated for error correction (Supplementary Table 3). Based on the k -mer distribution of Illumina reads, the genome sizes of C. quadricarinatus were estimated to be 6.02 Gb (Supplementary Fig. 1). The C. quadricarinatus genome was assembled into contigs with Nanopore reads using Shasta and WTDBG2, respectively [ 23 , 24 ]. The resulted contigs were assembled into longer sequences using quickmerge [ 25 ], and scaffolded using proximity ligation data from the Hi-C libraries to yield genome assembly (Supplementary Fig. 2; Supplementary Table 4). The final genome assembly of C. quadricarinatus consisted of 7,344 scaffolds (contig N50: 739.45 kb, scaffold N50: 33.93 Mb) assembled into 100 pseudomolecules, resulting in a total assembly size of 3.954 Gb (Fig. 1 andTable 1 ). We compared the sequence consistency and integrity of our assembly with the previously published genome assembly of C. quadricarinatus (GCF_026875155.1) [ 14 ]. First, syntenic analysis showed high collinearity between the chromosomes of the previously published assembly and our assembly (Supplementary Fig. 3). Benchmarking Universal Single-Copy Orthologs (BUSCO) analysis indicated that 88.5% of conserved single-copy arthropod ( Arthropoda ) genes (odb10) were captured in our assembly, compared to 81.2% were captured in previously published assembly (Supplementary Table 5) [ 26 ]. Additionally, Merqury evaluation indicated that the consensus quality value (QV) of our assembly was 32.81, compared to 18.75 of previously published assembly, suggesting our assembly is of better quality (Supplementary Table 6) [ 27 ].

Genome assembly of C. quadricarinatus . Circos plot of the distribution of genomic elements in C. quadricarinatus . a GC content; b Density of tandem repeats; c Density of transposable elements; d Gene density

Repetitive DNA in the C. quadricarinatus genome is exceptionally abundant, represented 93.36% (3.69 Gb) of the assembly (Supplementary Table 7). Transposable elements (TEs) account for 67.03% (2.65 Gb) of the C. quadricarinatus genome assembly. Long interspersed nuclear elements (LINEs) were the largest class of annotated TEs, making up 30.42% of the genome. Long terminal repeat (LTR) retrotransposons, which were the second largest class of TEs, represented 659.41 Mb (16.68%) of the genome. The proportion of LINEs and LTR retrotransposons in the C. quadricarinatus genome was relatively higher than that in the genomes of other crustaceans, which is derived from a recent expansion of retrotransposons (Supplementary Fig. 4 and 5; Supplementary Table 8). Protein-coding genes in the genomes were identified through a combination of ab initio, homology-based, and RNA-seq-based prediction approaches. In total, 17,698 protein-coding genes were identified in the C. quadricarinatus genome. BUSCO analysis identified 894 (88.3%) complete conserved single-copy arthropod ( Arthropoda ) genes (odb10) in the predicted gene models of C. quadricarinatus (Supplementary Table 9). In total, 16,509 (93.28%) gene models in the C. quadricarinatus genome can be annotated in at least one database (NCBI non-redudant, InterPro, KEGG, and eggNOG) (Supplementary Table 10).

Transposable elements and tandem repeats drive the expansion of decapod crustacean genomes

Crustaceans are characterized by having some of the most variable genome sizes among animals. The diverse genome sizes are associated with the physiological and life-history traits of these species, contributing to their adaptation [ 28 ]. Increases in genome sizes can be driven by several processes, including whole-genome duplication (WGD), TE expansion, intron expansion, and tandem gene duplication [ 29 ]. Of these, TE expansion and WGD are considered to be the most important factors [ 30 ]. To investigate the genome size evolution of decapod crustaceans, we compared the TE content among 10 decapod species with chromosome-level genome assemblies ( Procambarus virginalis , Procambarus clarkii , Homarus americanus , Scylla paramamosain , Eriocheir sinensis , Macrobrachium nipponense , Penaeus monodon , Fenneropenaeus chinensis , Litopenaeus vannamei , C. quadricarinatus ). A positive correlation was identified between genome size and TE content in these species (Fig. 2 A and B). In addition to TE, the genome size of decapods is positively correlated with the content of tandem repeats (TRs) (Fig. 2 A and C). The content of minisatellite is significantly higher in P. clarkii , C. quadricarinatus , and H. americanus compared to other decapods ( P < 0.01), while M. nipponense , P. monodon , F. chinensis , and L. vannamei have higher content of simple sequence repeats (SSRs) than satellites and minisatellites (Fig. 2 A). A previous study found the content of SSRs was significantly higher in penaeid shrimp species than in other decapods [ 31 ]. Our results suggested that TE and TR expansions drove the expansion of the genomes of decapod species.

The evolution of genome sizes in decapod crustaceans. A The comparison of genome sizes, transposable element (TE) content, and tandem repeat (TR) content among decapod species. B The correlation between TE content and genome sizes in decapod species. C The correlation between TR content and genome sizes in decapod species. D Frequency distribution of the synonymous substitution rates ( K s) among syntenic gene pairs of decapod species

To investigate the contribution of WGD to the genome size evolution of decapods, the synonymous substitution rates ( K s) of syntenic gene pairs were estimated in C. quadricarinatus and 8 other decapod crustaceans (Fig. 2 D). No obvious peak was identified in the K s distributions of all 9 species. Additionally, limited WGD-derived duplicated gene pairs were identified in the genomes of these 9 species using DupGen_finder [ 32 ] (Supplementary Fig. 6; Supplementary Table 11). These results indicated that these 9 decapod crustaceans have not undergone WGD during evolution.

Gene families related to immunity expanded in the genome of C. quadricarinatus

Gene family expansion and contraction analyses were performed to dissect the genetic basis of adaptation in C. quadricarinatus . A maximum-likelihood (ML) phylogenetic tree of C. quadricarinatus and 12 arthropods was reconstructed with Drosophila melanogaster as the outgroup (Fig. 3 A). Cherax quadricarinatus formed a clade with P. clarkii and P. virginalis . And H. americanus appeared sister to this clade. The ancestor of C. quadricarinatus diverged from the ancestors of P. clarkii and P. virginalis approximately 176 million years (Ma) ago (CI: 149.89–202.20 Ma). The divergence time of H. americanus and other three crayfish species was estimated to be approximately 214 Ma (CI: 183.85–245.06 Ma), consistent with the results of previous studies [ 33 , 34 ].

The evolution of gene families in decapod crustaceans. A A species tree of 13 arthropod species with D. melanogaster as outgroup. B Gene family expansion and contraction analysis of decapod crustaceans. Numbers of expanded gene families are marked in green, and numbers of contracted gene families are marked in red. The number below the MRCA (most recent common ancestor) represents the total number of orthologs from OrthoMCL analysis used as input for CAFE expansion/contraction analysis. C The comparison of gene numbers within the nine expanded immune-related gene families among arthropod species. D A phylogenetic tree of IgLectin Proteins from C. quadricarinatus , P. clarkii , S. paramamosain , E. sinensis . E Genomic distribution of IgLectin genes in C. quadricarinatus

Gene family analysis was performed based on the phylogenetic tree (Fig. 3 B). Compared with other arthropods, 146 gene families were expanded, and 147 gene families were contracted in C. quadricarinatus ( P < 0.01). Interestingly, nine gene families related to immunity were significantly expanded in the genome of C. quadricarinatus (Fig. 3 C). The invertebrate immune system primarily relies on specific pattern recognition receptors (PRRs) to recognize the pathogen-associated molecular patterns (PAMPs) of invading pathogens [ 35 ]. PPR-PAMP interaction initiate a series of immune responses, including prophenoloxidase (proPO) activated system, antimicrobial peptides (AMP) synthesis, phagocytosis, encapsulation, blood clotting, and reactive oxygen species production [ 36 ]. Within the nine expanded gene families related to immunity in C. quadricarinatus , four encode PPRs (Ig domain-containing C-type lectin, ficolin, scavenger receptor cysteine-rich domain containing protein, Down syndrome cell adhesion molecule), two encode AMPs (anti-lipopolysaccharide factor, c-type lysozyme), and three encode other immune-related proteins (Laccase-1, Peritrophin-1, Phenoloxidase-activating factor 2) [ 37 ]. The expansion of these immune-related gene families in C. quadricarinatus confers a strong ability of pathogen recognition and clearance, thereby contributing to the strong disease resistance in this species.

The C-type lectin family, characterized by its signature C-type lectin-like domain, promotes antibacterial host defense across numerous animal species [ 38 ]. While most invertebrate C-type lectins contain a single carbohydrate-recognition domain, certain arthropods were found to have C-type lectins possessing additional functional domains [ 39 , 40 ]. Vertebrate adaptive immunity primarily relies on immunoglobulins (Igs) belonging to the Ig superfamily [ 41 ]. Several proteins containing Ig-like domains play a crucial role in the innate immune response of invertebrates [ 42 ]. Recent studies identified a new C-type lectin protein possessing an Ig-like domain and a C-type lectin domain in E. sinensis and P. clarkii [ 43 , 44 ]. It triggers strong antibacterial activities through regulating phagocytosis in hemocytes and maintaining microbiota homeostasis in the intestine. To study the evolution of the Ig domain-containing C-type lectin (IgLectin), we identified the genes encoding this special C-type lectin in the major groups of vertebrates and invertebrates. IgLectin genes were only identified in four decapod species ( C. quadricarinatus , P. clarkii , S. paramamosain , E. sinensis ) (Fig. 3 D). While other decapods possess only one or two IgLectin genes, seven IgLectin genes formed a gene cluster on chromosome 10 of C. quadricarinatus (Fig. 3 E). This result indicated that the family of IgLectin genes was expanded in the genome of C. quadricarinatus . The expansion of IgLectin genes, which are exclusively found in decapods, may played a crucial role in disease resistance in C. quadricarinatus .

KDM5A is crucial for the hypoxia tolerance of C. quadricarinatus

Positively selected genes (PSGs) were identified in the genome of C. quadricarinatus to investigate the genetic basis of stress resistance within this species. In total, 27 PSGs were identified in the C. quadricarinatus genome compared to twelve arthropod species ( P. virginalis , P. clarkii , H. americanus , S. paramamosain , E. sinensis , M. nipponense , P. monodon , F. chinensis , L. vannamei , H. azteca , D. magna , D. melanogaster ) (Supplementary Table 12). Intriguingly, three genes ( KDM3A ; KDM5A ; Heme oxygenases-2, HMOX2 ) related to hypoxic response were positively selected in C. quadricarinatus .

KDM5A, a member of the JmjC domain-containing histone demethylase family, serves as a direct sensor of hypoxia. Hypoxia inactivates KDM5A in cancer cells, inhibiting the removal of a methyl group from H3K4me3 in the promoters of hypoxia-inducible genes. This leads to the activation of these genes and the induction of a hypoxic response [ 21 ]. As CqKDM5A has the highest omega score among the three identified positively selected hypoxia-related genes, we examined its function in the hypoxia response of C. quadricarinatus . Tissue distribution of CqKDM5A expression in C. quadricarinatus was examined using quantitative PCR (qPCR). The expression of CqKDM5A was detected in the eyestalk, intestine, muscle, hemocyte, hepatopancreas, stomach, and heart of C. quadricarinatus , but not in the gill and epidermis (Supplementary Fig. 7). To study whether Cq KDM5A involved in the hypoxic response of C. quadricarinatus , we examined CqKDM5A expression in the hemocyte of C. quadricarinatus reared under hypoxic and normoxic conditions using quantitative PCR (qPCR) (Fig. 4 A). The expression of CqKDM5A was significantly upregulated in the hemocyte of C. quadricarinatus after 12 h of hypoxia exposure. To further investigate the role of CqKDM5A in the hypoxic response, we used RNA interference (RNAi) to suppress CqKDM5A expression in C. quadricarinatus (Fig. 4 A). RNAi-treated and control groups of C. quadricarinatus were maintained at hypoxic and normoxic conditions. After 120 h of hypoxic exposure, the survival rate in the RNAi-treated group was significantly lower compared to the control groups (χ 2 : 16.37, P < 0.0001) (Fig. 4 B). This result indicated that suppressing CqKDM5A impaired the hypoxia tolerance of C. quadricarinatus . Furthermore, hypoxic tolerance was assessed in both adult and juvenile C. quadricarinatus . The expression of CqKDM5A was significantly higher in adult C. quadricarinatus than that in juveniles (Fig. 4 C). The survival rate of adult crayfishes was significantly higher than that of juveniles after 120 h of hypoxic exposure (χ 2 : 21.20, P < 0.0001) (Fig. 4 D). Taken together, these results suggested that upregulating CqKDM5A plays a critical role in the hypoxic tolerance of C. quadricarinatus .

The role of KDM5A in the hypoxic tolerance of C. quadricarinatus and L. vannamei . A RNA interference (RNAi) of the CqKDM5A gene in C. quadricarinatus . B Survival of wild-type, KDM5A-silenced, and GFP dsRNA treated C. quadricarinatus reared under normoxic and hypoxic conditions. C Quantification of the expression levels of CqKDM5A in adult and juvenile C. quadricarinatus . D Survival of adult and juvenile C. quadricarinatus when exposed to hypoxic stresses

In this study, we generated a chromosome-level genome sequence of C. quadricarinatus , which has a large genome size of over 6 Gb. Our final assembly is 3.95 Gb, approximately 2.07 Gb smaller than the estimated genome size. The previously published assembly (GCF_026875155.1) is 5.26 Gb [ 14 ], about 0.76 Gb smaller than the estimated genome size and 1.31 Gb larger than our assembly. However, BUSCO evaluation suggested that the completeness of our assembly is higher than the previously published assembly, and the quality value of our assembly assessed by Mercury software is much higher. This suggests that future studies are needed to generate a C. quadricarinatus genome assembly that includes the sequences that are missing from both current assemblies. In this study, we investigated the genome size evolution of 10 decapod crustaceans. We found that the genome size of these decapods is positively correlated with the content of both TEs and TRs, suggesting that the expansion of TEs and TRs contributes to the genome size expansion of decapod species. A previous study found the content of SSRs was significantly higher in penaeid shrimp species than in other decapods [ 31 ]. Consistent with this result, we found that three penaeid shrimp species ( P. monodon , F. chinensis , and L. vannamei ) and M. nipponense have a higher content of SSRs than satellites and minisatellites. In addition, we found that the content of minisatellite is significantly higher in P. clarkii , C. quadricarinatus , and H. americanus compared to other decapods. This suggests that minisatellites contribute to the genome evolution of certain decapod crustaceans.

Freshwater systems cover approximately 0.8% of the Earth’s surface but host almost 6% of the Earth’s described species. Despite this richness, freshwater ecosystems are currently the most endangered ecosystems. The decline in biodiversity within freshwater ecosystems surpasses that of other ecosystems [ 45 ]. The recent oxygen depletion in water resulted from rising global temperatures and anthropogenic eutrophication severely impacts the functioning and services of freshwater ecosystems [ 46 ]. Additionally, the increased connectivity of the global human population has amplified the frequency of biological invasion and pathogen transmission, potentially leading to the extinctions of freshwater species [ 47 ]. Crayfish are a diverse taxonomic group of freshwater crustaceans that includes both critically endangered endemic species and highly successful invasive species [ 48 ]. Due to their importance in aquaculture and popularity in the aquarium pet trade, numerous translocations of the Redclaw crayfish have occurred. The strong environmental tolerance and resistance to diseases of C. quadricarinatus allow it to establish self-sustaining populations in wild, negatively impacting the ecosystems it invades [ 8 ]. In this study, we investigated the genetic basis of disease resistance and hypoxia tolerance in C. quadricarinatus by generating a chromosome-level reference sequence. The strong disease resistance of C. quadricarinatus was found to be attributed to the expansion of nine immune-related gene families. Additionally, three genes critical for hypoxic response were identified as being subjected to positive selection in C. quadricarinatus , contributing to the hypoxia resistance in this species. These results elucidated the genetic basis underlying the invasive potential of C. quadricarinatus , thereby facilitating the development of preventative strategies to control its spread and mitigate its ecological impacts. Additionally, our results shed light on the protection of freshwater crayfish species that are susceptible to challenging environments.

Crayfish exhibit remarkable adaption to low dissolved oxygen conditions in water. Specifically, C. quadricarinatus have evolved adaptive molecular response to hypoxia, capable of surviving in low dissolved oxygen concentrations below 1 mg/L [ 8 ]. Hypoxia influences the activity of histone-modifying enzymes, which modulate the posttranslational modification of histones as well as nonhistone proteins [ 49 ]. A recent study shows that chromatin can sense oxygen independently of HIF through KDM5A [ 21 ]. The demethylase activity of KDM5A is suppressed in human cultured cells exposed to hypoxic conditions, leading to a global increase in the levels of H3K4me3 and elevated expression levels of hypoxia-inducible genes. Knockdown of KDM5A results in increased H3K4me3 and expression levels of the STAG2 and LOX genes, mirroring the hypoxia-induced cellular responses. Here, we performed in vivo analysis to investigate the role of KDM5A in the hypoxic response of C. quadricarinatus . Hypoxia upregulated KDM5A expression in C. quadricarinatus . And siRNA mediated silencing of KDM5A impaired the hypoxic response of this decapod species. These results suggested that while suppressing KDM5A expression triggers hypoxic response in human cultured cells, increasing KDM5A expression is essential for hypoxia tolerance in C. quadricarinatus . Due to their aquatic habitats, crustaceans encounter hypoxic conditions more frequently than mammals. Consequently, the mechanisms of hypoxic response may be different between these two animal groups. Our results elucidated the role of KDM5A in the hypoxia tolerance of C. quadricarinatus , highlighting the complexity of this HIF-independent hypoxic response mechanism.

In conclusion, this study found that the expansion of nine immune-related gene families contributed to the strong disease resistance of C. quadricarinatus . Furthermore, three genes crucial for hypoxic response ( KDM3A , KDM5A , HMOX2 ) were found to be subjected to positive selection in C. quadricarinatus . In vivo analyses revealed that upregulating KDM5A expression plays a crucial role in the hypoxic response of C. quadricarinatus . Our results provided the genetic basis of developing management strategies for C. quadricarinatus , a species with invasive potential.

Sampling and genome sequencing

One female C. quadricarinatus individual collected from the experimental pond of Zhejiang Institute of Freshwater Fisheries in Zhejiang Province, China was used for genome sequencing. High-quality DNA was extracted from muscle cells of C. quadricarinatus using DNeasy Blood & Tissue Kits (Qiagen) in accordance with the manufacturer’s protocol. DNA Quality and quantity were measured via standard agarose-gel electrophoresis and a Qubit 3.0 Fluorometer (Invitrogen), respectively. Nanopore sequencing libraries of C. quadricarinatus were constructed and sequenced using the Nanopore PromethION platform (Oxford Nanopore Technologies) (90X raw-read coverage). For Illumina sequencing, short-insert paired-end (PE) (150 bp) DNA libraries of C. quadricarinatus were constructed in accordance with the manufacturer’s instruction. Sequencing of PE libraries were performed (2 $\times$ 150 bp) on the Illumina NovaSeq 6000 platform (Illumina).

