Age-related pathologies are accelerated in mice with mLOY in blood

We constructed a mouse model of hematopoietic mLOY by using CRISPR-Cas9 gene editing to target repeat DNA sequences that are specific to the centromere of the Y chromosome. Guide RNAs (gRNAs) and a tomato red fluorescent protein (tRFP) marker were delivered to lineage-negative bone marrow cells through the lentivirus vector before transplantation into lethally irradiated, wildtype male mice (Fig. 1A). To decouple theCas9 gene from the lentivirus vector and to maximize gene-editing efficiency, donor bone marrow cells were isolated from ROSA26-Cas9 knock-in mice that express the Cas9 endonuclease systemically (11). Two LOY-gRNAs, LOY-gRNA1 and LOY-gRNA2, targeting different repeat sequences within the centromere were evaluated for the efficiency of Y chromosome ablation by fluorescent in situ hybridization (FISH) analysis of X and Y chromosomes in tRFP⁺ blood cells collected from mice reconstituted with bone marrow cells. As a control, we used a lentivirus encoding a gRNA designed to not target any region of the genome in the ROSA26-Cas9 knock-in donor cells. Y chromosome ablation efficiency was ~95 and 80% for LOY-gRNA1 and LOY-gRNA2, respectively (Fig. 1,B andC). Thus, unless otherwise indicated, the mouse model of mLOY used a lentivirus vector that expresses the LOY-gRNA1 transcript and the tRFP⁺ marker protein to assess cell transduction. Y chromosome ablation in vivo was validated by karyotype analysis using lineage-negative bone marrow cells isolated from mLOY and control mice that were immortalized by lentivirus-mediated HoxB8 overexpression (Fig. 1D). Consistent with the ablation of the male sex chromosome, the Y chromosome–encoded transcriptsKdm5d,Uty,Eif2s3y, andDdx3y were not detectable in the circulating tRFP⁺ leukocytes of the mLOY mice (Fig. 1E). Because of the inefficiency of bone marrow progenitor cell transduction by the lentivirus vector, blood chimerism, defined as a percentage of tRFP⁺ white blood cells, ranged from 49 to 81% in various experiments (mean = 64.9 ± 4.0%), which is consistent with the levels of mLOY in men that have been associated with various disease processes (2,9). These levels of chimerism were maintained for 12 months (fig. S1A). Focusing on each immune cell population, chimerism was higher in myeloid cells compared with B and T cells, consistent with observations of mLOY in men (12) (fig. S1B).

The phenotypic consequences of mLOY were evaluated in aging male mice. No obvious hematological abnormalities were observed in mLOY or control mice during the follow-up period (fig. S1C). However, mice with the mLOY condition displayed shorter life spans compared with control mice (Fig. 1F). Serial echocardiographic analyses revealed the development of an accelerated age-associated cardiomyopathy in the mLOY mice, with greater cardiac dysfunction detected in older mice (Fig. 1G andfig. S2A). Consistent with the development of an age-associated cardiomyopathy, mLOY mice displayed a small increase in heart mass at the termination of the experiment, but there were no differences in body weights between the mLOY and control genotypes over the course of the experiment (fig. S2B). The mLOY condition led to an increase in myocardial fibrotic area (Fig. 1H), as determined by quantitative analysis with picrosirius red staining of cardiac sections (fig. S2C), and an increase in the quantity of myocardial MEF-SK4⁺ fibroblasts, as assessed by flow cytometry (Fig. 1I). Consistent with this fibrotic response, the mLOY condition also led to increased left ventricular filling pressure, which is indicative of diastolic dysfunction (fig. S2D). These cardiac changes were observed despite modest reductions in blood pressure in the mLOY mice (fig. S2E) and no changes in the serum concentrations of renin 1 and angiotensin II (fig. S2F). Although further analyses of how mLOY affects the renin-angiotensin system are warranted, these experimental results are consistent with observations of diminished self-reported hypertension in cardiovascular disease patients with mLOY (10). At the 15-month time point after bone marrow transplantation (BMT), mLOY mice also showed accelerated fibrotic response in the lung interstitium (fig. S3A), and greater pulmonary fibrosis could also be observed in young mice after the intratracheal administration of bleomycin (fig. S3B). Histological analysis of the kidneys also revealed a greater degree of fibrosis in the mLOY mice compared with control mice at 15 months after BMT (fig. S3C). Finally, assessments of cognitive function revealed that aging mLOY mice (15 months after BMT) had short-term working memory deficits in the Y-maze and novel object recognition tests, but these phenotypes were not observed in young mLOY mice (2 months after BMT) (fig. S4). Collectively, these results indicate that mLOY mice can recapitulate aspects of the mLOY phenotype observed in men and suggest that accelerated tissue fibrosis could be a mechanistic feature of this condition. Because a potential relationship between mLOY and cardiac dysfunction has not been reported previously, further investigations focused on the role of mLOY in heart failure are warranted.

