DNMT1 Regulates Testosterone Synthesis in Aged Rat Leydig Cells via FTO‐Mediated HSP90AA1 m6A Levels

Testosterone is far more than the hormone of male virility. It sustains sperm production, modulates immune function, preserves bone density, supports cardiovascular health, and even influences sleep quality. When levels fall, as they inevitably do with advancing age, the consequences ripple across nearly every system in the male body. Now, a team of researchers writing in the journal Aging Cell has uncovered a previously hidden molecular cascade that connects two of biology’s most talked-about epigenetic layers—DNA methylation and RNA methylation—to the slow decline of testosterone synthesis in aging testes. The work, conducted in rats, reveals that a single enzyme called DNMT1 acts as a master switch governing how much testosterone the testes can produce, and it does so through an unexpected route involving the RNA demethylase FTO and a molecular chaperone known as HSP90AA1.

The story begins with a simple observation. When the researchers profiled rat testicular tissue at 2, 6, 12, and 18 months of age, they found that both overall DNA methyltransferase activity and circulating testosterone levels followed strikingly parallel trajectories, with a pronounced concurrent decline beginning at 12 months. The enzymes that actually build testosterone—CYP11A1, CYP17A1, and the cholesterol transporter StAR—mirrored this drop precisely. Intriguingly, when the team mined existing human multi-tissue transcriptomic datasets, they discovered that DNMT1, the principal maintenance methyltransferase responsible for preserving genomic methylation patterns during cell division, is preferentially and abundantly expressed in the testis compared with 31 other human tissues. This positioned DNMT1 as a prime candidate regulator of testicular function, and indeed, complementary RT-qPCR, Western blotting, and immunohistochemical staining all confirmed a progressive age-related decline in DNMT1 expression in rat testes.

To isolate the cell-intrinsic relationship between DNMT1 and testosterone production without the confounding influences of the living organism, the researchers established an in vitro aging model using primary Leydig cells—the specialized interstitial cells that serve as the testis’s sole source of testosterone. Isolated from juvenile rats and serially passaged to induce replicative senescence, these cells developed a robust aged phenotype, with more than 70 percent staining positive for senescence-associated β-galactosidase. They also upregulated the senescence marker p16 and secreted elevated levels of inflammatory SASP factors including TNF-α, IL-1β, and IL-6. Crucially, these senescent Leydig cells produced markedly less testosterone and expressed significantly less DNMT1, recapitulating the age-related declines observed in living animals and supporting DNMT1 as a cell-intrinsic correlate of Leydig cell functional aging.

Gain- and loss-of-function experiments then established DNMT1 as a genuine positive regulator of steroidogenesis rather than a mere bystander. When the researchers knocked down DNMT1 using lentiviral shRNA, testosterone secretion into the culture medium plummeted, and the protein expression of CYP11A1, CYP17A1, and StAR all fell in concert. Conversely, overexpressing DNMT1 produced the reciprocal phenotype, elevating testosterone output and upregulating the same rate-limiting enzymes. DNMT1 knockdown also drove Leydig cells toward senescence, increasing the proportion of β-galactosidase-positive cells and upregulating p16 and SASP factors. Overexpression of DNMT1 in already-senescent cells reversed this program, reducing senescence markers and restoring testosterone secretion. Together, these findings position DNMT1 as an epigenetic safeguard against the functional decay of the aging testis.

The real surprise came when the team performed RNA sequencing on DNMT1-deficient Leydig cells to identify what DNMT1 actually controls. Among the altered transcripts, the mRNA encoding FTO—the fat mass and obesity-associated protein, better known as a demethylase that erases N6-methyladenosine (m6A) marks from messenger RNA—was significantly upregulated when DNMT1 was silenced. Methylation-specific PCR revealed that the Fto promoter became significantly hypomethylated following DNMT1 knockdown, and chromatin immunoprecipitation coupled with qPCR confirmed that DNMT1 physically binds to the CpG island within the Fto promoter, enriching approximately 398-fold over a control antibody. In other words, DNMT1 normally keeps FTO switched off by methylating its promoter, and when DNMT1 levels fall with age, FTO is derepressed and begins to accumulate.

FTO, it turns out, is not a benign consequence of DNMT1 loss but an active agent of decline. When the researchers overexpressed FTO in Leydig cells, testosterone production dropped and the steroidogenic enzymes CYP11A1, CYP17A1, and StAR were all downregulated—phenocopying the effect of DNMT1 knockdown. Dual knockdown experiments then established the direction of causality: silencing FTO in DNMT1-deficient cells rescued testosterone production and restored enzyme expression, whereas silencing DNMT1 in FTO-deficient cells produced the opposite outcome. Notably, FTO overexpression did not feed back to alter DNMT1 levels, confirming that the regulatory axis is unidirectional, with DNMT1 upstream and FTO downstream.

To find FTO’s actual target, the team turned to computational prediction and RNA immunoprecipitation. Using the catRAPID protein-RNA interaction database, they identified Hsp90aa1—the gene encoding heat shock protein 90α, a molecular chaperone that stabilizes and activates the pro-survival kinase AKT—as the transcript with the highest binding propensity for FTO. RIP-qPCR using an FTO-specific antibody confirmed that Hsp90aa1 mRNA is a direct binding substrate of FTO. MeRIP-qPCR then showed that FTO overexpression reduced the m6A modification level of Hsp90aa1 transcripts, and this demethylation was accompanied by downregulation of Hsp90aa1 mRNA and HSP90AA1 protein. Conversely, FTO knockdown elevated both m6A marks and HSP90AA1 abundance. Importantly, DNMT1 knockdown did not alter the DNA methylation status of the Hsp90aa1 promoter, yet it still reduced HSP90AA1 expression—and this reduction was rescued by concurrent FTO knockdown. The message is clear: DNMT1 regulates HSP90AA1 indirectly, through the FTO-mediated m6A axis, rather than through direct promoter methylation.