PacBio HiFi reads were generated to correct errors in the draft genome assembly. To construct PacBio HiFi sequencing libraries, high-quality DNA was extracted from muscle cells of C. quadricarinatus using DNeasy Blood & Tissue Kits (Qiagen) in accordance with the manufacturer’s protocol. The extracted genomic DNA was sheared by g-TUBEs (Covaris) according to the expected size of the fragments for library construction. Sheared DNA fragments were damage repaired, end repaired, and ligated with the hairpin adaptors for PacBio sequencing. The sequencing libraries were size selected by the BluePippin system (Sage Science) and purified by AMPure PB beads (Pacific Bioscience). Quality of the sequencing libraries were measured by the Agilent 2100 Bioanalyzer (Agilent technologies). Sequencing was performed on a PacBio Sequel II instrument (Pacific Bioscience).

Hi-C library was constructed based on a previously published procedure [ 50 ]. In brief, muscle sample of C. quadricarinatus were cut into 2 cm pieces and vacuum infiltrated in nuclei isolation buffer supplemented with 2% formaldehyde. Crosslinking was stopped by adding glycine and additional vacuum infiltration. Fixed tissue was frozen in liquid nitrogen and grounded to powder before re-suspending in nuclei isolation buffer to obtain a suspension of nuclei. The purified nuclei were digested with HindIII and marked by incubating with biotin-14-dCTP. The ligated DNA was sheared into 300 − 600 bp fragments, blunt-end repaired, A-tailed, and purified through biotin-streptavidin-mediated pull-down. Hi-C library was sequenced (2 $\times$ 150 bp) on the Illumina NovaSeq 6000 platform (Illumina).

Transcriptome sequencing

Eye, gill, heart, intestine, hepatopancreas, and muscle samples were collected from the C. quadricarinatus specimen to construct sequencing libraries for strand-specific RNA-sequencing (RNA-seq). Total RNA was extracted with TRIzol reagent (Invitrogen). Purity and integrity of RNA were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific) and Bioanalyzer 2100 system (Agilent), respectively. The mRNA was enriched from total RNA using poly-T oligo-attached magnetic beads. rRNA was removed using a TruSeq Stranded Total RNA Library Prep kit (Illumina). PE libraries were constructed using a VAHTSTM mRNA-seq V2 Library Prep Kit for Illumina (Vazyme) and sequenced (2 $\times$ 150 bp) using the Illumina NovaSeq 6000 platform (Illumina).

Genome assembly

Low-quality (≥ 10% unidentified nucleotide and/or ≥ 50% nucleotides with a Phred score < 5) and sequencing adapter-contaminated Illumina reads were filtered and trimmed with Fastp (v0.21.0) [ 51 ]. The resulting high-quality Illumina reads were used in the following analyses. The sizes and heterozygosity of C. quadricarinatus genomes were estimated using high-quality Illumina reads based on k -mer frequency-distribution. The number of k -mers and the peak depth of k -mer sizes at 17 was obtained using Jellyfish (v2.3.0) [ 52 ] with the - C setting. Genome size was estimated based on a previously described k -mer analysis [ 53 ]. The heterozygosity of C. quadricarinatus genome was determined by fitting the k -mer distribution of Arabidopsis thaliana using Kmerfreq implemented in SOAPdenovo2 (r242) [ 54 ].

Low-quality Nanopore reads were filtered using custom Python script. The filtered reads were then corrected using NextDenovo (v1.0) ( https://github.com/Nextomics/NextDenovo ). Two draft genome assemblies were generated using filtered and corrected Nanopore reads with Shasta (v0.8.0) [ 23 ] and WTDBG2 (v2.5) [ 24 ], respectively. The contigs of the two draft assemblies were subject to error correction using PacBio HiFi reads with Racon (v1.4.16) three times [ 55 ]. The corrected contigs were then polished with high-quality Illumina reads with Nextpolish (v1.2.4) three times [ 56 ]. Haplotypic duplications in the error-corrected contigs were identified and removed using purge_dups (v1.2.3) [ 57 ]. The resulted contigs were assembled into longer sequences using quickmerge (v0.3) [ 25 ].

We used Hi-C to correct misjoins, to order and orient contigs, and to merge overlaps. Low-quality Hi-C reads were filtered using fastp (v0.21.0) with default parameters [ 51 ]. Filtered Illumina reads were aligned to the assembled contigs using Bowtie2 (v2.3.2) [ 58 ]. Scaffolding was accomplished using LACHESIS with parameters “CLUSTER MIN RE SITES = 100, CLUSTER MAX LINK DENSITY = 2.5, CLUSTER NONINFORMATIVE RATIO = 1.4, ORDER MIN N RES IN TRUNK = 60” [ 59 ].

The completeness and quality of the assembly was evaluated using QUAST (v5.0.2), and Benchmarking Universal Single Copy Orthologs (BUSCO, v4.0.5) against the conserved Arthropoda dataset (obd10) [ 26 , 60 ]. Additionally, Merqury (v1.3) [ 27 ] was used to assess the completeness and quality of the four assemblies with k -mer set to 17.

Genome annotation

Repetitive elements in the assembly were identified by de novo predictions using RepeatMasker (v4.1.0) ( https://www.repeatmasker.org/ ). RepeatModeler (v2.0.1) was used to build the de novo repeat libraries of C. quadricarinatus [ 61 ]. To identify repetitive elements, sequences from the assembly were aligned to the de novo repeat library using RepeatMasker (v4.1.0). Additionally, repetitive elements in the C. quadricarinatus genome assembly were identified by homology searches against known repeat databases using RepeatMasker (v4.1.0). A repeat landscape of C. quadricarinatus genome was obtained using an R script that was modified from https://github.com/ValentinaBoP/TransposableElements .

Protein-coding genes in the C. quadricarinatus genome were predicted using RNA-seq-based approach. Short RNA-seq reads were aligned to genome assembly using HISAT2 (v2-2.1). Gene models were predicted based on the alignment results of HISAT2 using StringTie (v2.1.4) [ 62 ], and coding regions were identified using TransDecoder (v5.5.0) [ 63 ]. The completeness of predicted gene models was evaluated using BUSCO (v4.0.5) against the conserved Arthropoda dataset (odb10) [ 26 ]. To assign functions to the predicted proteins, we aligned the C. quadricarinatus protein models against NCBI nonredundant (NR) amino acid sequences and UniProt databases using BLASTP with an E-value cutoff of 10 −5 . Protein models were also aligned against the eggNOG database using eggNOG-Mapper [ 64 ], and against the InterPro database using InterProScan [ 65 ]. Finally, Kyoto Encyclopedia of Genes and Genomes (KEGG) annotation of the protein models was performed using BlastKOALA [ 66 ].

Tandem repeats identification

Tandem repeats in the genomes of D. melanogaster , D. magna , H. azteca , L. vannamei , F. chinensis , P. monodon , M. nipponense , E. sinensis , S. paramamosain , H. americanus , P. clarkii , P. virginalis , and C. quadricarinatus were identified using tandem repeat finder (TRF) (v4.09.1) [ 67 ]. In addition to TRF, we used GMATA (v2.3) and SciRoKo (v3.3) to identify SSRs in genomes of these arthropod species [ 68 , 69 ]. TRF-, GMATA- and SciRoKo-identified SSRs were integrated into a nonredundant set using a custom Python script.

Whole genome duplication and duplicated gene analysis

To determine whether recent WGD occurred in C. quadricarinatus as well as 8 decapods with chromosome-level genome assemblies ( P. clarkii , H. americanus , S. paramamosain , E. sinensis , M. nipponense , P. monodon , F. chinensis , and L. vannamei ), we estimated the distribution of K s for each paralog in using wgd (v1.1.2) [ 70 ].

All-versus-all protein blast among 9 decapod species was performed using MCScanX [ 71 ]. Results of all-versus-all protein blast were inputted into DupGen_finder to determine modes of duplicated gene pairs [ 32 ].

Phylogenetic reconstruction

Protein sequences of 12 arthropod species ( Drosophila melanogaster , Daphnia magna , Hyalella azteca , Litopenaeus vannamei , Fenneropenaeus chinensis , Penaeus monodon , Macrobrachium nipponense , Eriocheir sinensis , Scylla paramamosain , Homarus americanus , Procambarus clarkii , Procambarus virginalis ) were downloaded from NCBI. Protein sequences shorter than 50 amino acids were removed. OrthoFinder (v 2.5.4) [ 72 ] was applied to determine and cluster gene families among these 12 species and C. quadricarinatus . Gene clusters with > 100 gene copies in one or more species were removed. Single-copy orthologs in each gene cluster were aligned using MAFFT (v7.490) [ 73 ]. Alignments were trimmed using ClipKit (v1.2.0) with “gappy” mode [ 74 ]. The phylogenetic tree was reconstructed with the trimmed alignments using a maximum-likelihood method implemented in IQ-TREE2 (v2.1.2) with D. melanogaster as outgroup [ 75 ]. The best-fit substitution model was selected using ModelFinder algorithm [ 76 ]. Branch supports were assessed using the ultrafast bootstrap (UFBoot) approach with 1,000 replicates [ 77 ].

To estimate the divergence time between species or clade, the trimmed alignments of single-copy orthologs were concatenated using PhyloSuite (v1.2.2) [ 78 ]. Concatenated alignment was used to estimated divergent times among species using the MCMCtree module of the PAML package (v4.9) [ 79 ]. MCMCtree analysis was performed using the maximum-likelihood tree reconstructed by IQ-TREE2 as a guide tree and calibrated with the divergent time obtained from the TimeTree database (minimum = 58 million years and soft maximum = 108 million years between L. vannamei and P. monodon ; minimum = 154 million years and soft maximum = 242 million years between C. quadricarinatus and P. clarkii ; minimum = 526 million years and soft maximum = 578 million years between H. azteca and D. magna ) [ 80 ].

Gene family expansion and contraction analysis

CAFÉ (v5) was applied to determine the significance of gene-family expansion and contraction among the 13 arthropod species based on the MCMCtree-generated ultrametric tree and OrthoMCL-determined gene clusters used for species tree reconstruction [ 81 ].

The IgLectin protein family was significantly expanded in the C. quadricarinatus genome compared with other decapod species. Phylogenetic tree reconstruction was performed to investigate the evolutionary relationships of IgLectin proteins from C. quadricarinatus and other decapod species. The IgLectin protein sequences from P. clarkii (AGL46986.1) and E. sinensis (UIS31342.1) were downloaded from NCBI [ 43 , 44 ]. And the Ig-Lectin protein sequences from S. paramamosain were extracted from published genome sequence [ 82 ]. Protein sequences were aligned using MAFFT (v7.490) [ 83 ]. The phylogenetic tree was reconstructed using maximum-likelihood alignment implemented in IQ-TREE2 (v2.2.0), with a C-type Lectin protein (XP_045031378.1) from D. magna as the outgroup. The best-fit substitution model was selected using the ModelFinder algorithm [ 76 ]. Branch supports were assessed using UFBoot with 1000 replicates [ 77 ].

Identification of positively selected genes

PSGs in the C. quadricarinatus genome were identified using PosiGene (v0.1) [ 84 ] with parameters “-as = D.melanogaster , -ts = C. quadricarinatus -rs = D.melanogaster , -nhsbr”. Genes with a P -value < 0.05 were identified to have been subject to positive selection.

Tissue distribution and pattern during development of CqKDM5A expression

Juvenile crayfish (1.5 ± 0.3 g) and adult crayfish (250 ± 3 g) were obtained from Zhongshan, Guangdong Province, China., and cultivated in aerated tanks at 26 °C for a minimum of two weeks before being used in the experiments.

To investigate the tissue distribution of CqKDM5A expression, samples of gill, epidermis, heart, stomach, hepatopancreases, hemocyte, muscle, intestine, and eyestalk were collected from juvenile crayfish. In addition, samples of hemocyte were collected from adult crayfish and juvenile crayfish to investigate the expression patten of CqKDM5A during growth. The expression levels of CqKDM5A were determined using qPCR assay, with Cqβ-actin (GenBank No. AY430093.1) used as the internal control for normalization (Supplementary Table 13) [ 85 ]. The primer pair efficiency was evaluated following the MIQE method, with a tenfold logarithmic dilution of a cDNA mixture used to generate a linear standard curve [ 86 ].

RNAi knockdown of CqKDM5A

Specific dsRNA targeting the CqKDM5A and CqGFP genes were synthesized through in vitro transcription using the T7 RiboMAX Express RNAi System kit (Promega, cat. no. P1700, USA) (Supplementary Table 13). The experimental groups were injected with CqKDM5A dsRNA (2 μg/g per individual), and the control groups were injected with GFP dsRNA.

Crayfish without dsRNA injection, crayfish injected with GFP dsRNA, and crayfish injected with CqKDM5A dsRNA were maintained in both normoxic and hypoxic conditions. For hypoxia experiments, crayfish were reared in sealed aquariums covered with plastic films and filled with cooled boiling water (1 ± 0.1 mg/L dissolved oxygen, 0‰ salinity, 25 °C). The dissolved oxygen levels were measured using a dissolved oxygen analyzer (Yieryi, DO9100, China). Every twelve hours, crayfish were transferred to new tanks with freshly treated water to maintain the hypoxic environment. Expression levels of CqKDM5A were determined using qPCR assay, with Cqβ-actin (GenBank No. AY430093.1) used as the internal control for normalization (Supplementary Table 13).

Availability of data and materials

Raw reads and genome assemblies are accessible in NCBI under BioProject number PRJNA904538. Raw reads and genome assemblies are also available at the CNGB Sequence Archive (CNSA) of China National GeneBank DataBase (CNGBdb) with accession number CNP0005505.

Abbreviations

Lysine demethylase 3A

Lysine demethylase 5A

Heme oxygenases-2

Hypoxia-inducible factor

Oxygen-dependent prolyl hydroxylase

Factor-inhibiting HIF

Aryl hydrocarbon nuclear translocator

2-Oxoglutarate-dependent dioxygenase

Lysine demethylase 6A

Lysine demethylase 6B

Positively selected gene

Benchmarking Universal Single-Copy Orthologs

Quality value

Transposable element

Long interspersed nuclear element

Long terminal repeat

Whole-genome duplication

Tandem repeat

Simple sequence repeat

Synonymous nucleotide substitutions

Maximum-likelihood

Million years ago

Pattern recognition receptor

Pathogen-associated molecular pattern

Prophenoloxidase

Antimicrobial peptide

Immunoglobulin

Quantitative PCR

RNA interference

Nonredundant

Kyoto Encyclopedia of Genes and Genomes

Tandem repeat finder

Reid AJ, Carlson AK, Creed IF, Eliason EJ, Gell PA, Johnson PTJ, Kidd KA, MacCormack TJ, Olden JD, Ormerod SJ, et al. Emerging threats and persistent conservation challenges for freshwater biodiversity. iol Rev Camb Philos Soc. 2019;94(3):849–73.

Article Google Scholar

Simberloff D, Martin JL, Genovesi P, Maris V, Wardle DA, Aronson J, Courchamp F, Galil B, Garcia-Berthou E, Pascal M, et al. Impacts of biological invasions: what’s what and the way forward. Trends Ecol Evol. 2013;28(1):58–66.

Article PubMed Google Scholar

Patoka J, Fisakova MN, Kalous L, Skrdla P, Kuca M. Earliest Evidence for Human Consumption of Crayfish. Crustaceana. 2014;87(13):1578–85.

Gherardi F. Towards a sustainable human use of freshwater crayfish (Crustacea, Decapoda, Astacidea). Knowl Manag Aquat Ec. 2011;401:02.

Momot WT. Redefining the role of crayfish in aquatic ecosystems. Rev Fish. 1995;3(1):33–63.

Riek EF. The Australian freshwater crayfish (Crustacea : Decapoda : Parastacidae), with descriptions of a new species. Aust J Zool. 1969;17(5):855–918.

FAO. The state of world fisheries and aquaculture 2022: towards blue transformation. In : Rome: FAO; 2022.

Haubrock PJ, Oficialdegui FJ, Zeng Y, Patoka J, Yeo DCJ, Kouba A. The redclaw crayfish: a prominent aquaculture species with invasive potential in tropical and subtropical biodiversity hotspots. Rev Aquacult. 2021;13:1488–530.

Douthwaite RJ, Jones EW, Tyser AB, Vrdoljak SM. The introduction, spread and ecology of redclaw crayfish Cherax quadricarinatus in the Zambezi catchment. Afr J Aquat Sci. 2018;43(4):353–66.

Ahyong ST, Yeo DCJ. Feral populations of the Australian Red-Claw crayfish (Cherax quadricarinatus von Martens) in water supply catchments of Singapore. Biol Invasions. 2007;9(8):943–6.

Powers QM, Aranguren LF, Fitzsimmons KM, McLain JE, Dhar AK. Crayfish (Cherax quadricarinatus) susceptibility to acute hepatopancreatic necrosis disease (AHPND). J Invertebr Pathol. 2021;186:107554.

Edgerton B, Owens L. Age at first infection of Cherax quadricarinatus by Cherax quadricarinatus bacilliform virus and Cherax Giardiavirus-like virus, and production of putative virus-free crayfish. Aquaculture. 1997;152(1):1–12.

Hayakijkosol O, La Fauce K, Owens L. Experimental infection of redclaw crayfish (Cherax quadricarinatus) with nodavirus, the aetiological agent of white tail disease. Aquaculture. 2011;319(1–2):25–9.

Chen H, Zhang R, Liu F, Shao C, Liu F, Li W, Ren J, Niu B, Liu H, Lou B. The chromosome-level genome of Cherax quadricarinatus. Sci Data. 2023;10(1):215.

Article CAS PubMed PubMed Central Google Scholar

Tan MH, Gan HM, Lee YP, Grandjean F, Croft LJ, Austin CM. A Giant Genome for a Giant Crayfish (Cherax quadricarinatus) With Insights Into cox1 Pseudogenes in Decapod Genomes. Front Genet. 2020;11:201.

Hammarlund EU, Flashman E, Mohlin S, Licausi F. Oxygen-sensing mechanisms across eukaryotic kingdoms and their roles in complex multicellularity. Science. 2020;370(6515):eaba3512.