mLOY is associated with death from cardiovascular disease and heart failure in men

mLOY has been associated with preexisting cardiovascular disease (8,10). However, a prospective analysis of mLOY with mortality from cardiovascular disease with substantial follow-up time is lacking. Thus, we investigated male survival data extracted from the UK Biobank in November 2020 (median follow-up time, 11.5 years) and conducted three different types of analysis to evaluate whether mLOY in leukocytes was associated with risk for death caused by cardiovascular diseases. The percentage of blood cells lacking chromosome Y in each male participant at baseline was estimated using intensity data from single-nucleotide polymorphism array experiments (fig. S5 andtable S1). In the first analysis, a continuous estimate of mLOY was used as an explanatory variable in a multivariable-adjusted Cox proportional hazards regression for mLOY and death caused by cardiovascular disease, adjusted for the confounding variables outlined intable S2. The dependent variable in this primary model was mortality from any cardiovascular disease during follow-up, as indicated by a registered primary or secondary cause of death from the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD10) codes I00 to I99. This survival analysis established that mLOY in leukocytes is associated with death from overall cardiovascular disease in men [hazard ratio (HR) = 1.0054 per 1% increase in LOY, 95% confidence interval (CI) = 1.0022 to 1.0087, adjustedP = 0.0010, events = 6222). Using a penalized splines approach, the association could be illustrated over the lull spectrum of LOY mosaicism (Fig. 2A), and a dose-response relationship was observed, indicating that mLOY might have a direct physiological effect. Exploratory Cox models using a binary LOY variable (defined from the spline analysis as described infig. S6A) showed that men with LOY in >40% of leukocytes at study entry displayed 31% increased risk of dying from any disease of the circulatory system during follow-up (Fig. 2B, HR = 1.31, 95% CI = 1.02 to 1.70, adjustedP = 0.0382, events = 6222). To identify the specific cardiovascular outcomes that might be driving the overall association, we next investigated associations between mLOY and each of the ICD10 codes within I00 to I99 and with at least 200 observed events using analogous Cox models (table S3). The specific causes of death associated with mLOY included hypertensive heart disease (I11; HR = 3.48), heart failure (I50; HR = 1.76), congestive heart failure (I50.0; HR = 2.42), and aortic aneurysm and dissection (I71; HR = 2.76) (Fig. 2B). Because comorbidities are common among cardiovascular conditions, other registered cardiovascular-related causes of death showing association with mLOY (but normally not considered as primary causes of death) are summarized infig. S6B. Next, cardiovascular diseases pooled on the basis of medical features and clinical resemblance were investigated using analogous survival models (table S4). In these extended analyses, there was no significant association between mLOY and death from diseases encompassing multiple etiologies of heart failure (HR = 1.65, 95% CI = 0.95 to 2.86, adjustedP = 0.0745, events = 1102). However, mLOY was associated with death caused by associated aortic and peripheral arterial diseases (HR = 2.79, 95% CI = 1.30 to 5.97, adjustedP = 0.0082, events = 279). Finally, corroborating prior analyses from the UK Biobank that used a shorter median (7-year) follow-up period (8), it was found that men with mLOY in >40% of leukocytes at study entry displayed a 41% increased risk of dying from any cause during the follow-up (Fig. 2B).

