Vitamin D3 for Life

Live Enlightened

Physiology

Vitamin D, Chromatin Accessibility and Epigenetic Marks: Full Evidence Review

Previous related pageVitamin D, Chromatin Accessibility and Epigenetic Marks: Study SummaryNext related pageTissue-Specific Vitamin D Metabolism and Gene Responses: Full Study
Study at a glance

What question did the studies ask?

Researchers asked whether vitamin D affects how DNA is packaged and regulated. DNA carries genetic information. Chromatin is DNA together with its associated proteins. Studies measured how accessible DNA was, chemical marks on DNA or proteins, and where proteins that regulate genes were attached to DNA. These are different measurements.

What did the researchers find?

Cell experiments found changes that depended on cell type and timing. Many genes could change activity at DNA regions that were already accessible. In one cell-line study, some accessibility changes became smaller later. That does not establish permanence. A selected one-person blood study found responses over one or two days. Repeated samples were not independent participants.

A pregnancy trial found small differences at selected DNA sites in newborn cord tissue. Overall differences did not meet its standard for a clear result after accounting for the number of sites tested. A larger observational newborn study found no clear associations after a similar statistical adjustment. Neither established lasting changes in children’s gene activity or health.

What does this mean?

Other experiments investigated proteins that package DNA and developmental effects in mice. Cell experiments often used calcitriol, the active vitamin D hormone. Some used a laboratory-made substance that activates the vitamin D receptor, a protein that responds to vitamin D. These differ from oral vitamin D3. The studies do not establish lasting changes that improve people’s health.

Read the full article below ↓

Vitamin D research measures several distinct features of gene regulation: chemical marks on DNA, modifications of DNA-associated proteins, how accessible DNA is, and where regulatory proteins bind. The evidence shows context-specific changes in selected experimental systems, alongside important human null findings. It does not establish that all vitamin D-responsive genes require newly opened chromatin or that the changes are permanent. For the shorter account, read Vitamin D, Chromatin Accessibility and Epigenetic Marks: Study Summary.

Four measurements that should not be merged

Chromatin consists of DNA and its associated proteins. Accessibility assays identify regions more readily reached by the experimental probe; they do not directly show that a gene is expressed. DNA methylation measures chemical groups attached to selected DNA bases. Histone assays examine modifications of proteins around which DNA is organised. Protein-occupancy assays ask where a regulator is bound. These measurements can be related, but one is not a substitute for the others. The general subject is introduced in Vitamin D and Epigenetic Regulation.

This focused evidence review compares primary studies across those measurements. It is not a pooled estimate of a universal epigenetic effect. A changed mark need not change expression, and nearby occupancy does not identify a functional gene target by itself. Messenger RNA, the intermediate copied from an active gene, supplies another measurement rather than completing every causal step. That distinction connects the experiments with Vitamin D and Gene Expression.

Most laboratory studies used calcitriol, the active vitamin D hormone, or a synthetic receptor ligand. Human supplementation studies generally used cholecalciferol, or vitamin D3; an observational blood-status study is a different design again. These interventions are not interchangeable with deleting the vitamin D receptor, abbreviated VDR. Vitamin D Receptors explains the receptor, while Vitamin D and Nuclear Receptors places it within the broader nuclear-receptor family. Research exposures describe the experiments, not personal dosing instructions.

Time-resolved accessibility in a monocyte cell line

Seuter and colleagues exposed THP-1 cells, a human leukaemia-derived monocyte model, to 100 nanomoles per litre calcitriol. Accessibility was measured at baseline and after two, 24 and 48 hours using an assay that enriches relatively exposed DNA. There were three biological replicates. The study identified 8,979 responsive regions among 62,231 consistently accessible regions, using nominal paired significance testing rather than a false-discovery-adjusted threshold for this responsive-region definition. Thousands of regions are therefore not thousands of independently replicated biological effects. [1]

The strongest accessibility changes were seen around 24 hours, with reclosure by 48 hours at many sites. This limits a permanently open-chromatin interpretation. Continued exposure is also different from a withdrawal experiment: the time course does not show that a new state survives after calcitriol is removed. Selected tests implicated CTCF, a protein involved in genome organisation, but did not establish that every responsive region or gene depends on it. Some three-dimensional contact information came from another cell line and was not a new direct map in these THP-1 cells. [1]