Article CAS PubMed Google Scholar

Epstein AC, Gleadle JM, McNeill LA, Hewitson KS, O’Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI, Dhanda A, et al. C. elegans EGL-9 and mammalian homologs define a family of dioxygenases that regulate HIF by prolyl hydroxylation. Cell. 2001;107(1):43–54.

Islam MS, Leissing TM, Chowdhury R, Hopkinson RJ, Schofield CJ. 2-Oxoglutarate-Dependent Oxygenases. Annu Rev Biochem. 2018;87:585–620.

Losman JA, Koivunen P, Kaelin WG Jr. 2-Oxoglutarate-dependent dioxygenases in cancer. Nat Rev Cancer. 2020;20(12):710–26.

Chakraborty AA, Laukka T, Myllykoski M, Ringel AE, Booker MA, Tolstorukov MY, Meng YJ, Meier SR, Jennings RB, Creech AL, et al. Histone demethylase KDM6A directly senses oxygen to control chromatin and cell fate. Science. 2019;363(6432):1217–22.

Batie M, Frost J, Frost M, Wilson JW, Schofield P, Rocha S. Hypoxia induces rapid changes to histone methylation and reprograms chromatin. Science. 2019;363(6432):1222–6.

Qian X, Li X, Shi Z, Bai X, Xia Bai, Zheng Y, Xu D, Chen F, You Y, Fang J, et al. KDM3A Senses Oxygen Availability to Regulate PGC-1α-Mediated Mitochondrial Biogenesis. Mol Cell. 2019;76(6):885–895 e887.

Shafin K, Pesout T, Lorig-Roach R, Haukness M, Olsen HE, Bosworth C, Armstrong J, Tigyi K, Maurer N, Koren S, et al. Nanopore sequencing and the Shasta toolkit enable efficient de novo assembly of eleven human genomes. Nat Biotechnol. 2020;38(9):1044–53.

Ruan J, Li H. Fast and accurate long-read assembly with wtdbg2. Nat Methods. 2020;17(2):155–8.

Chakraborty M, Baldwin-Brown JG, Long AD, Emerson JJ. Contiguous and accurate de novo assembly of metazoan genomes with modest long read coverage. Nucleic Acids Res. 2016;44(19):e147.

PubMed PubMed Central Google Scholar

Simao FA, Waterhouse RM, Ioannidis P, Kriventseva EV, Zdobnov EM. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics. 2015;31(19):3210–2.

Rhie A, Walenz BP, Koren S, Phillippy AM. Merqury: reference-free quality, completeness, and phasing assessment for genome assemblies. Genome Biol. 2020;21(1):245.

Iannucci A, Saha A, Cannicci S, Bellucci A, Cheng CLY, Ng KH, et al. Ecological, physiological and life-history traits correlate with genome sizes in decapod crustaceans. Front Ecol Evol. 2022;10. https://doi.org/10.3389/fevo.2022.930888 .

Lynch M. The Origins of Genome Architecture. Sunderland, MA: Sinauer Associates; 2007.

Google Scholar

Grover CE, Wendel JF. Recent Insights into Mechanisms of Genome Size Change in Plants. J Bot. 2010;2010:1–8.

Yuan J, Zhang X, Wang M, Sun Y, Liu C, Li S, Yu Y, Gao Y, Liu F, Zhang X, et al. Simple sequence repeats drive genome plasticity and promote adaptive evolution in penaeid shrimp. Commun Biol. 2021;4(1):186.

Qiao X, Li Q, Yin H, Qi K, Li L, Wang R, Zhang S, Paterson AH. Gene duplication and evolution in recurring polyploidization-diploidization cycles in plants. Genome Biol. 2019;20(1):38.

Article PubMed PubMed Central Google Scholar

Bracken-Grissom HD, Ahyong ST, Wilkinson RD, Feldmann RM, Schweitzer CE, Breinholt JW, Bendall M, Palero F, Chan TY, Felder DL, et al. The Emergence of Lobsters: Phylogenetic Relationships, Morphological Evolution and Divergence Time Comparisons of an Ancient Group (Decapoda: Achelata, Astacidea, Glypheidea, Polychelida). Syst Biol. 2014;63(4):457–79.

Wolfe JM, Breinholt JW, Crandall KA, Lemmon AR, Lemmon EM, Timm LE, Siddall ME, Bracken-Grissom HD. A phylogenomic framework, evolutionary timeline and genomic resources for comparative studies of decapod crustaceans. Proc Biol Sci. 1901;2019(286):20190079.

Janeway CA Jr, Medzhitov R. Innate immune recognition. Annu Rev Immunol. 2002;20:197–216.

Cerenius L, Soderhall K. The prophenoloxidase-activating system in invertebrates. Immunol Rev. 2004;198:116–26.

Huang Y, Ren Q. Research progress in innate immunity of freshwater crustaceans. Dev Comp Immunol. 2020;104:103569.

Brown GD, Willment JA, Whitehead L. C-type lectins in immunity and homeostasis. Nat Rev Immunol. 2018;18(6):374–89.

Gerardo NM, Altincicek B, Anselme C, Atamian H, Barribeau SM, de Vos M, Duncan EJ, Evans JD, Gabaldon T, Ghanim M, et al. Immunity and other defenses in pea aphids, Acyrthosiphon pisum. Genome Biol. 2010;11(2):R21.

Wang XW, Wang JX. Diversity and multiple functions of lectins in shrimp immunity. Dev Comp Immunol. 2013;39(1–2):27–38.

Marshall JS, Warrington R, Watson W, Kim HL. An introduction to immunology and immunopathology. Allergy Asthma Clin Immunol. 2018;14(Suppl 2):49.

Watson FL, Puttmann-Holgado R, Thomas F, Lamar DL, Hughes M, Kondo M, Rebel VI, Schmucker D. Extensive diversity of Ig-superfamily proteins in the immune system of insects. Science. 2005;309(5742):1874–8.

Zhang XW, Wang Y, Wang XW, Wang L, Mu Y, Wang JX. A C-type lectin with an immunoglobulin-like domain promotes phagocytosis of hemocytes in crayfish Procambarus clarkii. Sci Rep. 2016;6:29924.

Zhou K, Qin Y, Song Y, Zhao K, Pan W, Nan X, Wang Y, Wang Q, Li W. A Novel Ig Domain-Containing C-Type Lectin Triggers the Intestine-Hemocyte Axis to Regulate Antibacterial Immunity in Crab. J Immunol. 2022;208(10):2343–62.

Dudgeon D, Arthington AH, Gessner MO, Kawabata Z, Knowler DJ, Leveque C, Naiman RJ, Prieur-Richard AH, Soto D, Stiassny ML, et al. Freshwater biodiversity: importance, threats, status and conservation challenges. Biol Rev. 2006;81(2):163–82.

Jenny JP, Francus P, Normandeau A, Lapointe F, Perga ME, Ojala A, Schimmelmann A, Zolitschka B. Global spread of hypoxia in freshwater ecosystems during the last three centuries is caused by rising local human pressure. Glob Change Biol. 2016;22(4):1481–9.

Crowl TA, Crist TO, Parmenter RR, Belovsky G, Lugo AE. The spread of invasive species and infectious disease as drivers of ecosystem change. Front Ecol Environ. 2008;6(5):238–46.

Crandall KA, De Grave S. An updated classification of the freshwater crayfishes (Decapoda: Astacidea) of the world, with a complete species list. J Crustacean Biol. 2017;37(5):615–53.

Hancock RL, Dunne K, Walport LJ, Flashman E, Kawamura A. Epigenetic regulation by histone demethylases in hypoxia. Epigenomics. 2015;7(5):791–811.

Belton JM, McCord RP, Gibcus JH, Naumova N, Zhan Y, Dekker J. Hi-C: a comprehensive technique to capture the conformation of genomes. Methods. 2012;58(3):268–76.

Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90.

Marcais G, Kingsford C. A fast, lock-free approach for efficient parallel counting of occurrences of k-mers. Bioinformatics. 2011;27(6):764–70.

Star B, Nederbragt AJ, Jentoft S, Grimholt U, Malmstrom M, Gregers TF, Rounge TB, Paulsen J, Solbakken MH, Sharma A, et al. The genome sequence of Atlantic cod reveals a unique immune system. Nature. 2011;477(7363):207–10.

Luo R, Liu B, Xie Y, Li Z, Huang W, Yuan J, He G, Chen Y, Pan Q, Liu Y, et al. SOAPdenovo2: an empirically improved memory-efficient short-read de novo assembler. Gigascience. 2012;1(1):18.

Vaser R, Sovic I, Nagarajan N, Sikic M. Fast and accurate de novo genome assembly from long uncorrected reads. Genome Res. 2017;27(5):737–46.

Hu J, Fan J, Sun Z, Liu S. NextPolish: a fast and efficient genome polishing tool for long-read assembly. Bioinformatics. 2020;36(7):2253–5.

Guan D, McCarthy SA, Wood J, Howe K, Wang Y, Durbin R. Identifying and removing haplotypic duplication in primary genome assemblies. Bioinformatics. 2020;36(9):2896–8.

Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9.

Burton JN, Adey A, Patwardhan RP, Qiu R, Kitzman JO, Shendure J. Chromosome-scale scaffolding of de novo genome assemblies based on chromatin interactions. Nat Biotechnol. 2013;31(12):1119–25.

Mikheenko A, Prjibelski A, Saveliev V, Antipov D, Gurevich A. Versatile genome assembly evaluation with QUAST-LG. Bioinformatics. 2018;34(13):i142–50.

Flynn JM, Hubley R, Goubert C, Rosen J, Clark AG, Feschotte C, Smit AF. RepeatModeler2 for automated genomic discovery of transposable element families. Proc Natl Acad Sci U S A. 2020;117(17):9451–7.

Pertea M, Kim D, Pertea GM, Leek JT, Salzberg SL. Transcript-level expression analysis of RNA-seq experiments with HISAT. StringTie and Ballgown Nat Protoc. 2016;11(9):1650–67.

Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, Adiconis X, Fan L, Raychowdhury R, Zeng QD, et al. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol. 2011;29(7):644–U130.

Huerta-Cepas J, Forslund K, Coelho LP, Szklarczyk D, Jensen LJ, von Mering C, Bork P. Fast Genome-Wide Functional Annotation through Orthology Assignment by eggNOG-Mapper. Mol Biol Evol. 2017;34(8):2115–22.

Mulder N, Apweiler R. InterPro and InterProScan: tools for protein sequence classification and comparison. Methods Mol Biol. 2007;396:59–70.

Kanehisa M, Sato Y, Morishima K. BlastKOALA and GhostKOALA: KEGG Tools for Functional Characterization of Genome and Metagenome Sequences. J Mol Biol. 2016;428(4):726–31.

Benson G. Tandem repeats finder: a program to analyze DNA sequences. Nucleic Acids Res. 1999;27(2):573–80.

Wang X, Wang L. GMATA: An Integrated Software Package for Genome-Scale SSR Mining, Marker Development and Viewing. Front Plant Sci. 2016;7:1350.

Kofler R, Schlotterer C, Lelley T. SciRoKo: a new tool for whole genome microsatellite search and investigation. Bioinformatics. 2007;23(13):1683–5.

Zwaenepoel A, Van de Peer Y. wgd-simple command line tools for the analysis of ancient whole-genome duplications. Bioinformatics. 2019;35(12):2153–5.

Wang Y, Tang H, Debarry JD, Tan X, Li J, Wang X, Lee TH, Jin H, Marler B, Guo H, et al. MCScanX: a toolkit for detection and evolutionary analysis of gene synteny and collinearity. Nucleic Acids Res. 2012;40(7):e49.

Emms DM, Kelly S. OrthoFinder: phylogenetic orthology inference for comparative genomics. Genome Biol. 2019;20(1):238.

Katoh K, Misawa K, Kuma K, Miyata T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 2002;30(14):3059–66.

Steenwyk JL, Buida TJ 3rd, Li Y, Shen XX, Rokas A. ClipKIT: A multiple sequence alignment trimming software for accurate phylogenomic inference. PLoS Biol. 2020;18(12):e3001007.

Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, von Haeseler A, Lanfear R. IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol. 2020;37(5):1530–4.

Kalyaanamoorthy S, Minh BQ, Wong TKF, von Haeseler A, Jermiin LS. ModelFinder: fast model selection for accurate phylogenetic estimates. Nat Methods. 2017;14(6):587–9.

Hoang DT, Chernomor O, von Haeseler A, Minh BQ, Vinh LS. UFBoot2: Improving the Ultrafast Bootstrap Approximation. Mol Biol Evol. 2018;35(2):518–22.

Zhang D, Gao F, Jakovlic I, Zou H, Zhang J, Li WX, Wang GT. PhyloSuite: An integrated and scalable desktop platform for streamlined molecular sequence data management and evolutionary phylogenetics studies. Mol Ecol Resour. 2020;20(1):348–55.

Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol. 2007;24(8):1586–91.

Kumar S, Stecher G, Suleski M, Hedges SB. TimeTree: A Resource for Timelines, Timetrees, and Divergence Times. Mol Biol Evol. 2017;34(7):1812–9.

De Bie T, Cristianini N, Demuth JP, Hahn MWCAFE. a computational tool for the study of gene family evolution. Bioinformatics. 2006;22(10):1269–71.

Zhao M, Wang W, Zhang F, Ma C, Liu Z, Yang MH, et al. A chromosome-level genome of the mud crab (Scylla paramamosain estampador) provides insights into the evolution of chemical and light perception in this crustacean. Mol Ecol Resour. 2021;21:1299–317.

Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013;30(4):772–80.

Sahm A, Bens M, Platzer M, Szafranski K. PosiGene: automated and easy-to-use pipeline for genome-wide detection of positively selected genes. Nucleic Acids Res. 2017;45(11):e100.

Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCT) Method. Methods. 2001;25(4):402–8.

Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009;55(4):611–22.

Download references

Acknowledgements

We gratefully acknowledge the National Supercomputing Center in Guangzhou for provision of computational resources.

This study was supported by the earmarked fund of China Agriculture Research System for CARS-48, Key Scientific and Technological Grant of Zhejiang for Breeding New Agricultural Varieties (2021C02069-4–5), Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai) (SML2023SP234), Fundamental Research Funds for the Central Universities, Sun Yat-sen University (23ptpy23).

Author information

Ziwei Liu, Jianbo Zheng and Haoyang Li contributed equally to this work.

Authors and Affiliations

State Key Laboratory for Biocontrol, School of Marine Sciences, Sun Yat-Sen University, Zhuhai, 519000, China

Ziwei Liu, Ke Fang, Jian He, Dandan Zhou, Jianguo He & Muhua Wang

Key Laboratory of Genetics and Breeding, Zhejiang Institute of Freshwater Fisheries, Huzhou, 313001, China

Jianbo Zheng, Meili Chi, Zhimin Gu & Fei Li

China-ASEAN Belt and Road Joint Laboratory On Mariculture Technology, Guangdong Provincial Key Laboratory of Aquatic Economic Animals, School of Life Sciences, Sun Yat-Sen University, Guangzhou, 510275, China

Haoyang Li, Sheng Wang, Shaoping Weng, Jianguo He & Muhua Wang

Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, 519000, China

Haoyang Li, Jian He, Shaoping Weng, Jianguo He & Muhua Wang

Zhejiang Academy of Agricultural Sciences, Hangzhou, 310021, China

You can also search for this author in PubMed Google Scholar

Contributions

J.G.H., F.L., M.W. conceived of the project and designed research; J.Z., D.Z., S.P.W. collected the sample; Z.L., K.F. assembled and annotated the genomes; Z.L., J.Z., M.C., Z.G. conducted the evolutionary analyses; H.L., S.W., J.H. conducted the experiments of hypoxia tolerance; J.G.H., F.L., M.W. wrote the paper with contribution from all authors.

Corresponding authors

Correspondence to Jianguo He , Fei Li or Muhua Wang .

Ethics declarations

Ethics approval and consent to participate.

All animal experiments were approved by Institutional Animal Care and Use Committee of Sun Yat-Sen University (Approval No. SYSU-IACUC-2023-B0005). Crayfishes and shrimps were anaesthetized on ice prior to all experiments. All efforts were made to minimize animal suffering.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary material 1., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .

Reprints and permissions

About this article

Cite this article.

Liu, Z., Zheng, J., Li, H. et al. Genome assembly of redclaw crayfish ( Cherax quadricarinatus ) provides insights into its immune adaptation and hypoxia tolerance. BMC Genomics 25 , 746 (2024). https://doi.org/10.1186/s12864-024-10673-9

Download citation

Received : 22 April 2024

Accepted : 29 July 2024

Published : 31 July 2024

DOI : https://doi.org/10.1186/s12864-024-10673-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Cherax quadricarinatus
Immune-related gene family

BMC Genomics

ISSN: 1471-2164

Submission enquiries: [email protected]
General enquiries: [email protected]

OpenAI launches experimental GPT-4o Long Output model with 16X token capacity

Share on Facebook
Share on LinkedIn

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More

OpenAI is reportedly eyeing a cash crunch , but that isn’t stopping the preeminent generative AI company from continuing to release a steady stream of new models and updates.

Yesterday, the company quietly posted a webpage announcing a new large language model (LLM): GPT-4o Long Output, which is a variation on its signature GPT-4o model from May, but with a massively extended output size: up to 64,000 tokens of output instead of GPT-4o’s initial 4,000 — a 16-fold increase.

Tokens, as you may recall, refer to the numerical representations of concepts , grammatical constructions, and combinations of letters and numbers organized based on their semantic meaning behind-the-scenes of an LLM.

The word “Hello” is one token, for example, but so too is “hi.” You can see an interactive demo of tokens in action via OpenAI’s Tokenizer here . Machine learning researcher Simon Willison also has a great interactive token encoder/decoder .

By offering a 16X increase in token outputs with the new GPT-4o Long Output variant, OpenAI is now giving users — and more specifically, third-party developers building atop its application programming interface (API) — the opportunity to have the chatbot return far longer responses, up to about a 200-page novel in length.

Why is OpenAI launching a longer output model?

OpenAI’s decision to introduce this extended output capability stems from customer feedback indicating a need for longer output contexts.

An OpenAI spokesperson explained to VentureBeat: “We heard feedback from our customers that they’d like a longer output context. We are always testing new ways we can best serve our customers’ needs.”

The alpha testing phase is expected to last for a few weeks, allowing OpenAI to gather data on how effectively the extended output meets user needs.

This enhanced capability is particularly advantageous for applications requiring detailed and extensive output, such as code editing and writing improvement.

By offering more extended outputs, the GPT-4o model can provide more comprehensive and nuanced responses, which can significantly benefit these use cases.

Distinction between context and output

Already, since launch, GPT-4o offered a maximum 128,000 context window — the amount of tokens the model can handle in any one interaction, including both input and output tokens .

For GPT-4o Long Output, this maximum context window remains at 128,000.