mLOY worsens the outcome of experimental heart failure in mice

To further investigate the relationship between mLOY and cardiac dysfunction, 12- to 16-week-old (4 weeks after BMT) male mice were subjected to transverse aortic constriction (TAC) surgery (Fig. 3A). Although TAC in mice does not completely replicate nonischemic heart failure, this widely used model produces cardiac dysfunction, left ventricle hypertrophy, and fibrosis phenotypes that are prevalent in the elderly (13). In agreement with observations in the older unchallenged mice, echocardiographic analysis revealed a greater progressive decline in cardiac function in mLOY mice compared with control mice alter TAC (Fig. 3B andfig. S7A). The mLOY mice also displayed a marginally greater increase in the ratio of heart weight to tibia length, as well as an increase in the ratio of lung weight to tibia length that is indicative of lung congestion resulting from cardiac dysfunction (Fig. 3C). Transcripts encoding atrial natriuretic peptide A and the ratio of transcripts encoding myosin heavy chains β/α, a marker of advanced heart failure, were up-regulated in the hearts from mLOY mice compared with those from control mice (Fig. 3D). Histological analysis revealed greater interstitial and perivascular fibrosis in both the left ventricle and atrium of mLOY mice after TAC (Fig. 3E andfig. S7,B andC). Consistent with the increase in fibrosis, flow cytometric analysis of heart cells revealed that the number of MEF-SK4⁺ fibroblasts was higher in the hearts from TAC-treated mLOY mice compared with control (Fig. 3F). By contrast, there was no detectable difference in the number of cardiac endothelial cells or the average myocyte cross-sectional area between the experimental groups that had undergone TAC (Fig. 3,G andH), indicating that the predominant cellular effect of mLOY is on the fibroblast content of the heart.

Accelerated cardiac dysfunction could also be demonstrated in mice transplanted with lineage-negative bone marrow cells that were transduced with a gRNA that targets a different centromeric repeat in the Y chromosome (mLOY-gRNA2). The efficiency of Y chromosome ablation achieved by mLOY-gRNA2 was comparable but slightly less than that of mLOY-gRNA1 (Fig. 1B andfig. S8A). Transduction with mLOY-gRNA2 did not affect white blood cell counts, hemoglobin, or platelet counts (fig. S8B). However, TAC led to greater cardiac dysfunction, increased heart and lung weights, higher concentrations of serum brain natriuretic peptide (BNP), and higher numbers of cardiac fibroblasts in the mLOY-gRNA2–treated mice compared with control mice (fig. S8,C toF). These data corroborate the results with the mLOY-gRNA1 reagent and provide additional support for the finding that Y chromosome deletion in blood cells contributes to cardiac dysfunction and an accelerated flbrotic response.

Y chromosome deficiency modulates the transcriptional profile of cardiac macrophages

Myeloid cells typically display the greatest extents of Y chromosome deficiency in the blood of men (12). In the experimental model, analysis of the cardiac immune cell populations revealed higher numbers of CCR2⁺ cardiac macrophages in TAC-treated hearts from mLOY-gRNA1 mice compared with those from control mice (fig. S9A), suggesting that cardiac macrophages derived from Y chromosome–deficient hematopoietic stem cells have altered functional properties. To test whether Y chromosome deficiency in myeloid cells accounts for the accelerated heart failure phenotype, anti–granulocyte receptor-1 (anti–GR-1) antibody, which blocks neutrophil and monocyte recruitment to injured tissue (14), was administered to the different experimental groups of mice. In mLOY mice that underwent TAC, treatment with anti–GR-1 antibody attenuated the accelerated cardiac dysfunction (fig. S9B) and reversed the elevations in heart weight and serum BNP (fig. S9,C andD). Treatment with anti–GR-1 antibody also reversed the increase of the number of cardiac fibroblasts observed in the mLOY mice, but it did not affect the quantity of vascular endothelial cells (fig. S9E). Collectively, these results suggest that loss of the Y chromosome in myeloid cells can largely account for the pathological cardiac phenotype in the mLOY mouse model.