Acute human blood responses: repeated sampling is not broad replication

The VitDHiD study investigated one selected high-responder man after three monthly oral boluses of 80,000 international units of vitamin D3. Blood mononuclear-cell samples were analysed before treatment and 24 or 48 hours afterwards. Repeated challenges within one person are informative about repeatability in that person, not three independent participants. The study combined accessibility and RNA measurements in a mixture of blood-cell types. [2]

Its result table classified 3,538 of 33,836 regions as having increased accessibility at both post-treatment times under a false-discovery threshold of 0.10. Wording elsewhere about at least one time differs, so the table-based definition matters. A comparison cohort of 13 people received a single weight-based bolus. Its nominal accessibility overlap was limited rather than a broad replication of the selected individual. Cell-mixture changes, responder selection and lack of placebo also constrain attribution. [2]

Nearby RNA responses make a regulatory interpretation plausible, but a peak within a broad genomic window is not proof of physical contact or a causal enhancer-to-gene relationship. Some binding information was borrowed from THP-1 experiments. Participant or programme overlap with other acute blood studies also prevents treating every report as independent confirmation. These acute measurements do not establish persistent immune reprogramming or improved clinical immunity. [2]

Endometrial cells: expression at already accessible regions

Yi and colleagues studied an immortalised human endometrial stromal cell model, T-HESC, treated with two nanomoles per litre calcitriol. The accessibility experiment used three technical replicates per condition, not three independent donors. Published gene-expression and protein-binding datasets were overlaid using genomic proximity; their replicate structures should not be assumed to match the new accessibility experiment. Both more-accessible and less-accessible regions were reported, but many expression-associated sites were already accessible before treatment. This directly challenges equating every expression response with newly opened DNA. [3]

The mapping used proximity to genes rather than established regulatory contacts. Treatment-duration and accessibility-threshold reporting were insufficiently clear for precise protocol reconstruction, and some total-region counts differ between text and figure. These limits restrict fine numerical and causal claims. Actual decidualisation, a differentiated uterine state, and improved fertility were not established. The distinction between regulatory measurements and demonstrated cell fate is developed in Vitamin D and Cell Differentiation. [3]

Maternal supplementation and newborn DNA methylation

The MAVIDOS analysis compared 1,000 international units of vitamin D3 daily during pregnancy with placebo. It measured RXRA-region methylation in 453 umbilical-cord tissue samples; a smaller group also had newborn bone measurements. RXRA encodes a receptor partner involved in transcriptional regulation. Pregnant participants with very low vitamin D status were excluded, limiting generalisation to severe deficiency. Randomization strengthens exposure attribution compared with an observational status analysis. [4]

At one CpG site, methylation was an estimated 1.98 percentage points lower, with a 95% confidence interval from 3.65 to 0.32 percentage points lower and nominal P=0.02. A CpG is a DNA sequence context where methylation can be measured. This isolated result must be considered alongside testing of several sites: overall findings did not meet a 0.05 false-discovery threshold after adjustment. Summer-birth results appeared stronger while winter results were null, but these subgroup analyses do not establish a universal seasonal mechanism. [4]

Cord tissue contains several cell types. Its average methylation need not represent the same change in every cell, altered RXRA expression or a cause of bone outcomes. Overall neonatal bone benefit was not established. A separate placental-explant experiment used 25-hydroxyvitamin D, not randomized oral vitamin D3. These components cannot be merged into proof of one complete maternal-to-child mechanism. [4]

The PACE consortium analysed prenatal vitamin D status and cord-blood methylation across seven cohorts and more than 3,200 mother: child pairs. No CpG association met the relevant multiple-testing-adjusted significance criteria. This material human null limits broad claims that lower maternal vitamin D status reliably leaves a detectable newborn blood-methylation signature. It is not a randomized trial or a test of every tissue and every epigenetic mechanism. [5]

Main and tabulated sample totals differ slightly, so a more exact denominator is not securely reconstructed here. The analysis extends earlier work using some of the same cohorts, including ALSPAC and MoBa, rather than adding a new independent population for each publication. PACE and MAVIDOS measured different samples with different designs; the PACE null does not automatically invalidate a targeted cord-tissue finding. Together they limit a generalised newborn DNA-methylation signature. The measurement-to-effect distinction is explored in Vitamin D Status vs Vitamin D Effect. [5] [4]