So how is OpenAI able to increase the number of output tokens 16-fold from 4,000 to 64,000 tokens while keeping the overall context window at 128,000?

It call comes down to some simple math: even though the original GPT-4o from May had a total context window of 128,000 tokens, its single output message was limited to 4,000.

Similarly, for the new GPT-4o mini window, the total context is 128,000 but the maximum output has been raised to 16,000 tokens.

That means for GPT-4o, the user can provide up to 124,000 tokens as an input and receive up to 4,000 maximum output from the model in a single interaction. They can also provide more tokens as input but receive fewer as output, while still adding up to 128,000 total tokens.

For GPT-4o mini, the user can provide up to 112,000 tokens as an input in order to get a maximum output of 16,000 tokens back.

For GPT-4o Long Output, the total context window is still capped at 128,000. Yet, now, the user can provide up to 64,000 tokens worth of input in exchange for a maximum of 64,000 tokens back out — that is, if the user or developer of an application built atop it wants to prioritize longer LLM responses while limiting the inputs.

In all cases, the user or developer must make a choice or trade-off: do they want to sacrifice some input tokens in favor of longer outputs while still remaining at 128,000 tokens total? For users who want longer answers, the GPT-4o Long Output now offers this as an option.

Priced aggressively and affordably

The new GPT-4o Long Output model is priced as follows:

$6 USD per 1 million input tokens
$18 per 1 million output tokens

Compare that to the regular GPT-4o pricing which is $5 per million input tokens and $15 per million output, or even the new GPT-4o mini at $0.15 per million input and $0.60 per million output, and you can see it is priced rather aggressively, continuing OpenAI’s recent refrain that it wants to make powerful AI affordable and accessible to wide swaths of the developer userbase.

Currently, access to this experimental model is limited to a small group of trusted partners. The spokesperson added, “We’re conducting alpha testing for a few weeks with a small number of trusted partners to see if longer outputs help their use cases.”

Depending on the outcomes of this testing phase, OpenAI may consider expanding access to a broader customer base.

Future prospects

The ongoing alpha test will provide valuable insights into the practical applications and potential benefits of the extended output model.

If the feedback from the initial group of partners is positive, OpenAI may consider making this capability more widely available, enabling a broader range of users to benefit from the enhanced output capabilities.

Clearly, with the GPT-4o Long Output model, OpenAI hopes to address an even wider range of customer requests and power applications requiring detailed responses.

Stay in the know! Get the latest news in your inbox daily

By subscribing, you agree to VentureBeat's Terms of Service.

Thanks for subscribing. Check out more VB newsletters here .

An error occured.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

View all journals
Explore content
About the journal
Publish with us
Sign up for alerts
Open access
Published: 25 July 2024

Experimental demonstration of magnetic tunnel junction-based computational random-access memory

Yang Lv 1 ,
Brandon R. Zink 1 ,
Robert P. Bloom 1 ,
Hüsrev Cılasun 1 ,
Pravin Khanal 2 ,
Salonik Resch 1 ,
Zamshed Chowdhury 1 ,
Ali Habiboglu 2 ,
Weigang Wang 2 ,
Sachin S. Sapatnekar 1 ,
Ulya Karpuzcu 1 &
Jian-Ping Wang 1

npj Unconventional Computing volume 1 , Article number: 3 ( 2024 ) Cite this article

3516 Accesses

223 Altmetric

Metrics details

Computational science
Electrical and electronic engineering
Electronic and spintronic devices
Magnetic devices

The conventional computing paradigm struggles to fulfill the rapidly growing demands from emerging applications, especially those for machine intelligence because much of the power and energy is consumed by constant data transfers between logic and memory modules. A new paradigm, called “computational random-access memory (CRAM),” has emerged to address this fundamental limitation. CRAM performs logic operations directly using the memory cells themselves, without having the data ever leave the memory. The energy and performance benefits of CRAM for both conventional and emerging applications have been well established by prior numerical studies. However, there is a lack of experimental demonstration and study of CRAM to evaluate its computational accuracy, which is a realistic and application-critical metric for its technological feasibility and competitiveness. In this work, a CRAM array based on magnetic tunnel junctions (MTJs) is experimentally demonstrated. First, basic memory operations, as well as 2-, 3-, and 5-input logic operations, are studied. Then, a 1-bit full adder with two different designs is demonstrated. Based on the experimental results, a suite of models has been developed to characterize the accuracy of CRAM computation. Scalar addition, multiplication, and matrix multiplication, which are essential building blocks for many conventional and machine intelligence applications, are evaluated and show promising accuracy performance. With the confirmation of MTJ-based CRAM’s accuracy, there is a strong case that this technology will have a significant impact on power- and energy-demanding applications of machine intelligence.

A CMOS-integrated compute-in-memory macro based on resistive random-access memory for AI edge devices

A compute-in-memory chip based on resistive random-access memory

A four-megabit compute-in-memory macro with eight-bit precision based on CMOS and resistive random-access memory for AI edge devices

Introduction.

Recent advances in machine intelligence 1 , 2 for tasks such as recommender systems 3 , speech recognition 4 , natural language processing 5 , and computer vision 6 , have been placing growing demands on our computing systems, especially for implementations with artificial neural networks. A variety of platforms are used, from general-purpose CPUs and GPUs 7 , 8 , to FPGAs 9 , to custom-designed accelerators and processors 10 , 11 , 12 , 13 , to mixed- or fully- analog circuits 14 , 15 , 16 , 17 , 18 , 19 , 20 . Most are based on the Von Neumann architecture, with separate logic and memory systems. As shown in Fig. 1a , the inherent segregation of logic and memory requires large amounts of data to be transferred between these modules. In data-intensive scenarios, this transfer becomes a major bottleneck in terms of performance, energy consumption, and cost 21 , 22 , 23 . For example, the data movement consumes about 200 times the energy used for computation when reading three 64-bit source operands from and writing one 64-bit destination operand to an off-chip main memory 21 . This bottleneck has long been studied. Research aiming at connecting logic and memory more closely has led to new computation paradigms.

a , b Compared to a conventional computer architecture ( a ), which suffers from the memory-logic transfer bottleneck, CRAM ( b ) offers significant power and performance improvements. Its unique architecture allows for computation in memory, as well as, random access, reconfigurability, and parallel operation capability. c The CRAM could excel in data-intensive, memory-centric, or power-sensitive applications, such as neural networks, image processing, or edge computing ( c ).

Promising paradigms include “near-memory” and “in-memory” computing. Near-memory processing brings logic physically closer to memory by employing 3D-stacked architectures 24 , 25 , 26 , 27 , 28 , 29 . In-memory computing scatters clusters of logic throughout or around the memory banks on a single chip 14 , 15 , 16 , 17 , 18 , 19 , 20 , 30 , 31 , 32 , 33 , 34 , 35 . Yet another approach is to build systems where the memory itself can perform computation. This has been dubbed “true” in-memory computing 36 , 37 , 38 , 39 , 40 , 41 , 42 . The computational random-access memory (CRAM) 38 , 40 is one of the true in-memory computing paradigms. Logic is performed natively by the memory cells; the data for logic operations never has to leave the memory (Fig. 1b ). Additionally, CRAM operates in a fully digital fashion, unlike most other reported in-memory computing schemes 14 , 15 , 16 , 17 , 18 , 19 , 20 , which are partially or mostly analog. CRAM promises superior energy efficiency and processing performance for machine intelligence applications. It has unique additional features, such as random-access of data and operands, massive parallel computing capabilities, and reconfigurability of operations 38 , 40 . Also note that although the transistor-less (crossbar) architecture employed by most of the previous true-in-memory computing paradigms 36 , 37 , 39 , 42 allows for higher density, the maximum allowable size of the memory array is often severely limited due to the sneak path issues. CRAM includes transistors in each of its cells for better-controlled electrical accessibility and, therefore, a larger array size.

The CRAM was initially proposed based on the MTJ device 38 , an emerging memory device that relies on spin electronics 43 . Such “spintronic” devices, along with other non-volatile emerging memory devices, usually referred to as “X” for logic applications, have been intensively investigated over the past several decades for emerging memory and computing applications as “beyond-CMOS” and/or “CMOS + X” technologies. They offer vastly improved speed, energy efficiency, area, and cost. An additional feature that is exploited by CRAM is their non-volatility 44 . The MTJ device is the most mature of spintronic devices for embedded memory applications, based on endurance 45 , energy efficiency 46 , and speed 47 . We note that CRAM can be implemented based not only on spintronics devices but also on other non-volatile emerging memory devices.

In its simplest form, an MTJ consists of a thin tunneling barrier layer sandwiched between two ferromagnetic (FM) layers. When a voltage is applied between the two layers, electrons tunnel through the barrier, resulting in a charge current. The resistance of the MTJ is a function of the magnetic state of the two FM layers, due to the tunneling magnetoresistance (TMR) effect 48 , 49 , 50 . An MTJ can be engineered to be magnetically bi-stable. Accordingly, it can store information based on its magnetic state. This information can be retrieved by reading the resistance of the device. The MTJ can be electrically switched from one state to the other with a current due to the spin-transfer torque (STT) effect 51 , 52 . In this way, an MTJ can be used as an electrically operated memory device with both read and write functionality. A type of random-access memory, the STT-MRAM 53 , 54 , 55 , 56 has been developed commercially, utilizing MTJs as memory cells. A typical STT-MRAM consists of an array of bit cells, each containing one transistor and one MTJ. These are referred to as 1 transistor 1 MTJ (1T1M) cells.

A typical CRAM cell design, as shown in Fig. 2a , is a modification of the 1T1M STT-MRAM architecture 57 . The MTJ, one of the transistors, word line (WL), bit select line (BSL), and memory bit line (MBL) resemble the 1T1M cell architecture of STT-MRAM, which allows the CRAM to perform memory operations. In order to enable logic operations, a second transistor, as well as a logic line (LL) and a logic bit line (LBL), are added to each memory cell. During a logic operation, corresponding transistors and lines are manipulated so that several MTJs in a row are temporarily connected to a shared LL 40 . While the LL is left floating, voltage pulses are applied to the lines connecting to input MTJs, with that of the output MTJ being grounded. The logic operation is based on a working principle called voltage-controlled logic (VCL) 58 , 59 , which utilizes the thresholding effect that occurs when switching an MTJ and the TMR effect of MTJ. As shown in Fig. 2b , when a voltage is applied across the input MTJs, the different resistance values result in different current levels. The current flows through the output MTJ, which may or may not switch its state, depending on the states of the input MTJs. In this way, basic bitwise logic operations, such as AND, OR, NAND, NOR, and MAJ, can be realized. A unique feature of VCL is that the logic operation itself does not require the data in the input MTJs to be read-out through sense amplifiers at the edge of the array. Rather, it is used locally within the set of MTJs involved in the computation. This is fundamentally why the CRAM computation represents true-in-memory computing: the computation does not require data to travel out of the memory array. It is always processed locally by nearby cells. We note that this concept would also work with other two-terminal stateful passive memory devices, such as memristors. Accordingly, a CRAM could be implemented with such devices. A CRAM could also be implemented with three-terminal stateful devices, such as spin-orbit torque (SOT) devices. This could result in greater energy efficiency and reliability 60 . Although devices with progressive or accumulative switching behavior, such as spintronic domain wall devices 61 , 62 , can be adopted as well, CRAM would otherwise work best if adopting bi-stable memory devices with strong threshold switching behavior. As an oversimplified speculation, the performance comparison between CRAMs implemented by various emerging memory devices is expected to roughly follow the comparison between these for memory applications, since CRAM utilizes memory devices in similar manners as in-memory applications. For example, a CRAM implemented based on MTJs should be expected to offer high endurance and high speed. Also, generally, a CRAM logic operation should consume energy comparable to the energy consumption of a memory write operation, for the same emerging memory device operating at the same speed. However, a careful case-by-case analysis is necessary for CRAMs implemented by each emerging memory device technology. Also, note that we do not show a specific circuit design of CRAM peripherals because CRAM does not require significant circuit design change in sense amplifiers or peripherals compared to 1T1M STT-MRAM. And these in the STT-MRAM are already common and mature. Lastly, the true in-memory computing characteristic of CRAM is limited to within a continuous CRAM array: any computation that requires access to data across separate CRAM arrays will require additional data access and movement. The size of an array is ultimately limited by parasitic effects of interconnects 63 . However, these limitations are true for all other in-memory computing paradigms. CRAM is not at any disadvantage in this scenario.

a CRAM adopts the so-called 2 transistor 1 MTJ (2T1M) cell architecture. On top of the 1T1M cell architecture of STT-MRAM, an additional transistor, as well as the added logic line (LL) and logic bit line (LBL), allow the CRAM to perform logic operations. During a CRAM logic operation, the transistors and lines are manipulated to form an equivalent circuit, as shown in b . Although CRAM can be built based on various emerging memory devices, we use MTJs and MTJ-based CRAM as an example for illustration purposes. b The working principle of CRAM logic operation, the VCL, utilizes the thresholding effect that occurs when switching an MTJ and the TMR effect of the MTJ. With an appropriate V logic amplitude, the voltage is translated into different currents flowing through the output MTJ by the TMR effect of the input MTJs. Whether the output MTJ switches or not is dependent on the state of the input MTJs.

On top of the potential performance benefits that the emerging memory devices bring, at circuit and architecture level, CRAM fundamentally provides several benefits (Fig. 1b ): (1) the elimination of the costly performance and energy penalties associated with transferring data between logic and memory; (2) random access of data for the inputs and outputs to operations; (3) the reconfigurability of operations, as any of the logic operations, AND, OR, NAND, NOR, and MAJ can be programmed; and (4) the performance gain of massive parallelism, as identical operations can be performed in parallel in each row of the CRAM array when data is allocated properly. Based on analysis and benchmarking, CRAM has the potential to deliver significant gains in performance and power efficiency, particularly for data-intensive, memory-centric, or power-sensitive applications, such as bioinformatics 40 , 64 , 65 , image 66 and signal 67 processing, neural networks 66 , 68 , and edge computing 69 (Fig. 1c ). For example, a CRAM-based machine-learning inference accelerator was estimated to achieve an improvement on the order of 1000× over a state-of-art solution, in terms of the energy-delay product 70 . Another example shows that CRAM (at the 10 nm technology node) consumes 0.47 µJ and 434 ns of energy and time, respectively, to perform an MNIST handwritten digit classifier task. It is 2500× and 1700× less in energy and time, respectively, compared to a near-memory processing system at the 16 nm technology node 66 . And yet, to date, there have been no experimental studies of CRAM.

In this work, we present the first experimental demonstration of a CRAM array. Although based on a small 1 × 7 array, it successfully shows complete CRAM array operations. We illustrate computation with a 1-bit full adder. This work provides a proof-of-concept as well as a platform with which to study key aspects of the technology experimentally. We provide detailed projections and guidelines for future CRAM design and development. Specifically, based on the experimental results, models and calculations of CRAM logic operations are developed and verified. The results connect the CRAM gate-level accuracy or error rate to MTJ TMR ratio, logic operation pulse width, and other parameters. Then we evaluate the accuracy of a multi-bit adder, a multiplier, and a matrix multiplication unit, which are essential building blocks for many conventional and machine intelligence applications, including artificial neural networks.

Experiments

Figure 3 shows the experimental setup, consisting of both hardware and software. The hardware is built with a so-called ‘circuit-around-die’ approach 71 : semiconductor circuitry is built with commercially available components around the MTJ dies. This approach offers a more rapid development cycle and flexibility needed for exploratory experimental studies on CRAM arrays and potential new MTJ technologies, while the major foundries lack the specific process design kit available for making a CRAM array fully integrated with CMOS. The hardware is a 1 × 7 CRAM array, with the design of cells taken from the 2T1M CRAM cells 38 , 40 , modified for simplified memory access. Software on a PC controls the operation. It communicates with the hardware with basic commands: ‘open/close transistors’; ‘apply voltage pulses’ to perform write and logic operations; and ‘read cell resistance’. The software collects real-time measurements of the data associated with CRAM operations for analysis and visualization. All aspects of the 1 × 7 CRAM array are functional: memory write, memory read, and logic operations (more details in Methods section, and Supplementary Note S 1 ).

The setup consists of custom-built hardware and a suite of control software. It demonstrates a fully functioning 1 × 7 CRAM array. The hardware consists of a main board hosting all necessary electronics except for the MTJ devices; a connection board on which passive switches, connectors, and magnetic bias field mechanisms are hosted; and multiple cartridge boards that each have an MTJ array mounted and multiple MTJ devices that are wire bonded. The gray-scale scanning electron microscopy image shows the MTJ array used. The color optical photographs show the cartridge board and the entire hardware setup. The software is responsible for real-time measurements of the MTJs; configuration and execution of CRAM operations: memory write, memory read, and logic; and data collection. It is run on a PC, which communicates wirelessly with the main board.

MTJs with perpendicular interfacial anisotropy are used in the CRAM. They exhibit low resistance-area (RA) product and high TMR ratio—approximately 100%—when sized at 100 nm in diameter (more details in Supplementary Note S 2 ).

Device properties and CRAM memory operations

The experiments begin with measuring the resistance (R)–voltage (V) properties of each MTJ device and of each die. In order to compensate for device-to-device variations, the bias magnetic fields for each MTJ are adjusted so that the R–V properties are as close to each other as possible (more details in Supplementary Note S 2 ). As the processes of making CRAM arrays mature, bias magnetic fields are expected to be no longer needed and all CRAM cells will be able to be operated with uniform parameters and under uniform conditions. The resistance threshold for the MTJs logic states is also determined in this stage.

Then the seven MTJ cells are tested for memory operations with various write pulse amplitudes and widths. Based on the observed write error rates for memory write operations, appropriate pulse amplitudes and widths are configured, achieving reliable memory write operations with an average write error rate of less than 1.5 × 10 −4 (more details in Supplementary Note S 3 ). We designate logic ‘0’ and ‘1’ to the parallel (P) low resistance state and anti-parallel (AP) high resistance state of MTJ, respectively.

CRAM logic operations

Two-input logic operations are studied. The output cell is first initialized by writing ‘0’ to it. Then two input cells are connected to the output cell through the LL by turning on the corresponding transistors. Voltage pulses of amplitude of V logic , V logic , and 0, are simultaneously applied to the two input cells and the output cell, respectively. This is the same as grounding the output cell while applying a voltage pulse of V logic to the two input cells. Then, depending on the input cells’ states, the output cell will have a certain probability of being switched from ‘0’ to ‘1’. Such a cycle of operations is repeated n times, and the statistical mean of the output logic state, < D out >, is obtained. The entire process is repeated for different V logic values and input states. The basis for logic operations in the CRAM is the state-dependent resistance of the input cells. These shift and displace the output cell’s switching probability transfer curve. As a result, the output cell switches state based on specific input states, thereby implementing a logic function such as AND, OR, NAND, NOR, or MAJ. A specific initial state of the output cell and V logic value corresponds to one of these logic gates 66 . The time duration or pulse width of the voltage pulse applied during a logic operation is expected to contribute to most of the time required to complete a logic operation. In the following, we use the term logic speed to generally refer to the speed of a logic operation. Logic speed is approximately inversely proportional to the time duration of the voltage pulse used during a logic operation.