The downstream consequences of losing HSP90AA1 are what ultimately kill testosterone production. HSP90AA1 overexpression increased testosterone output and upregulated the steroidogenic enzymes, while simultaneously raising the levels of phosphorylated AKT and the anti-apoptotic protein BCL-2 and lowering the pro-apoptotic protein BAX. Pharmacological inhibition of HSP90AA1 with 17-AAG produced the opposite effects, reducing testosterone secretion, suppressing AKT phosphorylation, and tipping the balance toward cell death. Rescue experiments confirmed the hierarchical architecture of the pathway: FTO knockdown reversed the inhibitory effects of DNMT1 deficiency on HSP90AA1 expression, AKT phosphorylation, and testosterone production, whereas HSP90AA1 overexpression rescued the functional impairments induced by FTO overexpression. TUNEL staining corroborated these molecular findings, confirming that the DNMT1–FTO–HSP90AA1–AKT axis exerts hierarchical control over Leydig cell apoptosis.

When the researchers examined how this axis behaves during natural aging, the picture became even more coherent. In testicular tissue from 18-month-old rats, the Fto promoter was significantly hypomethylated compared with 2-month-old controls, FTO binding to Hsp90aa1 mRNA was markedly enhanced, and the m6A modification level of Hsp90aa1 transcripts was significantly reduced. FTO expression rose progressively with age while HSP90AA1 fell, and immunohistochemical staining revealed elevated FTO, reduced HSP90AA1, and increased pro-apoptotic Caspase-3 in aged testicular sections. TUNEL staining verified a significant age-dependent increase in apoptotic cells, and Western blot analysis showed that phosphorylated AKT and BCL-2 were significantly decreased while BAX was markedly upregulated. Senescent Leydig cells in vitro recapitulated every one of these molecular alterations, and primary Leydig cells isolated from aged rats displayed the same reduced Fto promoter methylation, reduced m6A modification of Hsp90aa1, and impaired AKT signaling.

To validate the pathway in a living organism, the team delivered adeno-associated virus serotype 9 carrying a Dnmt1-targeting construct directly into rat testes via local intratesticular injection. Three weeks later, serum testosterone levels were significantly reduced, the steroidogenic enzymes CYP11A1, CYP17A1, and StAR were downregulated, Fto promoter methylation was decreased, FTO was upregulated, HSP90AA1 was downregulated, and apoptotic cells within the testicular interstitium were markedly increased. Phosphorylated AKT and BCL-2 fell while BAX rose, and senescence markers including p16 and SASP factors were elevated. These in vivo observations mirror the in vitro findings almost exactly, confirming that the DNMT1–FTO–HSP90AA1–AKT cascade operates as a genuine regulatory circuit in the aging testis rather than an artifact of cell culture.

The implications of this work extend well beyond basic reproductive biology. Testosterone replacement therapy, currently the most widely applied intervention for age-related hypogonadism, carries notable contraindications and adverse effects including polycythemia, cardiovascular events, and prostate-related morbidity, and its efficacy remains suboptimal. The DNMT1–FTO–HSP90AA1 axis offers a mechanistic alternative: rather than supplying testosterone from outside the body, it suggests that pharmacological strategies aimed at preserving DNMT1 activity, restraining FTO expression, or sustaining HSP90AA1-dependent AKT signaling could enhance endogenous testosterone synthesis from within. Selective small-molecule inhibitors targeting DNMT1 and m6A modifiers provide feasible therapeutic entry points, and the identification of Leydig cells as among the earliest testicular cell types to undergo senescence makes them a particularly attractive target for interventions designed to postpone reproductive aging. The authors acknowledge several limitations, including the fact that the precise m6A residues on Hsp90aa1 transcripts have not yet been mapped, that the MSP assay lacked certain positive controls, and that the findings have not yet been validated in human testicular samples. Nevertheless, the study establishes a comprehensive molecular model in which the slow erosion of a single epigenetic enzyme—DNMT1—triggers a cascade of RNA-level changes that ultimately silence the testis’s capacity to produce its own testosterone, and it identifies the DNMT1–FTO–HSP90AA1–AKT axis as a promising therapeutic target for enhancing endogenous steroidogenesis during aging.

Subject of Research: Epigenetics and Reproductive Aging

Article Title: DNMT1 Regulates Testosterone Synthesis in Aged Rat Leydig Cells via FTO‐Mediated HSP90AA1 m6A Levels

Article References: Li, Y., Duan, Y., Tian, Z., Gao, Y., Qiao, Y., Zhao, S., Li, W., Gao, Z., & Niu, S. (2026). DNMT1 Regulates Testosterone Synthesis in Aged Rat Leydig Cells via FTO ‐Mediated HSP90AA1 m 6 A Levels. Aging Cell, 25(10), Article e70760. https://doi.org/10.1111/acel.70760

Image Credits: AI Generated

DOI: 10.1111/acel.70760

Keywords: DNMT1, FTO, HSP90AA1, m6A, testosterone, Leydig cells, epigenetics, aging, AKT signaling, reproductive biology, Regulates, Synthesis