Histone-related mechanisms need their own evidence

Shen and colleagues treated 786-O kidney-cancer cells with 100 nanomoles per litre calcitriol for 48 hours. JMJD3, a histone-demethylating enzyme, increased alongside reduced global H3K27me3, a methylation mark on histone H3. Selected p16 and beta-galactosidase measurements also increased, and JMJD3 gene silencing attenuated parts of the response. This supplies a histone-related perturbation experiment rather than just an association between vitamin D status and a DNA mark. [6]

It does not establish editing at a particular regulatory locus or prove that the global histone change caused every expression and cell-state result. The global histone-mark blot was normalized to beta-actin rather than total histone H3, limiting how specifically the signal can be attributed to modification alone. Figure-label inconsistencies also constrain fine interpretation, and irreversible arrest was not fully tested. A malignant cell is not a normal ageing cell. The relation to Vitamin D and Cellular Senescence is therefore a specific mechanistic question, not evidence that the same intervention rejuvenates healthy tissues. [6]

Chromatin complexes and context-dependent protein cooperation

Wei and colleagues combined biochemical interaction experiments with rat INS1 insulin-secreting cells under inflammatory cytokine stress and mouse diabetes models. Acetylation, a chemical modification, at lysine 91 of VDR supported recognition by BRD9 and BRD7. Ligand treatment reduced VDR association with BRD9 while increasing association with BRD7; BRD9 inhibition helped favour PBAF over the alternative BAF chromatin-remodelling complex. This is a receptor-protein modification and cofactor exchange, rather than DNA methylation. Accessibility measurements showed that ligand and inhibitor countered selected stress-associated reductions, with a broader response under the combination. H3K27 acetylation also changed, with increases and decreases depending on the genomic locus and experimental comparison. The result was not universal opening or acetylation of every region. [7]

The mouse outcome experiments combined calcipotriol, a synthetic VDR ligand, with a BRD9 inhibitor. Improvements in selected glucose and insulin outcomes under that combination are not evidence for ordinary vitamin D3 efficacy. The experiments did not isolate the proposed physical interface as the sole in-vivo mediator, and other-tissue or off-target effects remain possible. Repeated assays are not independent clinical confirmations. The study supplies a mechanistic chromatin example rather than a demonstrated human diabetes treatment. [7]

Ding and colleagues studied hepatic stellate cells, the liver cells involved in fibrosis. TGF-beta stimulation changed the VDR binding landscape, with VDR and SMAD proteins occupying related genomic regions. At selected loci, ligand-associated VDR recruitment was followed by lower SMAD3 occupancy. Reducing VDR reversed the tested occupancy reduction and transcriptional repression; reducing SMAD also reduced inducible VDR binding. Sequential binding assays and this perturbation/time-course evidence support context-dependent genomic antagonism beyond motif overlap alone. Calcipotriol also modified fibrosis-associated structure in an injured-mouse model, but the data do not identify every occupied region as a causal target or establish human organ benefit. [8]

This is different from Ito and colleagues’ renal-fibrosis work, which included purified VDR and SMAD3 domains and experiments showing reduced SMAD3 DNA binding. That study supplies a direct physical-interaction comparison. Genomic cooccupancy in the liver model, purified-domain interaction in the renal model and a histone modification are distinct evidence types. Protein association alone does not establish a particular epigenetic mark. These distinctions are developed more generally in Vitamin D and Transcription Factors. [8] [9]

Developmental DNA marks: methylation and expression must be separated

Oh and colleagues transplanted whole fetal liver cells from vitamin D-deficient or sufficient mouse dams into sufficient recipients, then examined later immune-cell programmes. Targeted bisulfite/pyrosequencing of 69 CpG sites in groups of four found greater methylation at selected Jarid2 intron-1 regions alongside lower Jarid2 expression. Pathway perturbations investigated how altered regulation might arise, but did not selectively test those DNA marks themselves. JARID2 is an accessory protein associated with the PRC2 regulatory complex, not itself its catalytic methyltransferase. The work supports a locus-associated regulatory mechanism but did not edit the CpG marks selectively to prove that they alone caused the later programme. [10]