The experimental results are shown in Fig. 4 a, b . Generally, for a given input state, < D out > increases with increasing V logic . The < D out > response curves are input state-dependent. The four input states can be divided into three groups:

The ‘00’ input state yields the lowest resistance at the two input cells, so the output cell switches from ‘0’ to ‘1’ first (with the lowest V logic ).

The ‘11’ input state yields the highest resistance at the two input cells, so the output cell switches from ‘0’ to ‘1’ last (with the highest V logic ).

The ‘01’ and ‘10’ input states both yield resistance that falls in between that of ‘00’ and ‘11’so that the output cell’s response curve falls in between that of ‘00’ and ‘11’.

a Output logic average, D out , vs. logic voltage, V logic . In a 2-input logic operation, two input cells and one output MTJ cell are involved. The output cell’s terminal is grounded, while the common line is left floating. A logic operation voltage pulse is applied to the two input cells’ terminals for a fixed duration (pulse width) of 1 ms. Before each logic operation, input data is written to the input cells. After each logic operation, the output cell’s state is read. Each curve corresponds to a specific input state. Each data point represents the statistical average of the output cell’s logic state, < D out >, sampled by 1000 repeats ( n = 1000) of the operations. The separation between the < D out > curves indicates the margins for NOR or NAND operation, highlighted in blue and red, respectively. b Accuracy of 2-input NAND operation vs. logic voltage, V logic . The results in a can be converted into a more straightforward metric, accuracy, for the NAND truth table. The curve labeled ‘mean’ and ‘worst’ indicates the average and the worst-case accuracy across all input states, respectively. So, for NAND operation, the optimal logic voltage is indicated in such a plot where the mean or worst accuracy is maximized. c , d Accuracy of MAJ3 ( c ) and MAJ5 ( d ) logic operations vs. logic voltage, V logic . Each curve corresponds to an input state or a group of input states. And each data point represents the statistical average of the output MTJ logic state sampled by n = 1000 and n = 250, for c and d , respectively.

Figure 4a shows the experiment results. The two regions highlighted in blue and red that fall in between the three groups of response curves are suitable for NOR and NAND operations, respectively. For example, in the red region, the ‘11’ input has a high probability of yielding a ‘0’ output, while the other three input states have a high probability of yielding a ‘1’ output. This matches the expected truth table for a NAND logic gate. Therefore, if V logic is chosen carefully—within the red region for the CRAM 2-input logic operation—the operation performed is highly likely to be NAND.

The experimental results of < D out > can be converted into a straightforward format representing the accuracy for a specified logic function. This translation can be computed by simply subtracting < D out > from 1 for those input states where a ‘0’ output is expected in the truth table of the logic function. Figure 4b shows the NAND accuracy of the same 2-input CRAM logic operation. The ‘mean’ and ‘worst’ plots are based on the average value and minimum value of the accuracy, respectively, across all input state combinations at a fixed value for V logic . Based on the experimental results, if V logic = 0.624 or 0.616 V, the CRAM delivers a NAND operation with a best mean and a worst-case accuracy of about 99.4% and 99.0%, respectively. From a circuit perspective, both increasing the effective TMR ratio of input cells and/or making the output cell’s response curve steeper would increase the vertical separation of these input state-dependent curves, resulting in higher accuracy. For example, a higher effective TMR ratio of input cells results in a larger contrast of current in the output cell between different input states. Therefore, there is more ‘horizontal’ room to separate the < D out > curves associated with different input states so that for the inputs with which the output is expected to be ‘0’ or ‘1’, the < D out > of the output cell is closer to the expected value (‘0’ or ‘1’). Also note that for a logic operation, the ‘accuracy’ and ‘error rate’ are essentially two quantities describing the same thing: how true the logic operation is, statistically. By definition, the sum of accuracy and error rate is always 1. The higher or closer to 1 the accuracy is, the better. The lower or closer to 0 the error rate is, the better. Lastly, to facilitate better visualization of how the resistance changes of different input cell states are translated into voltage differences on the output cell resulting in it being switched or unswitched, we list the equivalent resistance of the two input cells combined in parallel and the resulting voltage on the output cell as follows: With V logic = 0.620 V, the equivalent resistance of input cells and the resulting voltage on the output cell are 0.4133 V and 1120 Ω, 0.3753 V, and 1461 Ω, and 0.3248 V and 2037 Ω, for input states ‘00’, ‘01’ or ‘10’, and ‘11’, respectively. Note that these values are estimated by the experiment-based modeling, which is introduced in the later part of this paper.

With more input cells, we also studied 3-input and 5-input majority logic operations. Figure 4c shows the accuracy of a 3-input MAJ3 logic operation. At V logic = −0.464 V, both the optimal mean and the worst-case accuracy are observed to be 86.5% and 78.0%, respectively. Similarly, for a 5-input MAJ5 logic operation, shown in Fig. 4d , both the optimal mean and the worst-case accuracy are observed to be 75% and 56%, respectively. As expected, comparing 2-input, 3-input, and 5-input logic operations, the accuracy decreases with an increasing number of inputs (more discussions and explanations in Supplementary Note S 4 ).

CRAM full adder

Having demonstrated fundamental elements of CRAM—memory write operations, memory read operations, and logic operations—we turn to more complex operations. We demonstrate a 1-bit full adder. This device takes two 1-bit operands, A and B, as well as a 1-bit carry-in, C, as inputs, and outputs a 1-bit sum, S, and a 1-bit carry-out, C out . A variety of implementations exist. We investigate two common designs: (1) one that uses a combination of majority and inversion logic gates, which we will refer to as a ‘MAJ + NOT’ design; and (2) one that uses only NAND gates, which we will refer to as an ‘all-NAND’ design. Figures 5 a and 5b illustrate these designs. Supplementary Note S 5 provides more details.

a , b Illustrations of the ‘MAJ + NOT’ and ‘all-NAND’ 1-bit full adder designs. Green and orange letter symbols indicate input and output data for the full adder, respectively. From left to right, numbered by ‘logic step,’ each drawing shows the intended input (green rectangle) and output (orange rectangle) cells involved in the logic operation. The text in purple under each drawing indicates the intended function of the logic operation (MAJ3, NAND, or MAJ5). c – f Experimental ( c , d ) and simulation ( e , f ) results of the output accuracy of 1-bit full adder operations by CRAM with the MAJ + NOT ( c , e ) and all-NAND ( d , f ) designs. The CRAM adder’s outputs, S and C out , are assessed against the expected values, i.e., their truth table, for all input states of A, B, and C. The accuracy of each result for each input state is shown by the numerical value in black font, as well as, represented by the color of the box with red (or blue) indicating wrong (or correct), or accuracy of 0% (100%). The accuracy is calculated based on the statistical average of outputs obtained by repeating the full adder execution n times, for n = 10,000. The experimental results for the MAJ + NOT ( c ) and all-NAND ( d ) designs are obtained by repeatedly executing the operation for all input states and observing the output states. The simulation results for the MAJ + NOT ( e ) and all-NAND ( f ) designs are obtained with probabilistic modeling, using Monte Carlo methods. The accuracy of individual logic operations is set to what was observed experimentally.

Figure 5c, f shows the experimental and simulation results for the MAJ + NOT and the all-NAND designs, respectively. Each plot is a colormap that lists the accuracy of the output bits S and C out , with each input state coded as [ABC]. The blue (red) indicates good/desired (bad/undesired) accuracy. In the boxes of colormap, results in saturated blue are the most desirable. The numerical values of accuracy are also labeled accordingly. From the experimental results for the MAJ + NOT design full adder shown in Fig. 5c , we make two observations:

The accuracy of C out is generally higher than that of S. This is because C out is directly produced by the first MAJ3 operation from inputs A, B, and C, while S is produced after additional logic operations. We also note that since C out is produced earlier than S, it is less impacted by error propagation and accumulation during each step; and the MAJ5 involved in producing S is inherently less accurate than the MAJ3.

Both C out and S have higher accuracy when the input [ABC] = 000 or 111 than in the other cases. This is expected since the input states of all ‘0’s and all ‘1’s yield higher accuracy than those with mixed numbers of ‘0’s and ‘1’s for both MAJ3 and MAJ5.

The experimental results for the all-NAND design are shown in Fig. 5d . The same observations regarding accuracy vs. inputs as the MAJ + NOT design apply. However, it is clear that the accuracy of the all-NAND full adder, at 78.5%, is higher than that of the MAJ + NOT full adder, at 63.8%. This is likely due to the fact that 2-input NAND operations are inherently more accurate than MAJ3 and MAJ5 operations. This offsets the impact of the additional steps required in the all-NAND design. We note that the accuracy of all computation blocks will improve as the underlying MTJ technology evolves. Accordingly, the relative accuracy of the all-NAND versus the MAJ + NOT designs may change 66 .

Modeling and analysis of CRAM logic accuracy

To understand the origin of errors, how they accumulate, and how they propagate, we performed numerical simulations of the full adder designs. These are based on probabilistic models of logic operations, implemented using Monte Carlo methods. Figure 5 e, f shows the simulation results for the MAJ + NOT and all-NAND designs, respectively. In these simulations, the accuracy of individual logic operations was set to match what was experimentally observed. The simulation results for the overall designs of the full adders correspond well to what was observed experimentally for these, confirming the validity of the proposed probabilistic models (more details in the Methods section and Supplementary Note S 6 ).

We note that beyond the inherent inaccuracy of logic operations, other factors such as device drift and device-to-device variation in MTJ devices will contribute to error in a CRAM. Specifically, drifts in temperature, external magnetic field, MTJ anisotropy, and MTJ resistance can lead to drift of the response curve, < D out >. Most likely, any such drift will result in a reduction (increase) of accuracy (error rate). More discussion regarding device-to-device variation is provided in Supplementary Note S 7 .

On the other hand, the accuracy of logic operations will significantly benefit from improvements in TMR ratio as MTJ technology evolves. To project the future accuracy of CRAM operations, we employ various types of physical modeling informed by existing experimental results (more details are provided in the Methods section and Supplementary Note S 8 ).

Three sets of assumptions on the accuracies (or error rates) of NAND logic operations underlie the following studies.

The ‘experimental’ assumptions are based on the best accuracy experimentally observed among the 9 NAND steps involved with the all-NAND 1-bit full adder. These are adjusted linearly to ensure that the error for inputs ‘01’ and ‘10’ equals that for input ‘11’. In reality, as supported by the experimental results shown in Fig. 4a , such a condition can be reached by properly tuning the V logic . Therefore, assuming the gate-level error rate is already optimized by tuning the V logic , then the per-input state NAND accuracies can be further simplified so that an error rate, δ (0 ≤ δ ≤ 1), can be used to characterize the error, accuracy, and probabilistic truth table of NAND operations in a CRAM. The NAND accuracy is [1, 1–δ, 1–δ, 1–δ], and the NAND probabilistic truth table is [1, 1– δ , 1– δ , δ ], both being a function of δ. Through the above-mentioned modeling and calculations, the ‘experimental’ assumptions yield δ = 0.0076, which corresponds to a TMR ratio of approximately 109%, based on experiments.

Two additional sets of assumptions, labeled as ‘production’ and ‘improved’, assume MTJ TMR ratios of 200% and 300%, respectively. These two assumptions yield δ = 2.1 × 10 −4 , and δ = 7.6 × 10 −6 , respectively, based on modeling and calculations. The ‘production’ assumptions represent the current industry-level TMR ratios developed for STT-MRAM technologies. The ‘improved’ assumptions present reasonable expectations for near-future MTJ developments.

CRAM NAND error rates vs. TMR ratio with various logic voltage pulse widths are shown in Fig. 6a . Higher TMR ratios and faster logic speed—so shorter V logic pulse widths—lead to smaller error rates. Further details can be found in Supplementary Note S 8 and in Supplementary Figure S 5 . Also included is an analysis of error rates vs. effective TMR ratio, which is independent of the specific TMR modeling. Note that, for all subsequent results, we will use the NAND error rate at the assumed TMR ratios, with pulse widths of 1 ms. This is more conservative but is consistent with the pulse widths used in the experimental results reported above.

a NAND gate minimum error rate vs. MTJ TMR ratio with various V logic pulse widths. For a given TMR ratio, the error rate is a function of V logic . So, the ‘minimum error rate’ represents the minimum error rate achievable with an appropriate V logic value. All subsequent studies use the error rates observed with 1 ms pulse widths (to be consistent with the earlier experimental studies) at assumed TMR ratios. b The NED error of a 4-bit dot-product matrix multiplier vs. TMR ratio. TMR ratios of 109%, 200%, and 300% are adopted for the ‘experimental,’ ‘production,’ and ‘improved’ assumptions, respectively. The size of the input matrix is indicated in the legend of the plot.

Analysis of CRAM multi-bit adder, multiplier, and matrix multiplier

With these defined sets of assumptions, we provide projections of CRAM accuracy at a larger scale for meaningful applications. First, we evaluate ripple-carry adders and array multipliers 72 operating on scalar operands, with up to 6 bits. To evaluate the results, we adopt the normalized error distance (NED) metric 73 to represent the error of these primitives, as it has been shown to be more suitable for arithmetic primitives in the presence of computational error. We will refer to the error for a given primitive as ‘NED error’. We also define a complementary metric of ‘NED accuracy’ as the NED subtracted from 1 and then multiplied by 100%, to facilitate a more intuitive visualization of the error values. While the ‘experimental’ assumptions with a TMR ratio of 109% yield good overall accuracy for adders and multipliers, as the TMR ratio increases, the ‘production’ assumption with a TMR ratio of 200%, and the ‘improved’ assumption with a TMR ratio of 300%, yield significantly better or higher accuracies. Specifically, a 4-bit adder produces NED error of 2.8 × 10 −2 , 8.6 × 10 −4 , and 3.3 × 10 −5 , or NED accuracy of 97.2%, 99.914%, and 99.9967%, for the ‘experimental’, ‘production’, and ‘improved’ assumptions, respectively. A 4-bit multiplier produces NED error of 5.5 × 10 −2 , 1.8 × 10 −3 , and 6.6 × 10 −5 , or NED accuracy of 94.5%, 99.82%, and 99.9934%, for the three sets of assumptions, respectively. It is expected that when comparing the adder to the multiplier, since the latter is more complex and involves more gates, its accuracy is generally lower than that of the adder. Similarly, as the bit width of the adder or multiplier increases, their accuracy decreases. Further details and results with bit width up to 6-bit are provided in the Methods section and in Supplementary Note S 9 .

Then, using these primitives, we evaluate dot-product operations, which form the basis of matrix multiplication. They are heavily employed in many applications in both conventional domains and machine intelligence. Dot products consist of element-wise multiplication of two unsigned integer vectors, followed by addition. We perform additions with binary trees to maintain smaller circuit depth. Figure 6b shows the NED error of a 4-bit 4 × 4 dot-product matrix multiplier with respect to various TMR ratio assumptions. Like the adders and multipliers, as the TMR ratio increases, the NED error decreases, or the NED accuracy improves. Specifically, a 4-bit 4 × 4 dot-product matrix multiplier produces an NED error of 0.11, 3.4 × 10 −3 , and 1.2 × 10 −4 , or NED accuracy of 89%, 99.66%, and 99.988%, for the ‘experimental’, ‘production’, and ‘improved’ assumptions, respectively. Also, when comparing different input sizes (e.g., 1 × 1 to 4 × 4), as expected, the NED error is larger for larger input sizes due to the increased number of gates involved. Further details and results with bit width up to 5-bit are provided in the Methods section and in Supplementary Note S 9 .

Discussions

To summarize the experimental work, an MTJ-based 1 × 7 CRAM array hardware was experimentally demonstrated and systematically evaluated. The basic memory write and read operations of CRAM were achieved with high reliability. The study on CRAM logic operations began with 2-input logic operations. It was found that a 2-input NAND operation could be performed with accuracy as high as 99.4%. As the number of input cells was increased, for example, for 3-input MAJ3 and 5-input MAJ5 operations, the accuracy decreased to 86.5% and 75%, respectively. The decrease was attributed to having too many levels corresponding to the input states crowding a limited operating margin. Next, two versions of a 1-bit full adder were experimentally demonstrated using the 1 × 7 CRAM array: an all-NAND version and a MAJ + NOT version. The all-NAND design achieved an accuracy of 78.5%, while the seemingly simpler MAJ + NOT, which involves 3- and 5-input MAJ operations, only achieved an accuracy of 63.8%. Note that although each type of logic operation achieves optimal accuracy performance with a specific voltage value, the value is expected to only need to be static or constant. Therefore, only a finite number of power rails is needed to accommodate the logic operations of the CRAM array. Also, if the multiple logic pulse duration is allowed by a peripheral design, it is possible to operate the CRAM array with a single set of power rails for both memory write and logic operations.

A probabilistic model was proposed that accounts for the origin of errors, their propagation, and their accumulation during a multi-step CRAM operation. The model was shown to be effective when matched with the experimental results for the 1-bit full adder. The working principles of this model were adopted for the rest of the studies.

A suite of MTJ device circuit models was fitted to the existing experimental data and used to project CRAM NAND gate-level accuracy in the form of error rates. The gate-level error rates were shown to be 7.6 × 10 −6 , with reasonable expectations of TMR ratio improvement as MTJ technology develops. Other device properties, such as the switching probability transfer curve, could also significantly affect the CRAM gate-level error rate. This calls for improvements or new discoveries of the physical mechanisms for memory read-out and memory write. Error is an inherent property of any physical hardware, including CMOS logic components, which are commonly perceived as deterministic. As the development of CRAM proceeds, the gate-level error rate of CRAM will be further reduced over time. For now, while the error rate of CRAM is still higher compared to that of CMOS logic circuits, CRAM is naturally more suitable for applications that require less precision but can still benefit from the true-in-memory computing features and advantages of CRAM, instead of those that require high precision and determinism. Additionally, logic operations with many inputs, such as majority, may be desirable in certain scenarios. And yet, these were shown to have lower accuracy than 2-input operations. So, a tradeoff might exist.