To address the mechanistic aspects of mLOY-mediated cardiac dysfunction, single-cell RNA sequencing was performed on BMT-derived (RFP⁺) immune cells (CD45⁺) from mLOY and control mice at 7 days after TAC. Seurat analysis of gene expression enabled the clustering of immune cell populations, and the reduced expression of Y chromosome–encoded genes could be detected in various clusters (fig. S10A). CD68⁺ macrophages accounted for the largest portion of immune cells, which displayed polarization based on the expression of the marker genesIl1b andCcr2^hi, which will henceforth define an inflammatory subpopulation, and the marker genesLyve1 andMrc1, which will henceforth define a fibrotic subpopulation (Fig. 4A andfig. S10B). PHATE (potential of heat diffusion for affinity-based trajectory embedding) analysis to assess differentiation progression and branching revealed a trajectory from nonactivated macrophages to a continuum of inflammatory and fibrotic macrophages (fig. S10C), with the fibrotic macrophage subpopulation containing a greater percentage of mLOY cells and the inflammatory macrophage subpopulation containing a greater percentage of control cells (Fig. 4B). Consistent with the interpretation that mLOY promotes macrophage polarization toward a fibrotic phenotype, analysis with the SCENIC (single-cell regulatory network inference and clustering) regulatory gene network algorithm revealed that profibrotic regulons were enriched by the mLOY condition in the fibrotic macrophages (fig. S10D). Conversely, prionflammatory regulons were suppressed by the mLOY condition in the inflammatory macrophage subpopulation. Further analyses revealed that the mLOY condition promoted the enrichment of regulons specifically associated with transforming growth factor–β1 (TGFβ1) signaling in the fibrotic macrophage subpopulation, whereas regulons associated with interleukin-1β (IL-1β) signaling were down-regulated by mLOY in the inflammatory subpopulation (Fig. 4C). The mLOY condition promoted the expression of the transcript encoding TGFβ1 in the fibrotic macrophages, but not in the inflammatory macrophages, whereas the transcript encoding galectin-3, another secreted profibrotic protein, was up-regulated by the mLOY condition in both macrophage populations (fig. S10E).

The phenotypic transition of Y chromosome–deficient cardiac macrophages was also evident in a bulk transcriptome analysis of BMT-derived (RFP⁺) cardiac macrophages. As shown in the principal components analysis plots, mLOY macrophages isolated from heart displayed a transcriptomic profile distinct from that of control cardiac macrophages (fig. S11A). Gene set enrichment analysis (GSEA) revealed that differentially expressed transcripts related to “TGFβ signaling” were enriched in mLOY macrophages (fig. S11B). In addition, GSEA revealed that differentially expressed transcripts related to “TGFβ binding,” including LTBP1, LTBP3, LTBP4, and thrombospondin, which facilitate the localization, secretion, and activation of TGFβ1, as well as other profibrotic factors, were up-regulated in macrophages in the mLOY condition (fig. S11C). In marked contrast, bulk or single-cell analysis revealed little or no differences in transcriptomes of cardiac monocytes, cardiac neutrophils, or blood neutrophils between the mLOY and control conditions (fig. S11,D toG).

Consistent with the transcriptome analyses, elevated TGFβ1 protein and SMAD2 phosphorylation in the myocardium could be detected by immunoblot analysis in the mLOY condition compared with control at 1 week after TAC (fig. S12,A andB). Elevated TGFβ1 protein in the mLOY condition could also be detected by fluorescence immunohistochemistry (fig. S12C). Although TGFβ1/macrophage colocalization was found to a greater extent in the mLOY condition, most of the TGFβ1 signal was localized to other cell types or matrix in the myocardium, consistent with its putative role in a feedforward signaling loop and its affinity for extracellular matrix. Furthermore, the total number of phosphorylated SMAD2⁺ cells was higher in mLOY mice compared with control mice at 1 week after TAC, as confirmed by quantitative analysis of phosphorylated SMAD2⁺ fibroblasts identified by co-immunostaining with vimentin (Fig. 4D).

TGF β1 neutralization reverses the cardiac dysfunction observed in mLOY mice

To assess the role of profibrotic signaling in accelerated cardiac dysfunction caused by the experimental mLOY condition, mice that had undergone TAC were treated with control immunoglobulin G or a monoclonal antibody directed toward TGFβ1. Sequential echocardiographic analysis revealed that treatment with anti-TGFβ1 monoclonal antibody partly reversed the accelerated cardiac dysfunction observed in mLOY mice (Fig. 4E andfig. S12D). The anti-TGFβ1 antibody also reversed the mLOY-mediated increases in serum BNP concentration, heart weight, and lung congestion (fig. S12,E andF). Consistent with the recognized functions of TGFβ1 in promoting fibroblast proliferation and their conversion to myofibroblasts that produce higher amounts of matricellular proteins, the neutralizing anti-TGFβ1 antibody suppressed the increase in fibroblast number and extracellular matrix deposition that was associated with the mLOY condition but did not alter endothelial cell number (Fig. 4,F andG, andfig. S12G).