Xue and colleagues studied maternal depletion and imprinted loci across mouse generations. Methylation patterns differed by tissue, sex and strain cross; changes could disappear or new ones emerge rather than one fixed mark being transmitted intact. Most differences were small, and no statistically significant overall dietary effect on total expression of the tested imprinted genes was detected in second-generation neonatal liver. A DNA mark is therefore not a direct readout of a persistent gene-expression effect. [11]

These developmental studies answer different questions from acute cell accessibility or newborn human methylation. Persistence requires follow-up after exposure context changes; inheritance requires careful accounting for directly exposed offspring and germ cells. Neither can be inferred simply because an assay is described as epigenetic. The strongest findings remain tied to the particular molecular measurement, organism and perturbation used. [10] [11]

What each assay can establish

MeasurementWhat was observed in selected studiesWhat does not follow automatically
DNA methylationLocus-specific animal findings, small randomized cord-tissue differences and a larger human nullChanged expression, lasting memory or clinical benefit
Histone modificationGlobal H3K27me3 and JMJD3-associated responses in malignant cellsEditing at a specific causal locus or effects in normal ageing tissue
AccessibilityTime: and model-specific exposed DNA regionsEvery expression response needs new opening, or opening is permanent
Protein occupancy or bindingContext-dependent cooccupancy and a separate purified interactionEvery nearby gene is a target or a particular epigenetic modification occurred

Vitamin D-related signalling can influence chromatin and associated regulation, but the evidence supports several specific mechanisms rather than one universal epigenetic programme. Human intervention, observational newborn data, cell models and animal developmental experiments differ in directness. Null findings, pre-existing accessibility and incomplete causal isolation remain part of the scientific answer. The studies do not establish permanent beneficial reprogramming in people.

Scientific sources

[1] Epigenome-wide effects of vitamin D and their impact on the transcriptome of human monocytes involve CTCF. Seuter S, Neme A, Carlberg C. Nucleic acids research. 2016;44(9):4090-4104. DOI 10.1093/nar/gkv1519; PMID 26715761.

[2] Chromatin and transcriptional dynamics underlying the immune-modulatory effects of vitamin D3 in vivo. Rybiński M, Ghosh Dastidar R, Zawrotna N, Gospodarska E, Carlberg C. Scientific reports. 2026;16:2997. Published online 18 December 2025. DOI 10.1038/s41598-025-32831-z; PMID 41413700; PMCID PMC12830676.

[3] Increased chromatin accessibility following 1α,25-dihydroxyvitamin D3 treatment in human endometrial stromal cells. Yi M, Bostan H, DeMayo FJ. Toxicology and applied pharmacology. 2026;516:118036. DOI 10.1016/j.taap.2026.118036; PMID 42767517; PMCID PMC13670113.

[4] Gestational Vitamin D Supplementation Leads to Reduced Perinatal RXRA DNA Methylation: Results From the MAVIDOS Trial. Curtis EM, Krstic N, Cook E, D’Angelo S, Crozier SR, Moon RJ, Murray R, Garratt E, Costello P, Cleal J, Ashley B, Bishop NJ, Kennedy S, Papageorghiou AT, Schoenmakers I, Fraser R, Gandhi SV, Prentice A, Javaid MK, Inskip HM, Godfrey KM, Bell CG, Lillycrop KA, Cooper C, Harvey NC, MAVIDOS Trial Group. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 2019;34(2):231-240. DOI 10.1002/jbmr.3603; PMID 30321476; PMCID PMC6372078.

[5] Epigenome-wide meta-analysis of prenatal vitamin D insufficiency and cord blood DNA methylation. Diemer EW, Tuhkanen J, Sammallahti S, Heinonen K, Neumann A, Robinson SL, Suderman M, Jin J, Page CM, Fore R, Rifas-Shiman SL, Oken E, Perron P, Bouchard L, Hivert MF, Räikköne K, Lahti J, Yeung EH, Guan W, Mumford SL, Magnus MC, Håberg S, Nystad W, Parr CL, London SJ, Felix JF, Tiemeier H. Epigenetics. 2024;19(1):2413815. DOI 10.1080/15592294.2024.2413815; PMID 39418282; PMCID PMC11487971.

[6] Calcitriol induces cell senescence of kidney cancer through JMJD3 mediated histone demethylation. Shen Y, Yu D, Qi P, Wang X, Guo X, Zhang A. Oncotarget. 2017;8(59):100187-100195. DOI 10.18632/oncotarget.22124; PMID 29245970; PMCID PMC5725012.