Lastly, building on the experimental demonstration and evaluation of the 1-bit full adder designs, simulation and analysis were performed for larger functional circuits: scalar addition and multiplication up to 6 bits and matrix multiplication up to 5 bits with input size up to 4 × 4. These are essential building blocks for many conventional and machine intelligence applications. The parameters for the simulations were experimentally measured values as well as reasonable projections for future MTJ technology. The results show promising accuracy performance of CRAM at a functional building block level. Furthermore, as device technologies progress, improved performance or new switching mechanisms could further reduce the gate-level error rate of CRAM. Error correction techniques could also be employed to suppress CRAM gate errors.

In summary, this work serves as the first step in experimentally demonstrating the viability, feasibility, and realistic properties of MTJ-based CRAM hardware. Through modeling and simulation, it also lays out the foundation for a coherent view of CRAM, from the device physics level up to the application level. Prior work had established the potential of CRAM through numerical simulation only. Accordingly, there had been considerable interest in the unique features, speed, power, and energy benefits of the technology. This study puts the earlier work on a firm experimental footing, providing application-critical metrics of gate-level accuracy or error rate and linking it to the application accuracy. It paves the way for future work on large-scale applications, in conventional domains as well as new ones emerging in machine intelligence. It also indicates the possibility of making competitive large-scale CMOS-integrated CRAM hardware.

MTJ fabrication and preparation

The MTJ thin film stacks were grown by magnetron sputtering in a 12-source deposition system with a base pressure of 5 × 10 −9 Torr. The MgO barrier was fabricated by RF sputtering, while all the metallic layers were fabricated by DC sputtering. The stack structure is Si/SiO 2 /Ta(3)/Ru(6)/Ta(4)/Mo(1.2)/Co 20 Fe 60 B 20 (1)/MgO(0.9)/Co 20 Fe 60 B 20 (1.4)/Mo(1.9)/Ta(5)/Ru(7), where numbers in brackets indicate the thickness of the layer in nm. The stack was then annealed at 300 °C for 20 minutes in a rapid thermal annealing system under an Ar atmosphere (more information on the MTJ stack fabrication can be found in refs. 74 , 75 ).

The MTJ stacks were fabricated using three rounds of lithography similar to those described in ref. 76 . First, the bottom contacts were defined using photolithography followed by Ar+ ion milling etching. Then, the MTJ pillars were patterned into 120-nm circular nano-pillars via E-beam lithography and etched through Ar+ ion milling. After etching, SiO 2 was deposited via plasma-enhanced chemical vapor deposition (PECVD) to protect the nano-pillars. Finally, the top contacts were defined using photolithography, and the metallic electrodes of Ti (10 nm)/Au (100 nm) were deposited using electron beam evaporation.

The die of the MTJ array was diced into smaller pieces, with each piece containing about 10 MTJ devices. Each of the small pieces was mounted on a cartridge board, and up to 8 MTJ devices were wire-bonded to the electrodes of the cartridge board. Seven cartridge boards were inserted into the connection board, providing MTJs to the CRAM. The MTJ in each CRAM cell is selected among up to 8 MTJs on the corresponding cartridge board. In total, seven MTJs are selected from up to 56 MTJs. This method allows the user to find a collection of seven MTJs with minimum device-device variations.

CRAM experiment

An individual bias magnetic field was implemented for each of the seven MTJs on the connection board by positioning a permanent magnet at a certain distance from the MTJ devices. The bias magnetic field was used to compensate for intrinsic magnetic exchange bias and stray fields in the MTJ devices, thereby restoring the balance between the P and AP states. Additionally, slight rotation of bias field in the device plane was used to effectively adjust the switching voltage of each MTJ. More details can be found in Supplementary Note S 2 .

The connection board with seven MTJs was connected to the main board. On the main board, necessary active and passive electronic components were populated on the custom-designed PCB. The CRAM demo hardware circuit implemented a 1 × 7 CRAM array with a modified architecture to emphasize logic operations while compromising on memory operations bandwidth for simplicity. It was modified from the full-fledged 2T1M 40 architecture. It was equivalent to a 2T1M CRAM in logic mode, but it only had serial access to all cells for memory read and write operations (more details in Supplementary Note S 1 ). The hardware was powered by a battery and communicated with the controller PC wirelessly via Bluetooth®. In this way, the entire hardware was electrically isolated from the environment so that the risk of ESD to these sensitive MTJs was minimized.

The experiment control software running on a PC was implemented using National Instruments’ LabVIEW™. It was responsible for real-time measurements and control of the experiments, as well as necessary visualizations. Certain results were further analyzed post-experiment.

CRAM modeling and simulations

The simulation studies of accuracy as well as error origination, accumulation, and propagation began with a simple probabilistic model of each NAND logic operation. A probabilistic truth table was used to describe the expected statistical average of the output logical state. Then, the 1-bit full adder designs and operations were simulated using the Monte Carlo method with assumed probabilistic truth tables for each of the logic steps (see Supplementary Note S 6 ).

The experiment-based physics modeling and calculations for obtaining the projected CRAM logic operation accuracies began with an MTJ resistance-voltage model 77 , which was fit to the experimental data of TMR vs. bias voltage. The coefficients of this model were scaled accordingly to model projected TMR ratios higher than those observed experimentally. Then, a thermal activation model 78 , 79 of MTJ switching probability was fit to experimental data and was used to calculate the switching probability of the output MTJ cell under various bias voltages. Finally, the average of the output state, < D out >, could be calculated under various V logic , and the optimal NAND accuracies could be obtained in a manner similar to that discussed with Fig. 4 (more details in Supplementary Note S 8 ).

Further simulation studies of a ripple-carry adder, a systolic multiplier, and the dot-product operation of a matrix multiplication for various numbers of bits as well as matrix sizes were carried out using the same methods. More details can be found in Supplementary Note S 9 .

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files.

Lecun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521 , 436–444 (2015).

Article ADS CAS PubMed Google Scholar

Jordan, M. I. & Mitchell, T. M. Machine learning: trends, perspectives, and prospects. Science 349 , 255–260 (2015).

Article ADS MathSciNet CAS PubMed Google Scholar

Adomavicius, G. & Tuzhilin, A. Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions. IEEE Trans. Knowl. Data Eng. 17 , 734–749 (2005).

Article Google Scholar

Hinton, G. et al. Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Signal Process. Mag. 29 , 82–97 (2012).

Article ADS Google Scholar

Collobert, R. et al. Natural language processing (almost) from scratch. J. Mach. Learn. Res. 12 , 2493–2537 (2011).

Google Scholar

Krizhevsky, A., Sutskever, I. & Hinton, G. E. ImageNet classification with deep convolutional neural networks. Commun. ACM 60 , 84–90 (2017).

Oh, K. S. & Jung, K. GPU implementation of neural networks. Pattern Recognit. 37 , 1311–1314 (2004).

Strigl, D., Kofler, K. & Podlipnig, S. Performance and scalability of GPU-based convolutional neural networks. In 2010 18th Euromicro Conference on Parallel, Distributed and Network-based Processing 317–324 (IEEE, 2010).

Nurvitadhi, E. et al. Accelerating binarized neural networks: comparison of FPGA, CPU, GPU, and ASIC. In 2016 International Conference on Field-Programmable Technology (FPT) 77–84 (IEEE, 2017).

Sawada, J. et al. TrueNorth ecosystem for brain-inspired computing: scalable systems, software, and applications. In SC ’16: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis 130–141 (IEEE, 2016).

Jouppi, N. P. et al. In-datacenter performance analysis of a tensor processing unit. In Proceedings of the 44th Annual International Symposium on Computer Architecture 1–12 (ACM, 2017).

Chen, Y. H., Krishna, T., Emer, J. S. & Sze, V. Eyeriss: an energy-efficient reconfigurable accelerator for deep convolutional neural networks. IEEE J. Solid-State Circuits 52 , 127–138 (2017).

Yin, S. et al. A high energy efficient reconfigurable hybrid neural network processor for deep learning applications. IEEE J. Solid-State Circuits 53 , 968–982 (2018).

Borghetti, J. et al. Memristive switches enable stateful logic operations via material implication. Nature 464 , 873–876 (2010).

Chi, P. et al. PRIME: a novel processing-in-memory architecture for neural network computation in ReRAM-based main memory. In 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA) 27–39 (ACM, 2016).

Shafiee, A. et al. ISAAC: a convolutional neural network accelerator with in-situ analog arithmetic in crossbars. In 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA) , 14–26 (2016).

Hu, M. et al. Dot-product engine for neuromorphic computing: programming 1T1M crossbar to accelerate matrix-vector multiplication. In 2016 53nd ACM/EDAC/IEEE Design Automation Conference (DAC) 1–6 (IEEE, 2016).

Seshadri, V. et al. Ambit: in-memory accelerator for bulk bitwise operations using commodity DRAM technology. In 2017 50th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) 273–287 (IEEE, 2017).

Yao, P. et al. Fully hardware-implemented memristor convolutional neural network. Nature 577 , 641–646 (2020).

Jung, S. et al. A crossbar array of magnetoresistive memory devices for in-memory computing. Nature 601 , 211–216 (2022).

Keckler, S. W., Dally, W. J., Khailany, B., Garland, M. & Glasco, D. GPUs and the future of parallel computing. IEEE Micro 31 , 7–17 (2011).

Bergman, K. et al. ExaScale Computing Study: Technology Challenges in Achieving Exascale Systems . www.cse.nd.edu/Reports/2008/TR-2008-13.pdf (2008).

Horowitz, M. Computing’s energy problem (and what we can do about it). In 2014 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC) 10–14 (IEEE, 2014).

Kim, D., Kung, J., Chai, S., Yalamanchili, S. & Mukhopadhyay, S. Neurocube: a programmable digital neuromorphic architecture with high-density 3D memory. In 2016 ACM/IEEE 43rd Annual International Symposium on Computer Architecture (ISCA) , 380–392 (2016).

Huang, J. et al. Active-routing: compute on the way for near-data processing. In 2019 IEEE International Symposium on High Performance Computer Architecture (HPCA) 674–686 (IEEE, 2019).

Nair, R. et al. Active memory cube: a processing-in-memory architecture for exascale systems. IBM J. Res. Dev. 59 , 17:1–17:14 (2015).

Pawlowski, J. T. Hybrid memory cube (HMC). In 2011 IEEE Hot Chips 23 Symposium (HCS) 1–24 (IEEE, 2011).

Gao, M., Ayers, G. & Kozyrakis, C. Practical near-data processing for in-memory analytics frameworks. In 2015 International Conference on Parallel Architecture and Compilation (PACT) 113–124 (IEEE, 2015).

Gao, M., Pu, J., Yang, X., Horowitz, M. & Kozyrakis, C. TETRIS: scalable and efficient neural network acceleration with 3D memory. SIGARCH Comput. Arch. News 45 , 751–764 (2017).

Aga, S. et al. Compute caches. In 2017 IEEE International Symposium on High Performance Computer Architecture (HPCA) 481–492 (IEEE, 2017).

Zidan, M. A., Strachan, J. P. & Lu, W. D. The future of electronics based on memristive systems. Nat. Electron. 1 , 22–29 (2018).

Jeon, K., Ryu, J. J., Jeong, D. S. & Kim, G. H. Dot-product operation in crossbar array using a self-rectifying resistive device. Adv. Mater. Interfaces 9 , 2200392 (2022).

Article CAS Google Scholar

Matsunaga, S. et al. Fabrication of a nonvolatile full adder based on logic-in-memory architecture using magnetic tunnel junctions. Appl. Phys. Express 1 , 091301 (2008).

Hanyu, T. et al. Standby-power-free integrated circuits using MTJ-based VLSI computing. Proc. IEEE 104 , 1844–1863 (2016).

Li, S. et al. Pinatubo: a processing-in-memory architecture for bulk bitwise operations in emerging non-volatile memories. In 2016 53rd ACM/EDAC/IEEE Design Automation Conference (DAC) 1–6 (IEEE, 2016).

Kvatinsky, S. et al. Memristor-based material implication (IMPLY) logic: design principles and methodologies. IEEE Trans. Very Large Scale Integr. Syst. 22 , 2054–2066 (2014).

Kvatinsky, S. et al. MAGIC—memristor-aided logic. IEEE Trans. Circuits Syst. II Express Briefs 61 , 895–899 (2014).

Wang, J.-P. & Harms, J. D. General structure for computational random access memory (CRAM). US patent 14/259,568 (2015).

Gupta, S., Imani, M. & Rosing, T. FELIX: fast and energy-efficient logic in memory. In 2018 IEEE/ACM International Conference on Computer-Aided Design (ICCAD) 1–7 (IEEE, 2018).

Chowdhury, Z. et al. Efficient in-memory processing using spintronics. IEEE Comput. Archit. Lett. 17 , 42–46 (2018).

Gao, F., Tziantzioulis, G. & Wentzlaff, D. ComputeDRAM: in-memory compute using off-the-shelf DRAMs. In 2019 52nd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) 100–113 (IEEE, 2019).

Truong, M. S. Q. et al. RACER: Bit-pipelined processing using resistive memory. In MICRO-54: 54th Annual IEEE/ACM International Symposium on Microarchitecture 100–116 (ACM, 2021).

Žutić, I., Fabian, J. & Das Sarma, S. Spintronics: fundamentals and applications. Rev. Mod. Phys. 76 , 323–410 (2004).

Nikonov, D. E. & Young, I. A. Benchmarking of beyond-CMOS exploratory devices for logic integrated circuits. IEEE J. Explor. Solid-State Comput. Devices Circuits 1 , 3–11 (2015).

Lee, T. Y. et al. World-most energy-efficient MRAM technology for non-volatile RAM applications. In 2022 International Electron Devices Meeting (IEDM) 10.7.1–10.7.4 (IEEE, 2022).

Jan, G. et al. Demonstration of ultra-low voltage and ultra low power STT-MRAM designed for compatibility with 0x node embedded LLC applications. In 2018 IEEE Symposium on VLSI Technology 65–66 (IEEE, 2018).

Zhao, H. et al. Sub-200 ps spin transfer torque switching in in-plane magnetic tunnel junctions with interface perpendicular anisotropy. J. Phys. D. Appl. Phys. 45 , 025001 (2012).

Julliere, M. Tunneling between ferromagnetic films. Phys. Lett. A 54 , 225–226 (1975).

Parkin, S. S. P. et al. Giant tunnelling magnetoresistance at room temperature with MgO (100) tunnel barriers. Nat. Mater. 3 , 862–867 (2004).

Yuasa, S., Nagahama, T., Fukushima, A., Suzuki, Y. & Ando, K. Giant room-temperature magnetoresistance in single-crystal Fe/MgO/Fe magnetic tunnel junctions. Nat. Mater. 3 , 868–871 (2004).

Berger, L. Emission of spin waves by a magnetic mulitlayer traversed by a current. Phys. Rev. B 54 , 9353–9358 (1996).

Article ADS CAS Google Scholar

Slonczewski, J. C. Current-driven excitation of magnetic multilayers. J. Magn. Magn. Mater. 159 , L1–L7 (1996).

Wei, L. et al. A 7Mb STT-MRAM in 22FFL FinFET technology with 4ns read sensing time at 0.9V using write-verify-write scheme and offset-cancellation sensing technique. In 2019 IEEE International Solid- State Circuits Conference - (ISSCC) 214–216 (IEEE, 2019).

Gallagher, W. J. et al. 22nm STT-MRAM for reflow and automotive uses with high yield, reliability, and magnetic immunity and with performance and shielding options. In 2019 International Electron Devices Meeting (IEDM) 2.7.1-2.7.4 (IEEE, 2019).

Chih, Y. Der et al. A 22nm 32Mb embedded STT-MRAM with 10ns read speed, 1M cycle write endurance, 10 years retention at 150 °C and high immunity to magnetic field interference. In 2020 IEEE International Solid- State Circuits Conference - (ISSCC) 222–224 (IEEE, 2020).

Edelstein, D. et al. A 14 nm embedded STT-MRAM CMOS technology. In 2020 International Electron Devices Meeting (IEDM) 11.5.1-11.5.4 (IEEE, 2020).

Chun, K. C. et al. A scaling roadmap and performance evaluation of in-plane and perpendicular MTJ based STT-MRAMs for high-density cache memory. IEEE J. Solid-State Circuits 48 , 598–610 (2013).

Lilja, D. J. et al. Systems and methods for direct communication between magnetic tunnel junctions. US patent 13/475,544 (2014).

Lyle, A. et al. Direct communication between magnetic tunnel junctions for nonvolatile logic fan-out architecture. Appl. Phys. Lett. 97 , 152504 (2010).

Zabihi, M. et al. Using spin-Hall MTJs to build an energy-efficient in-memory computation platform. In 20th International Symposium on Quality Electronic Design (ISQED) 52–57 (IEEE, 2019).

Currivan-Incorvia, J. A. et al. Logic circuit prototypes for three-terminal magnetic tunnel junctions with mobile domain walls. Nat. Commun. 7 , 1–7 (2016).

Alamdar, M. et al. Domain wall-magnetic tunnel junction spin-orbit torque devices and circuits for in-memory computing. Appl. Phys. Lett. 118 , 112401 (2021).

Zabihi, M. et al. Analyzing the effects of interconnect parasitics in the STT CRAM in-memory computational platform. IEEE J. Explor. Solid-State Comput. Devices Circuits 6 , 71–79 (2020).

Chowdhury, Z. I. et al. A DNA read alignment accelerator based on computational RAM. IEEE J. Explor. Solid-State Comput. Devices Circuits 6 , 80–88 (2020).

Chowdhury, Z. I. et al. CRAM-Seq: accelerating RNA-Seq abundance quantification using computational RAM. IEEE Trans. Emerg. Top. Comput. 10 , 2055–2071 (2022).

Zabihi, M. et al. In-memory processing on the spintronic CRAM: from hardware design to application mapping. IEEE Trans. Comput. 68 , 1159–1173 (2019).

Article MathSciNet Google Scholar

Cilasun, H. et al. CRAFFT: High resolution FFT accelerator in spintronic computational RAM. In 2020 57th ACM/IEEE Design Automation Conference (DAC) 1–6 (IEEE, 2020).

Resch, S. et al. PIMBALL: Binary neural networks in spintronic memory. ACM Trans. Archit. Code Optim. 16 , 41 (2019).

Chowdhury, Z. I. et al. CAMeleon: reconfigurable B(T)CAM in computational RAM. In Proceedings of the 2021 on Great Lakes Symposium on VLSI 57–63 (ACM, 2021).

Resch, S. et al. MOUSE: inference in non-volatile memory for energy harvesting applications. In 2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) 400–414 (IEEE, 2020).

Lv, Y., Bloom, R. P. & Wang, J.-P. Experimental demonstration of probabilistic spin logic by magnetic tunnel junctions. IEEE Magn. Lett. 10 , 1–5 (2019).

Subathradevi, S. & Vennila, C. Systolic array multiplier for augmenting data center networks communication link. Cluster Comput. 22 , 13773–13783 (2019).