Funding:

This work was supported by the National Institutes of Health (NIH grants AG073249, AG072095, HL141256, HL139819, and HL142650 to K.W.; NIH grant HL152174 to S.S. and K.W.; Grant-in-Aid for Research Activity Start-up 21K20879 to S.S.; Grant-in-Aid for Scientific Research C 22K08162 to S.S.; and NIH Grant HL146056 to K.K.H.); the MSD Life Science Foundation (S.S.); the Ichiro Kanehara Foundation (S.S.); the Kenzo Suzuki Memorial Foundation (S.S); YOKOYAMA Foundation for Clinical Pharmacology (S.S.); the Cardiovascular Research Fund (S.S.); The Japanese Heart Failure Society (S.S.); the Osaka Medical Research Foundation for Intractable Diseases (S.S.); The European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Program (grants 679744 and 101001789 to L.A.F.); the Swedish Research Council (grant 2017-03762 to L.A.F.); The Swedish Cancer Society (grant 20-1004 to L.A.F.); Kjell och Märta Beijers Stiftelse (L.A.F.); Marcus Borgströms stiftelse (J.H.); and the American Heart Association (grant 20POST35210098 to M.A.E.).

• 1.Jobling MA, Tyler-Smith C, Nat. Rev. Genet18, 485–497 (2017). [DOI] [PubMed] [Google Scholar]
• 2.Forsberg LAet al. , Nat. Genet46, 624–628 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 3.Dumanski JPet al. , Science347, 81–83 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 4.Thompson DJet al. , Nature575, 652–657 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Ljungström Vet al. , Leukemia36, 889–891 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 6.Zink Fet al. , Blood130, 742–752 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7.Forsberg LAet al. , Nat. Genet51, 4–7 (2019). [DOI] [PubMed] [Google Scholar]
• 8.Loftfield Eet al. , Sci. Rep8, 12316 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Dumanski JPet al. , Am. J. Hum. Genet98, 1208–1219 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 10.Haitjema Set al. , Circ. Cardiovasc. Genet10, e001544 (2017). [DOI] [PubMed] [Google Scholar]
• 11.Sano Set al. , J. Vis. Exp (152): (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Dumanski JPet al. , Cell. Mol. Life Sci78, 4019–4033 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 13.Rockey DC, Bell PD, Hill JA, N. Engl. J. Med372, 1138–1149 (2015). [DOI] [PubMed] [Google Scholar]
• 14.Daley JM, Thomay AA, Connolly MD, Reichner JS, Albina JE, J. Leukoc. Biol83, 64–70 (2008). [DOI] [PubMed] [Google Scholar]
• 15.Rhee AJ, Lavine KJ, Annu. Rev. Physiol82, 1–20 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 16.Wang Yet al. , JCI Insight5, e135204 (2020). [Google Scholar]
• 17.Dorsheimer Let al. , JAMA Cardiol. 4, 25–33 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Yu Bet al. , J. Am. Coll. Cardiol78, 42–52 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Fuster JJet al. , Science355, 842–847 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 20.Sano Set al. , Circ. Res123, 335–341 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 21.Sano Set al. , J. Am. Coll. Cardiol71, 875–886 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 22.Yura Yet al. , Circ. Res129, 684–698 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23.Wijsenbeek M, Cottin V, N. Engl. J. Med383, 958–968 (2020). [DOI] [PubMed] [Google Scholar]
• 24.Kang TE Jr. et al. , N. Engl. J. Med370, 2083–2092 (2014). [DOI] [PubMed] [Google Scholar]
• 25.Lewis GAet al. , Nat. Med27, 1477–1482 (2021). [DOI] [PubMed] [Google Scholar]
• 26.Fujiwara Aet al. , Sci. Rep10, 10900 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
• 27.Aghajanian Het al. , Nature573, 430–433 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

Supplementary Materials

Data Availability Statement

All data are available in the main text or thesupplementary materials. Data were deposited into the National Center for Biotechnology Information’s Gene Expression Omnibus database under accession numbersGSE205091 andGSE204732.