[7] Vitamin D Switches BAF Complexes to Protect β Cells. Wei Z, Yoshihara E, He N, Hah N, Fan W, Pinto AFM, Huddy T, Wang Y, Ross B, Estepa G, Dai Y, Ding N, Sherman MH, Fang S, Zhao X, Liddle C, Atkins AR, Yu RT, Downes M, Evans RM. Cell. 2018;173(5):1135-1149.e15. DOI 10.1016/j.cell.2018.04.013; PMID 29754817; PMCID PMC5987229.

[8] A vitamin D receptor/SMAD genomic circuit gates hepatic fibrotic response. Ding N, Yu RT, Subramaniam N, Sherman MH, Wilson C, Rao R, Leblanc M, Coulter S, He M, Scott C, Lau SL, Atkins AR, Barish GD, Gunton JE, Liddle C, Downes M, Evans RM. Cell. 2013;153(3):601-613. DOI 10.1016/j.cell.2013.03.028; PMID 23622244; PMCID PMC3673534.

[9] A nonclassical vitamin D receptor pathway suppresses renal fibrosis. Ito I, Waku T, Aoki M, Abe R, Nagai Y, Watanabe T, Nakajima Y, Ohkido I, Yokoyama K, Miyachi H, Shimizu T, Murayama A, Kishimoto H, Nagasawa K, Yanagisawa J. The Journal of clinical investigation. 2013;123(11):4579-4594. DOI 10.1172/jci67804; PMID 24135137; PMCID PMC3809783.

[10] Embryonic vitamin D deficiency programs hematopoietic stem cells to induce type 2 diabetes. Oh J, Riek AE, Bauerle KT, Dusso A, McNerney KP, Barve RA, Darwech I, Sprague JE, Moynihan C, Zhang RM, Kutz G, Wang T, Xing X, Li D, Mrad M, Wigge NM, Castelblanco E, Collin A, Bambouskova M, Head RD, Sands MS, Bernal-Mizrachi C. Nature communications. 2023;14(1):3278. DOI 10.1038/s41467-023-38849-z; PMID 37311757; PMCID PMC10264405.

[11] Maternal vitamin D depletion alters DNA methylation at imprinted loci in multiple generations. Xue J, Schoenrock SA, Valdar W, Tarantino LM, Ideraabdullah FY. Clinical epigenetics. 2016;8:107. DOI 10.1186/s13148-016-0276-4; PMID 27777636; PMCID PMC5062906.

Related reading

Original sources

Source

Chromatin and transcriptional dynamics underlying the immune-modulatory effects of vitamin D3 in vivo

Vitamin D3 is an essential micronutrient that supports innate immunity and modulates adaptive responses, with in vitro studies implicating epigenetic mechanisms.

Publication details

[2] Chromatin and transcriptional dynamics underlying the immune-modulatory effects of vitamin D3 in vivo. Rybiński M, Ghosh Dastidar R, Zawrotna N, Gospodarska E, Carlberg C. Scientific reports. 2026;16:2997. Published online 18 December 2025. DOI 10.1038/s41598-025-32831-z; PMID 41413700; PMCID PMC12830676.

Read more →
Source

Increased chromatin accessibility following 1α,25-dihydroxyvitamin D3 treatment in human endometrial stromal cells.

Published in Yi M, Bostan H, DeMayo FJ. Toxicology and applied pharmacology (2026).

Publication details

[3] Increased chromatin accessibility following 1α,25-dihydroxyvitamin D3 treatment in human endometrial stromal cells. Yi M, Bostan H, DeMayo FJ. Toxicology and applied pharmacology. 2026;516:118036. DOI 10.1016/j.taap.2026.118036; PMID 42767517; PMCID PMC13670113.

Read more →
Source

Gestational Vitamin D Supplementation Leads to Reduced Perinatal RXRA DNA Methylation: Results From the MAVIDOS Trial

In this study, we used an existing randomized trial to test the hypothesis that maternal gestational vitamin D supplementation would lead to reduced perinatal RXRA locus DNA methylation.