Liang, J., Han, J. & Lombardi, F. New metrics for the reliability of approximate and probabilistic adders. IEEE Trans. Comput. 62 , 1760–1771 (2013).

Almasi, H. et al. Perpendicular magnetic tunnel junction with W seed and capping layers. J. Appl. Phys. 121 , 153902 (2017).

Xu, M. et al. Voltage-controlled antiferromagnetism in magnetic tunnel junctions. Phys. Rev. Lett. 124 , 187701 (2020).

Lyu, D. et al. Sub-ns switching and cryogenic-temperature performance of mo-based perpendicular magnetic tunnel junctions. IEEE Electron Device Lett. 43 , 1215–1218 (2022).

Kim, J. et al. A technology-agnostic MTJ SPICE model with user-defined dimensions for STT-MRAM scalability studies. In 2015 IEEE Custom Integrated Circuits Conference (CICC) 1–4 (IEEE, 2015).

Diao, Z. et al. Spin-transfer torque switching in magnetic tunnel junctions and spin-transfer torque random access memory. J. Phys. Condens. Matter 19 , 165209 (2007).

Heindl, R., Rippard, W. H., Russek, S. E., Pufall, M. R. & Kos, A. B. Validity of the thermal activation model for spin-transfer torque switching in magnetic tunnel junctions. J. Appl. Phys. 109 , 073910 (2011).

Download references

Acknowledgements

This work was supported in part by the Defense Advanced Research Projects Agency (DARPA) via No. HR001117S0056-FP-042 “Advanced MTJs for computation in and near random-access memory” and by the National Institute of Standards and Technology. This work was supported in part by NSF SPX grant no. 1725420 and NSF ASCENT grant no. 2230963. The work at the University of Arizona is supported in part by NSF grant no. 2230124. The authors also thank Cisco Inc. for the support. Portions of this work were conducted in the Minnesota Nano Center, which was supported by the National Science Foundation through the National Nanotechnology Coordinated Infrastructure (NNCI) under Award No. ECCS-2025124. The authors acknowledge the Minnesota Supercomputing Institute (MSI, URL: http://www.msi.umn.edu ) at the University of Minnesota for providing resources that contributed to the research results reported within this paper. The authors thank Prof. Marc Riedel and Prof. John Sartori from the Department of Electrical and Computer Engineering at the University of Minnesota for proofreading the manuscript. Yang Lv, Brandon Zink, and Hüsrev Cılasun were CISCO Fellows.

Author information

Authors and affiliations.

Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN, 55455, USA

Yang Lv, Brandon R. Zink, Robert P. Bloom, Hüsrev Cılasun, Salonik Resch, Zamshed Chowdhury, Sachin S. Sapatnekar, Ulya Karpuzcu & Jian-Ping Wang

Department of Physics, University of Arizona, Tucson, Arizona, 85721, USA

Pravin Khanal, Ali Habiboglu & Weigang Wang

You can also search for this author in PubMed Google Scholar

Contributions

J.-P.W. conceived the CRAM research and coordinated the entire project. Y.L. and J.-P.W. designed the experiments. Y.L. and R.P.B. designed and developed the demonstration hardware and software. P.K., A.H., and W.W. grew part of the perpendicular MTJ stacks. B.R.Z. fabricated the MTJ nanodevices. Y.L. conducted the CRAM demonstration experiments and analyzed the results. Y.L. studied the probabilistic model of CRAM operations and conducted simulations of a 1-bit full adder. Y.L., B.R.Z., and R.P.B. developed the device physics modeling of CRAM logic operations and gate-level error rates and conducted related calculations. H.C., S.R., Z.C., and U.K. carried out the simulation studies of the multi-bit adder, multiplier, and matrix multiplication. S.S. participated in discussions of modeling and simulation. All authors reviewed and discussed the results. Y.L. and J.-P.W. wrote the draft manuscript. All authors contributed to the completion of the manuscript.

Corresponding author

Correspondence to Jian-Ping Wang .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Lv, Y., Zink, B.R., Bloom, R.P. et al. Experimental demonstration of magnetic tunnel junction-based computational random-access memory. npj Unconv. Comput. 1 , 3 (2024). https://doi.org/10.1038/s44335-024-00003-3

Download citation

Received : 29 January 2024

Accepted : 29 May 2024

Published : 25 July 2024

DOI : https://doi.org/10.1038/s44335-024-00003-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

Explore articles by subject
Guide to authors
Editorial policies

Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Publications
Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

Advanced Search
Journal List
v.11(4); Oct-Dec 2021

Sugen, hypoxia and the lung circulation

Twenty years ago, the paper was published that first described the Sugen Hypoxia rat model, a novel experimental model of pulmonary vascular disease and severe pulmonary arterial hypertension (PAH). 1 With this short letter, we highlight some advantages and disadvantages inherent to the model and provide possible directions for future research. This piece is not intended as a comprehensive review of the model, but rather aims to celebrate two decades of work by many groups who have worked with an experimental model that was created by serendipity. We aim foremost to encourage all interested in the field to go back at the original paper and assess the road traveled since then.

The Sugen Hypoxia model first appeared in a paper published in the FASEB Journal, 1 after the manuscript had been rejected by the major pulmonary journals. The rejection of a manuscript describing a novel model is not too surprising and should not encourage young investigators to give up on their work but urge them to continue to find ways to get their data published. At the time of this writing, the paper has been cited more than 800 times (in Google Scholar, while cited 295 times in PubMed) and the model is established in many laboratories around the world. With passage of time, many investigators have stopped referencing the original publication and this may indicate that the model has become common place, a standard model like the chronic hypoxia model and the monocrotaline model.

In reminiscing, it may be of more than historical interest to illustrate how contentious the early acceptance of this model of severe pulmonary arterial hypertension PAH model was. After all, it was based on a wrong hypothesis, and the team of investigators was caught by surprise when examining the lung histology for the first time, being confronted with an obliterative pulmonary vasculopathy 1 that over time progresses to a plexogenic arteriopathy with striking resemblance to human PAH. 2 The question asked most often by reviewers and conference attendees was: “You are treating the animals with an inhibitor of angiogenesis, and then you end up with angiogenesis?” Because chronic hypoxia, via hypoxia inducible factor (HIF)-1alpha, increased the expression of vascular endothelial growth factor (VEGF) in the rat lungs and VEGF was also highly expressed in the plexiform lesions in the lung tissue from patients with idiopathic PAH, 3 the hypothesis was that the VEGF receptor blocker Sugen 5416 would inhibit chronic hypoxic pulmonary hypertension. That did not happen and the investigators needed to explain why severe angioproliferative PAH developed.

The Sugen/hypoxia model of angioobliterative PAH was not the first rodent model that had produced obliterative pulmonary vascular lesions. The pneumonectomy-monocrotaline model 4 did not gain wide acceptance, likely because of the technical problems with the surgery and also because it did not provide a new hypothesis. The Sugen/hypoxia model generated interest because the first publication had identified the lumen-obliterating cells as endothelial cells and had documented that initial (early) apoptosis was necessary for the pathogenesis, because treatment with a pan-caspase inhibitor prevented disease development. Without initial apoptosis no subsequent endothelial cell growth. 5 Follow-up publications provided evidence that the Sugen/hypoxia induced disease was remarkably restricted to the lung circulation, that targeted therapy had no influence on the pulmonary vascular lesions and that the number of obliterated small lung vessels predicted the level of right ventricular systolic pressure elevation. 6 , 7 With passage of time the rat model was further explored, and one important question was: must in this two hit model the VEGF receptor blocker be paired with hypoxia as an obligatory second hit? This answer was soon provided as it was shown that instead of hypoxia, pneumonectomy 8 and immune insufficiency in the athymic rat 9 could substitute for hypoxia. Reviewers had frequently questioned whether therapies targeting VEGF signaling in human patients would affect the lung circulation. More recently, data on cancer patients treated with the VEGF receptor antibody bevacizumab provided some clinical context for the “angiogenesis paradox” of PAH in the Sugen PAH models, 10 as it was shown that treatment of cancer patients with bevacizumab can induce proliferative pulmonary vascular changes. 11

The Sugen/hypoxia model also allows to study right heart failure and its relationship with myocardial capillary rarefaction. 12 The authors of the 2001 publication were concerned whether indeed the pathobiology which leads to severe PAH in this model could be solely explained by inhibited VEGF signaling. “Whether the action of Sugen 5416 is exclusively via VEGFR-2 blockade or via other effects is of great importance for the interpretation of our data and the mechanisms involved in the development of pulmonary hypertension.”

One component of the model highlighted more recently is the high lung tissue expression of genes encoding proteins of the arylhydrocarbon receptor (AhR)-cytochrome P450 axis. 13 See also Masaki et al. 14 and Dean et al. 15 for further reading on the role of the AhR in PAH. This aspect of the model continues to be generally overlooked but needs to be considered as a potentially important contributor to the vascular cell proliferation and cell phenotype changes which characterize this model. In addition, Sitapara et al. reported that Sugen 5416 affects BMPRII signaling, 16 providing yet another possible explanation for the development of PAH like vascular lesions after the double challenge of Sugen and hypoxia.

Now, 20 years later, investigators are still divided: some are critical and believe that the model does not represent all of the important features of the severe forms of human PAH, 17 others continue to work with the model as a disease-relevant model for the preclinical testing of novel therapies. 18 Today, as at the time of its first publication, 1 the Sugen-hypoxia model is still only partially understood and remains controversial. Funding agencies and journal reviewers have criticized a lack of mechanistic insight that comes with experimental rat models. Many have insisted that the future of pulmonary vascular research lies in studying genetically engineered mice? Apropos, the mouse. It is important to note that treatment of mice with Sugen 5416 does not recapitulate the lung pathology observed in rats, perhaps because the two rodent species do not share the same repertoire of cytochrome P450 genes. 13

We remain convinced that the Sugen-based models will continue to teach us, not just about the transition from the initial pulmonary endothelial cell damage (how does VEGF receptor blockade and/or activation of the AhR-cytochrome axis kill endothelial cells?) to exuberant proliferation of phenotypically altered cells (how do vascular cells become apoptosis-resistant?). The models can teach us about pulmonary intravascular inflammation and lung vascular immunity and right heart failure (how important is the impaired myocardial microcirculation?).

Beginning in 2000, when Peter Hirth at Sugen in South San Francisco provided the first batch of SU5416, until today, it has been an interesting journey. In our opinion, there are important questions that work with this model can answer. Here are just a few: Because the Sugen/hypoxia model is a “two hit model,” as are the subsequently developed Sugen-based rat models, 8 , 9 , 19 it is critical for a deeper understanding of the pathogenesis to investigate how two hits interact on a cellular and molecular level. Of note, it is generally accepted that for human severe PAH to develop, also two hits are necessary and one of which can be genetic. The Sugen model of a genetically susceptible rat strain which develops severe PAH without hypoxia, 20 could perhaps be a useful tool for such investigations.

While the PAH in Sugen/hypoxia model is unresponsive to conventional vasodilator treatments, 6 fasudil injection in anesthetized rats with established severe PAH revealed that—in spite of widespread pulmonary vessel obliteration—the pulmonary artery pressure could be acutely decreased. 7 The mechanism of this vaso-relaxing effect of fasudil is not understood. Investigation of the mechanism of action of this drug may lead to the discovery of new classes of pulmonary vasodilators.

Finally, it appears to us that intact signaling through the VEGF→Akt axis is critically important for the homeostasis of pulmonary vascular endothelial cells. In the Sugen/hypoxia model, treatment with Sugen5416 resulted in a significantly decreased expression of the genes encoding VEGF, VEGFR-2 and phospho Akt .Which are the feed-forward mechanisms that cause this effect?

Perhaps the next 20 years of investigating the Sugen models will provide some of the answers.

Conflict of Interest: The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors received speaking fees from Janssen and MSD; grant support from Janssen, MSD and Ferrer. The authors have served on advisory boards for Jannssen, VIVUS and Polarean.

Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: HJB was supported by the Dutch Cardiovascular Research Alliance, through the DOLPHIN-GENESIS and PHAEDRA-IMPACT grants. Research support includes Phaedra consortium 2012-17 and 2018-2023/Dutch CardioVascular Alliance (DCVA).

Author contributions: Both authors contributed equally to the concept, writing and revisions of this manuscript.

Guarantor: HJB.

ORCID iD: Harm J. Bogaard https://orcid.org/0000-0001-5371-0346

Design and experimental study of biomimetic microtextured hard alloy cutting tools based on sparse matrix and hydrodynamic lubrication theory

Metals & corrosion
Published: 05 August 2024

Cite this article

Zhixiong Tang 1 ,
Zhenghao Ge ORCID: orcid.org/0009-0002-7618-1576 1 &
Jie Li 1 , 2

Friction between the rake face of the tool and the chip is a primary contributor to tool rake face wear. Hence, reducing the friction coefficient between the tool’s rake face and the chip proves to be an effective method for wear mitigation. To minimize the friction coefficient of cutting tools, distinctive microtextures were laser-engraved onto the rake face, employing hydrodynamic lubrication principles. To minimize the adverse impact of microtextures on the tool and simultaneously achieve hydrodynamic lubrication, a sparse matrix arrangement method was introduced for microtexture placement. For further enhancement of the hydrodynamic lubrication effect, finite element analysis was utilized to optimize both the depth and arrangement of microtextures. Friction and cutting experiments were performed to validate the lubrication efficiency of biomimetic microtextures, with an accompanying analysis of the lubrication mechanism. In summary, biomimetic microtextures successfully diminish friction between the tool’s rake face and the chip, resulting in a significant reduction in three-dimensional cutting forces, notably decreasing the main cutting force and accomplishing the objective of reducing tool wear.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save.

Get 10 units per month
Download Article/Chapter or eBook
1 Unit = 1 Article or 1 Chapter
Cancel anytime

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Data availability

Data availability is not applicable to this article as no new data were created or analyzed in this study.

Ye W, Zhou Q, Shi Y et al (2024) Robust wear performance of graphene-reinforced high entropy alloy composites. Carbon. https://doi.org/10.1016/j.carbon.2024.119040

Article Google Scholar

Ren Y, Zhou Q, Hua D et al (2024) Wear-resistant CoCrNi multi-principal element alloy at cryogenic temperature. Sci Bull 69(2):227–236

Article CAS Google Scholar

Dongpeng H, Qing Z, Yeran S et al (2023) Revealing the deformation mechanisms of 〈110〉 symmetric tilt grain boundaries in CoCrNi medium-entropy alloy. Int J Plast 171:171103832

Google Scholar

Wu F, Liu N, Ma Y et al (2022) Research on the influence of diamond coating microtexture on graphitization law and friction coefficient. Diam Relat Mater 127:109153

Lian AY, Xie AC, Mu AC et al (2019) Preparation technology of micro-textured tools fabricated by inductively coupled plasma etching. Surf Coat Technol 370:177–186

Gropper D, Harvey TJ, Wang LA (2018) Numerical model for design and optimization of surface textures for tilting pad thrust bearings. Tribol Int 119:190–207

Costa HL, Hsu SM et al (2017) A critical assessment of surface texturing for friction and wear improvement. Wear Int J Sci Technol Frict Lubr Wear 372:21

Rosenkranz A, Costa HL, Baykara MZ et al (2021) Synergetic effects of surface texturing and solid lubricants to tailor friction and wear–A review. Elsevier 151:106792

Wu Z, Bao H, Liu L et al (2020) Numerical investigation of the performance of micro-textured cutting tools in the cutting of Ti-6Al-4V alloys. I J Adv Manuf Technol 108(10):463

Pang KWD (2020) Study on the performances of the drilling process of nickel-based superalloy Inconel 718 with differently micro-textured drilling tools. Int J Mech Sci 180(1):105658

Zhang K, Deng J, Sun J et al (2015) Effect of micro/nano-scale textures on anti-adhesive wear properties of WC/Co-based TiAlN coated tools in AISI 316 austenitic stainless steel cutting. Appl Surf Sci 355(15):602–614

Tomar AK, Sharma SC (2020) An investigation into surface texture effect on hole-entry hybrid spherical journal bearing performance. Tribol Int 151:106417

Xuan M, Jane Wang Q, Xiqun Mehta L, Viral S (2019) Piston surface design to improve the lubrication performance of a swash plate pump. Tribol Int 132:275

Yue H, Deng J, Ge D et al (2019) Effect of surface texturing on the tribological performance of sliding guideway under boundary lubrication. J Manuf Process 47:172–182. https://doi.org/10.1016/j.jmapro.2019.09.031

Fei D, Yun-Peng DU , Yong-Hong FU , et al. (2024) Simulation and experiment study on laser surface micro-texturing on cylinder liner-piston ring friction pair of internal combustion engine. Chinese Internal Combustion Engine Engineering [2024-01-18]

Chen L, Liu Z, Shen Q (2018) Enhancing tribological performance by anodizing micro-textured surfaces with nano-MoS 2 coatings prepared on aluminum-silicon alloys. Tribol Int 122:84–95

Li N, Xu E, Liu Z et al (2016) Tuning apparent friction coefficient by controlled patterning bulk metallic glasses surfaces. Sci Rep 6(1):39388

Article CAS PubMed PubMed Central Google Scholar

Khanali M, Farahnakian M, Elhami S et al (2022) Tribological properties of vibro-mechanical texturing during face-turning processes. Int J Lightweight Mater Manuf 5(1):91–101

CAS Google Scholar

Marchetto D et al (2008) AFM investigation of tribological properties of nano-patterned silicon surface—ScienceDirect. Wear 265(5–6):577–582

Pettersson U, Jacobson S (2004) Friction and wear properties of micro textured DLC coated surfaces in boundary lubricated sliding. Tribol Lett 17(3):553–559

Walker JC, Kamps TJ et al (2017) Tribological behavior of an electrochemical jet machined textured Al-Si automotive cylinder liner material. Wear 376:1611

Earl C, Castrejón-Pita JR, Hilton PA et al (2016) The dynamics of laser surface modification. J Manuf Process 21:214–223

Mao B, Siddaiah A, Liao Y et al (2020) Laser surface texturing and related techniques for enhancing the tribological performance of engineering materials: a review. J Manuf Process 53:153–173

Yu SY, Liu GD, Wu YT et al (2023) Morphological evolution of Laser-Induced concave/convex Micro-Textures on 304 stainless steel. Appl Laser 43(04):73–79

Wu Z, Xing Y, Huang P et al (2017) Tribological properties of dimple-textured titanium alloys under dry sliding contact. Surf Coat Technol 309:21–28

Yu H, Wang X, Zhou F (2010) Geometric shape effects of surface texture on the generation of hydrodynamic pressure between conformal contacting surfaces. Tribol Lett 37:123–130

Wang X, Adachi K, Otsuka K et al (2006) Optimization of the surface texture for silicon carbide sliding in water. Appl Surf Sci 253(3):1282–1286

Gropper D, Wang L, Harvey TJ (2016) Hydrodynamic lubrication of textured surfaces: a review of modeling techniques and key findings. Tribol Int 94:509–529

He C, Yang S, Zheng M (2022) Analysis of synergistic friction reduction effect on micro-textured cemented carbide surface by laser processing. Opt Laser Technol 155:108343

Wu F, Zhang N, Peng W et al (2023) A novel hybrid micro-texture for improving the wear resistance of PCD tools on cutting SiCp/Al composites. J Manuf Process 101:930–942

Tong X, Gou Y, Yu S (2021) Anti-wear and anti-friction performance of cemented carbide under the interaction of micro-circular texture preparation and geometric parameters. Mater Sci Eng B 268:115116

Wang Q, Yang Y, Yao P et al (2021) Friction and cutting characteristics of micro-textured diamond tools fabricated with femtosecond laser. Tribol Int 154:106720

Orra K, Choudhury SK (2018) Tribological aspects of various geometrically shaped micro-textures on cutting insert to improve tool life in hard turning process. J Manuf Process 31:502–513

Hou Q, Yang X, Li D et al (2022) Tribological performance of hydrophobic and micro/nano triangle textured rake face of cutting tools. Appl Surf Sci 571:151250

Download references

Acknowledgements

The authors greatly acknowledge National Key Research and Development Program under (Grant 2022YFB3205801), The Major Science and Technology Project of Anhui Province (Grant 2022e03020002), Natural Science Basic Research Program of Shaanxi (Grant 2024JC-YBQN-0455), Open Research Fund of State Key Laboratory of Mechanical Manufacturing Systems Engineering (Grant sklms2022016) for the financial and technical support of the present study.