Publication details

[4] Gestational Vitamin D Supplementation Leads to Reduced Perinatal RXRA DNA Methylation: Results From the MAVIDOS Trial. Curtis EM, Krstic N, Cook E, D’Angelo S, Crozier SR, Moon RJ, Murray R, Garratt E, Costello P, Cleal J, Ashley B, Bishop NJ, Kennedy S, Papageorghiou AT, Schoenmakers I, Fraser R, Gandhi SV, Prentice A, Javaid MK, Inskip HM, Godfrey KM, Bell CG, Lillycrop KA, Cooper C, Harvey NC, MAVIDOS Trial Group. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 2019;34(2):231-240. DOI 10.1002/jbmr.3603; PMID 30321476; PMCID PMC6372078.

Read more →
Source

Epigenome-wide meta-analysis of prenatal vitamin D insufficiency and cord blood DNA methylation

We aimed to evaluate whether maternal vitamin D insufficiency during pregnancy was conditionally associated with DNA methylation in the offspring cord blood.

Publication details

[5] Epigenome-wide meta-analysis of prenatal vitamin D insufficiency and cord blood DNA methylation. Diemer EW, Tuhkanen J, Sammallahti S, Heinonen K, Neumann A, Robinson SL, Suderman M, Jin J, Page CM, Fore R, Rifas-Shiman SL, Oken E, Perron P, Bouchard L, Hivert MF, Räikköne K, Lahti J, Yeung EH, Guan W, Mumford SL, Magnus MC, Håberg S, Nystad W, Parr CL, London SJ, Felix JF, Tiemeier H. Epigenetics. 2024;19(1):2413815. DOI 10.1080/15592294.2024.2413815; PMID 39418282; PMCID PMC11487971.

Read more →
Source

Vitamin D Switches BAF Complexes to Protect β Cells.

[7] Vitamin D Switches BAF Complexes to Protect β Cells.

Publication details

[7] Vitamin D Switches BAF Complexes to Protect β Cells. Wei Z, Yoshihara E, He N, Hah N, Fan W, Pinto AFM, Huddy T, Wang Y, Ross B, Estepa G, Dai Y, Ding N, Sherman MH, Fang S, Zhao X, Liddle C, Atkins AR, Yu RT, Downes M, Evans RM. Cell. 2018;173(5):1135-1149.e15. DOI 10.1016/j.cell.2018.04.013; PMID 29754817; PMCID PMC5987229.

Read more →
Source

A vitamin D receptor/SMAD genomic circuit gates hepatic fibrotic response.

[8] A vitamin D receptor/SMAD genomic circuit gates hepatic fibrotic response.

Publication details

[8] A vitamin D receptor/SMAD genomic circuit gates hepatic fibrotic response. Ding N, Yu RT, Subramaniam N, Sherman MH, Wilson C, Rao R, Leblanc M, Coulter S, He M, Scott C, Lau SL, Atkins AR, Barish GD, Gunton JE, Liddle C, Downes M, Evans RM. Cell. 2013;153(3):601-613. DOI 10.1016/j.cell.2013.03.028; PMID 23622244; PMCID PMC3673534.

Read more →
Source

A nonclassical vitamin D receptor pathway suppresses renal fibrosis.

[9] A nonclassical vitamin D receptor pathway suppresses renal fibrosis.

Publication details

[9] A nonclassical vitamin D receptor pathway suppresses renal fibrosis. Ito I, Waku T, Aoki M, Abe R, Nagai Y, Watanabe T, Nakajima Y, Ohkido I, Yokoyama K, Miyachi H, Shimizu T, Murayama A, Kishimoto H, Nagasawa K, Yanagisawa J. The Journal of clinical investigation. 2013;123(11):4579-4594. DOI 10.1172/jci67804; PMID 24135137; PMCID PMC3809783.

Read more →
Source

Embryonic vitamin D deficiency programs hematopoietic stem cells to induce type 2 diabetes

Environmental factors may alter the fetal genome to cause metabolic diseases.

Publication details

[10] Embryonic vitamin D deficiency programs hematopoietic stem cells to induce type 2 diabetes. Oh J, Riek AE, Bauerle KT, Dusso A, McNerney KP, Barve RA, Darwech I, Sprague JE, Moynihan C, Zhang RM, Kutz G, Wang T, Xing X, Li D, Mrad M, Wigge NM, Castelblanco E, Collin A, Bambouskova M, Head RD, Sands MS, Bernal-Mizrachi C. Nature communications. 2023;14(1):3278. DOI 10.1038/s41467-023-38849-z; PMID 37311757; PMCID PMC10264405.

Read more →