Author information

Authors and affiliations.

College of Mechanical and Electrical Engineering, Shaanxi University of Science and Technology, Xi’an, 710021, Shaanxi, China

Zhixiong Tang, Zhenghao Ge & Jie Li

State Key Laboratory for Manufacturing Systems Engineering, International Joint Laboratory for Micro/Nano Manufacturing and Measurement Technologies, Xi’an Jiaotong University, Xi’an, 710049, China

You can also search for this author in PubMed Google Scholar

Contributions

All authors contributed to the conception and design of the review. All authors approve of the final version to be submitted.

Corresponding author

Correspondence to Zhenghao Ge .

Ethics declarations

Conflict of interest.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Handling Editor: Zhao Shen.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Tang, Z., Ge, Z. & Li, J. Design and experimental study of biomimetic microtextured hard alloy cutting tools based on sparse matrix and hydrodynamic lubrication theory. J Mater Sci (2024). https://doi.org/10.1007/s10853-024-10012-z

Download citation

Received : 11 May 2024

Accepted : 12 July 2024

Published : 05 August 2024

DOI : https://doi.org/10.1007/s10853-024-10012-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Find a journal
Publish with us
Track your research

Algal Blooms & Hypoxia

Experimental Lake Erie Hypoxia Forecast

Forecast date: 2020-10-23.

The Experimental Lake Erie Hypoxia forecast is updated daily, and uses a hydrodynamic model and weather forecast data to predict the location and movement of hypoxic bottom water up to five days in advance. Hypoxia is a condition of low dissolved oxygen (

The Experimental Lake Erie Hypoxia Forecast is transitioning to operations at NOAA’s National Centers for Coastal Ocean Science (NCCOS). In support of the transition, GLERL will continue to update this webpage through the 2024 hypoxia season, but NOAA is directing forecast users to the new operational web page . Email updates regarding the hypoxia forecast and predicted hypoxic upwelling events will come from NCCOS. For additional information, or to be added to the email list, contact [email protected].

--> animation of forecasted change maps below

The Experimental Lake Erie Hypoxia forecast is a forecast model in development for transition to NOAA operations. Other panels provide additional information on lake conditions. For the latest images, refresh your browser and/or clear its cache.

Forecast bottom temperature and dissolved oxygen - This panel depicts modeled lake bottom temperatures in degrees Celcius and bottom dissolved oxygen, and also a cross-sectional view of Lake Erie forecast model data from Cleveland to Port Stanley (Canada). Red color is associated with higher values, and blue color with lower values, according to the color scale that appears beside the plot. Observations are plotted as white-edged circles, filled with a color according to the color scale. These observations are provided by the Great Lakes Observing System , Ontario Ministry of the Environment, Conservation and Parks , the Regional Science Consortium , the Real-Time Aquatic Ecosystem Observation Network (RAEON) at the University of Windsor , and collaborations with regional operators of water treatment plants including the Cleveland Water Department , Lake County Department of Utilities , as well as the Cleveland Water Alliance , LimnoTech , and Freeboard Technology (These links are external to NOAA-GLERL).

Forecast change in bottom temperature and dissolved oxygen - This panel depicts the Experimental Lake Erie Hypoxia 4-day Forecast. The color scale indicates changes in dissolved oxygen (D.O.) concentrations over a period of 4 days, beginning at noon yesterday (EDT). Areas in green indicate no change in dissolved oxygen levels, while areas in blue indicate a decrease and areas in red indicate an increase.

Wind Rose - This panel depicts the frequency of occurrence of wind directions and speeds for yesterday, today, tomorrow, and the following 3 days. The length of each spoke indicates how frequently a wind blows from a particular direction. Wind speeds are indicated by color, as given by the color scale at the bottom of the plot. The data were sampled from the wind data used to drive the hydrodynamic model simulation at locations around the central basin of Lake Erie, for additional detail see the Great Lakes Coastal Forecasting System (GLCFS)

We make every effort to ensure accuracy and minimize downtime, however, we cannot guarantee that data and products will always be available and/or accurate. Our forecasts and products should be considered experimental. Please review our

Senior Research Scientist
NOAA GLERL
(734) 741-2268 -->

(734)741-2235

Research Physical Scientist
NOAA GLERL
(734) 741-2172

Info Services Branch Chief
NOAA GLERL
(734) 741-2210

IMAGES

Experimental models for the investigation of hypoxia-induced modulation
Figure 1 from Design of Experimental Hypoxia Chamber for Cell Culture
Establishment of hypoxia induction in an in vivo animal replacement
Experimental design of the study. a Severe hypobaric hypoxia model. b
Model of chronic intermittent hypoxia and the experimental protocol
Fitting model to experimental data under hypoxia. Model time-course

VIDEO

Hypoxia Training CAMI/PROTE
Hypoxia and Hypoxemia Video No 357
Response to hypoxia inflight / Private Pilot
Benefits of controlled hypoxia in breathwork
HYPOXIA AND HYPOMAXIA 🫀🫁MEDICAL TERMINOLOGY#nursing #facts #nursingstudent ‎@Nursingstudent_GNM
Hypoxia? This can KILL everyone on board an aircraft! ✈️ #pilotlife #aviation

COMMENTS

Hypoxia modeling techniques: A review
Since each type of hypoxia can be a cause and a consequence of various physiological changes, the methods for modeling these hypoxias are also different. There are many techniques for modeling hypoxia under experimental conditions. The most common animal for modeling hypoxia is a rat.
Hypoxia modeling techniques: A review
Since each type of hypoxia can be a cause and a consequence of various physiological changes, the methods for modeling these hypoxias are also different. There are many techniques for modeling hypoxia under experimental conditions. The most common animal for modeling hypoxia is a rat.
Induction and Testing of Hypoxia in Cell Culture
Hypoxia is defined as the reduction or lack of oxygen in organs, tissues, or cells. This decrease of oxygen tension can be due to a reduced supply in oxygen (causes include insufficient blood vessel network, defective blood vessel, and anemia) or to ...
Technical Feasibility and Physiological Relevance of Hypoxic Cell
A search of the current literature on the topic revealed this was the case for many original studies pertaining to experimental models of hypoxia in vitro. Therefore, in this review, we present evidence mandating for the close control of oxygen levels in cell culture models of hypoxia.
The use of cobalt chloride as a chemical hypoxia model
The use of hypoxia models in cell culture has allowed the characterization of the hypoxia response at the cellular, biochemical and molecular levels. Although a decrease in oxygen concentration is the optimal hypoxia model, the problem faced by many researchers is access to a hypoxia chamber or a CO …
Technical Feasibility and Physiological Relevance of Hypoxic ...
A search of the current literature on the topic revealed this was the case for many original studies pertaining to experimental models of hypoxia in vitro. Therefore, in this review, we present evidence mandating for the close control of oxygen levels in cell culture models of hypoxia.
Human hypoxia models in aerospace medicine: Potential applications for
Aerospace medicine required controlled terrestrial models to investigate influences of altered atmosphere conditions, such as hypoxia, on human health and performance. These models could potentially ...
Hypoxia modeling techniques: A review
Since each type of hypoxia can be a cause and a consequence of various physiological changes, the methods for modeling these hypoxias are also different. There are many techniques for modeling hypoxia under experimental conditions. The most common animal for modeling hypoxia is a rat.
A simplified protocol to induce hypoxia in a standard incubator: A
Here, we describe simplified protocols for stabilizing cellular hypoxia-inducible factor-1α (HIF-1α) in cell culture using either a hypoxia chamber or CoCl 2. In addition, we also provide a detailed methodology to confirm hypoxia induction by the assessment of protein levels of HIF-1α, which accumulates in response to hypoxic conditions.
Nucleotide depletion in hypoxia experimental models of mouse myocardial
Experimental models to test the effective protection against cardiac ischemia injury are still challenging in pre-clinical studies. The use of myocardial slices creates a special link between testi...
Experimental Hypoxia as a Model for Cardiac Regeneration in Mice
Experimental hypoxia has been used for decades to examine the adaptive response to low-oxygen environments. Various models have been studied, including flies, worms, ...
Hypoxia-induced metastasis model in embryonic zebrafish
The average experimental time for this model is 7 d. Our protocol offers a remarkable opportunity to study molecular mechanisms of hypoxia-induced cancer metastasis.
Experimental Hypoxia
Numerous experimental hypoxia or ischemia/reperfusion models have suggested that oxygen radicals cause injury in the myocardium from mitochondrial electron transport.
The Latest in Animal Models of Pulmonary Hypertension and Right
To create a model presenting this characteristic, MCT has been combined with either chronic hypoxia exposure 19 or left pneumonectomy. 20 In these 2-hit models, a neointimal pattern of remodeling and severe RV hypertrophy occur providing a model that closely mimics the advanced stage of the disease.
Hypoxia: molecular pathophysiological mechanisms in human diseases
Recent studies on experimental models focused on the effect of hypoxia in tissue regeneration and repair. Yuji Nakada et al. reported that mice exposed to a state of systemic hypoxia inhibited the oxidative metabolism, decreased the production of ROS and oxidative damage to DNA, and reactivated the mitosis of cardiomyocytes.
PDF doi: 10.1007/978-1-0716-0668-1_25
Experimental hypoxia has been used for decades to examine the adaptive response to low-oxygen environ-ments. Various models have been studied, including flies, worms, fish, rodents, and humans.
Animal Models of Brain Hypoxia
Hypoxia (i.e., reduced oxygen availability) is a classical model of the metabolic encephalopathies or delirium. An understanding of how hypoxia alters brain function has implications for understanding other metabolic encephalopathies as well as aging and age-related...
Novel approach to study hypoxia enables identification of a marker for
Hypoxia is thus associated with a poor prognosis ... Experimental AI method boosts doctors' ability to diagnose cancers and precancers of the esophagus ... patient-derived neurons accurately model ...
Experimental Hypoxia as a Model for Cardiac Regeneration in Mice
Experimental hypoxia has been used for decades to examine the adaptive response to low-oxygen environments. Various models have been studied, including flies, worms, fish, rodents, and humans. Our lab has recently used this technology to examine the effect of environmental hypoxia on mammalian heart regeneration.
Experimental Lake Erie Hypoxia Forecast
The Experimental Lake Erie Hypoxia forecast is updated daily, and uses a hydrodynamic model and weather forecast data to predict the location and movement of hypoxic bottom water up to five days in advance. Hypoxia is a condition of low dissolved oxygen (< 2 mg/L) that can be harmful to aquatic life. Hypoxic upwelling events along the coast can ...
Pericytes recruited by CCL28 promote vascular normalization after anti
Our study showed that after undergoing anti-angiogenesis treatment, the tumor exhibited a state of ischemia and hypoxia, leading to an upregulation in the expression of CCL28 in hypoxic lung adenocarcinoma cells by the hypoxia-sensitive transcription factor CEBPB.
Comparison of specific expression profile in two in vitro hypoxia models
The microenvironment plays a fundamental role in carcinogenesis: Acidity and hypoxia are actively involved in this process. It is important to have in vitro models to study these mechanisms. The models that are most commonly referred to are the hypoxia ...
Genome assembly of redclaw crayfish (Cherax quadricarinatus) provides
Three hypoxia-related genes (KDM3A, KDM5A, HMOX2) were identified as being subjected to positive selection in C. quadricarinatus. ... Experimental infection of C. quadricarinatus with Macrobrachium rosenbergii nodavirus showed that C. quadricarinatus has low susceptibility ... (v2-2.1). Gene models were predicted based on the alignment results ...
Experimental Parameterization of a Model of Hypoxia Dynamics in
This paper estimates the parameters of a model of hypoxia in large animals (specifically, Yorkshire swine) from experimental data. The paper adopts a simple, control-oriented model of oxygen transport and consumption dynamics. The model uses three state variables to represent oxygen concentrations and/or partial pressures in the animal's ...
Google Gemini 1.5 Pro leaps ahead in AI race, challenging GPT-4o
Google's Gemini 1.5 Pro revolutionizes AI with top LMSYS scores, outperforming GPT-4 and Claude-3.5, promising transformative impact across industries while raising ethical questions.
OpenAI launches experimental GPT-4o Long Output model with 16X token
It is a variation on its signature GPT-4o model from May, but with a massively extended output size: up to 64,000 tokens of output.
Experimental demonstration of magnetic tunnel junction-based
The experiment-based physics modeling and calculations for obtaining the projected CRAM logic operation accuracies began with an MTJ resistance-voltage model 77, which was fit to the experimental ...
Sugen, hypoxia and the lung circulation
Sugen, hypoxia and the lung circulation. Twenty years ago, the paper was published that first described the Sugen Hypoxia rat model, a novel experimental model of pulmonary vascular disease and severe pulmonary arterial hypertension (PAH). 1 With this short letter, we highlight some advantages and disadvantages inherent to the model and provide ...
Design and experimental study of biomimetic microtextured ...
The experimental results indicated a substantial reduction in cutting force and friction coefficient for tools with microtextures, excluding concentric circle textures. Cutting experiments affirmed the effective enhancement of cutting performance for single-crystal diamond tools.
Experimental Lake Erie Hypoxia Forecast
The Experimental Lake Erie Hypoxia forecast is a forecast model in development for transition to NOAA operations. Other panels provide additional information on lake conditions. For the latest images, refresh your browser and/or clear its cache.

Experimental Hypoxia as a Model for Cardiac Regeneration in Mice

Figures ( 0 ) & Videos ( 0 )

Related articles

Novel approach to study hypoxia enables identification of a marker for ovarian cancers

The LTHY method: To reproduce hypoxic conditions in the laboratory

A new marker of aggressiveness and survival?

Hospital pneumonia diagnoses are uncertain, revised more than half the time, study finds

EMS training on key skills improves heart attack survival

Effective mental health care takes varying forms, says new study

New biomaterial regrows damaged cartilage in joints

A new drug could turn back the clock on multiple sclerosis

New method tracks how psychedelics affect neurons in minutes

Botanicals like turmeric, green tea are harming Americans' livers

In law enforcement, a survey finds a link between head injuries and depression, PTSD

Study finds omega-3 supplements reduce genetic risk of high total cholesterol, LDL and triglyceride levels

Honey added to yogurt supports probiotic cultures for digestive health

Molecular mechanisms regulating prostate cancer hypoxia identified

Team identifies protein key to cancer cells ability to spread

New research sheds light on why tumor cells become resistant to chemotherapy

Oxygen-starved tumor cells have survival advantage that promotes cancer spread

Smart hydrogel with oxygen-scavenging capability inhibits tumor growth and metastasis

Tumor resistance is promoted by anti-cancer protein

Recommended for you

Study finds noncoding RNAs dysregulated in several human cancers

Proposed lung cancer screening guideline could eliminate racial disparities, allow for early detection

Researchers develop promising therapy treatment that can kill glioblastoma cells in newly-discovered brain pathway

Researchers devise novel solution to preventing relapse after CAR T-cell therapy

Researchers highlight the genetic complexity of schizophrenia

E-mail the story

Newsletter sign up

Donate and enjoy an ad-free experience

E-mail newsletter

Save citation to file

Add to My Bibliography

Experimental Hypoxia as a Model for Cardiac Regeneration in Mice

Similar articles

Publication types

Related information

Grants and funding

LinkOut - more resources

Pericytes recruited by CCL28 promote vascular normalization after anti-angiogenesis therapy through RA/RXRA/ANGPT1 pathway in lung adenocarcinoma

Introduction

Materials and methods

Isolation and identification of pericytes

Hypoxic culture model and RNA sequence assay

RNA extraction and quantitative PCR

Luciferase reporter assay

Chromatin Immunoprecipitation (ChIP)

Liquid Chromatography-Mass Spectrometry (LC–MS)

Single-cell RNA sequencing

Animal model

Western blot and ELISA

Immunofluorescence and multiplex Immunohistochemistry (mIHC)

CRISPR-cas9 knock out

Knockdown of RDH13 and DHRS11

Bioinformatic analysis

Statistical analyses

CCL28 expression is up-regulated after anti-angiogenesis therapy by hypoxia-sensitive transcription factor CEBPB in lung adenocarcinoma

Tumor-derived CCL28 promotes vascular normalization and pericyte recruitment in the tumor microenvironment

Tumor-derived CCL28 promotes expression of angiopoietin-1 via CCR3 in pericytes

CCL28 activates retinoic acid signaling in pericytes through CCR3

Conclusions

Availability of data and materials

Abbreviations

Acknowledgements

Author information

Contributions

Corresponding authors

Ethics declarations

Consent for publication

Competing interests

Supplementary Information

About this article

Share this article

Journal of Experimental & Clinical Cancer Research

Genome assembly of redclaw crayfish ( Cherax quadricarinatus ) provides insights into its immune adaptation and hypoxia tolerance

Conclusions

Chromosome-level genome assembly of C. quadricarinatus

Transposable elements and tandem repeats drive the expansion of decapod crustacean genomes

Gene families related to immunity expanded in the genome of C. quadricarinatus