Abstract

   Spermatogonial stem cells (SSCs) possess the capacity for spontaneous
   reprogramming during in vitro culture, while the underlying mechanisms
   remain poorly understood, especially why the addition of epidermal
   growth factor (EGF), leukemia inhibitory factor (LIF) remarkably
   enhanced transition efficiency. Here we employed a multi-omics
   approach, integrating transcriptomics, metabolomics, and DNA
   methylation analyses to focus on the interplay between exogenous growth
   factors, metabolic pathways, and signaling cascades, particularly the
   role of SMAD3 in these networks. Our findings reveal that SSC
   reprogramming is contingent upon a metabolic shift from the
   tricarboxylic acid (TCA) cycle to aerobic glycolysis, modulated by
   fluctuating SMAD3 levels. SMAD3 downregulation activates HIF-1α,
   inducing aerobic glycolysis to supply energy and substrates for
   reprogramming. Subsequent SMAD3 reactivation promotes rapid cell
   proliferation, facilitating successful reprogramming. This study
   elucidates the pivotal role of SMAD3 in modulating glycometabolic
   pathways driving SSC transformation, emphasizing the necessity of
   aerobic glycolysis following SMAD3 fluctuations for effective
   reprogramming, which provides novel insights into the intricate
   interplay between energy metabolism and stem cell plasticity and
   potential applications in regenerative medicine and fertility
   treatments.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s13287-025-04541-w.

   Keywords: Aerobic glycolysis, Cell fate, Cellular plasticity,
   Germline-to-pluripotent transition, HIF-1A activation, Metabolic
   reprogramming, SMAD3 signaling

Introduction

   Spermatogonial stem cells (SSCs), derived from primordial germ cells
   (PGCs), robustly generate sperm under the regulatory effects of the
   testicular niche [[38]18]. In the last century, it was discovered that
   PGCs could spontaneously transition to a pluripotent state during in
   vitro culture [[39]29, [40]34], highlighting their inherent
   pluripotency. Similarly, SSCs, traditionally viewed as unipotent, have
   been found capable of abrogating lineage commitment and spontaneously
   converting to pluripotent cells after long-term in vitro culture. These
   reprogrammed cells share many features with embryonic stem cells (ESCs)
   derived from the inner cell mass (ICM), including the ability to induce
   teratoma formation and contribute to chimeric animals [[41]13, [42]35].
   This unique reprogramming event distinguishes SSCs from induced
   pluripotent stem (iPS) cells, as it occurs without the introduction of
   exogenous genes or gene products, suggesting a unique potential in
   cellular therapy with minimal risk.

   However, controlling the efficiency of SSC transformation remains
   challenging, as the mechanisms underlying their spontaneous
   reprogramming are largely unknown. Notably, the loss of p53 remarkably
   accelerates SSC transformation [[43]13], potentially linked to
   methylation changes mediated by DNMT1 [[44]39]. Our recent studies have
   analyzed chromatin accessibility changes triggered by p53 deletion,
   revealing an increase in the accessibility of the SMAD3 binding domain
   during the initial phase of cell transformation, alongside a notable
   decrease in SMAD3 protein levels. As the culture period extends, SMAD3
   expression begins to rise, and the cell proliferation rate starts to
   accelerate, culminating in pluripotent transformation [[45]22].
   Additionally, we found that SMAD3 can directly bind to the promoter of
   the key pluripotency gene, Nanog [[46]22]. Consistent with these
   findings, exogenous activation of SMAD3 in the early stage of SSC
   culture (germline commitment) led to cell death, whereas its activation
   of SMAD3 in transforming SSCs (with a pluripotent cell fate) increased
   transformation efficiency [[47]22]. This oscillatory change in SMAD3
   expression—initially decreasing and subsequently increasing—suggests
   that inhibiting SMAD3 expression in the early stage of transformation
   might be a prerequisite for spontaneous reprogramming of SSCs. However,
   the molecular underpinnings of SMAD3 fluctuations and cell fate
   transformation remain unclear.

   A more recent study from our laboratory reported that addition of
   EGF/LIF to SSC culture medium significantly improved the transformation
   efficiency of SSCs maintained on fresh mouse embryonic fibroblast (MEF)
   feeder layers, particularly noting that EGF signaling activates Smad3
   expression through RAC1 during SSC transformation [[48]44]. Moreover,
   transcriptomic analysis revealed significant changes in the expression
   of metabolism-associated genes, especially those related to
   carbohydrate and energy metabolism, during the initiation stage of SSC
   transformation [[49]44]. Increased glycolytic activity, a hallmark of
   rapid proliferation in both tumor cells and stem cells, is
   characterized by a preference for glycolysis even in the presence of
   adequate oxygen—a phenomenon known as aerobic glycolysis or the Warburg
   effect. This dysregulation of glycolysis provides cells with essential
   nutrients and ATP, restructuring the microenvironment to facilitate
   rapid growth and metastasis by acidifying the extracellular milieu and
   disrupting the extracellular matrix [[50]11, [51]23]. These
   observations imply some potential connections between SMAD3 and energy
   metabolism in SSCs spontaneous reprogramming.

   SMAD3 is a key transcription factor downstream of classical TGF-β
   signaling and plays a pivotal role in the regulation of cell
   proliferation, apoptosis, differentiation, epithelial–mesenchymal
   transition, tumor suppression and cancer promotion [[52]7, [53]42].
   Accumulating evidence suggests that SMAD3 activity is closely
   associated with metabolites, for example, the reduced de novo fatty
   acid synthesis and increased fatty acid β-oxidation lead to
   intracellular acetyl-CoA accumulation, which promotes SMAD3 acetylation
   and entry into the nucleus to regulate endoderm differentiation
   [[54]47]. Based on these observations, we proposed that SMAD3 might
   function as a molecule that connects energy metabolism and cell fate
   determination in the SSCs' spontaneous reprogramming.

   Notably, the shift from TCA to aerobic glycolysis entails a
   reprogramming of glucometabolic gene expression, orchestrated by
   Hypoxia-inducible factor (HIF)-1α, a transcription factor with a
   helix-loop-helix structure that is activated in response to hypoxia and
   homeostasis. HIF-1α directly or indirectly regulates the expression of
   several genes involved in cell differentiation, including Ldha (lactate
   dehydrogenase-A), Vegf (vascular endothelial growth factor), Flt1
   (FMS-associated tyrosine kinase 1), Pdgf-β (platelet-derived growth
   factor-β), bFgf (basic fibroblast growth factor), and other genes that
   affect glycolysis [[55]17, [56]33], suggesting that it plays a role in
   the regulation of glycolytic activity. More importantly, HIF-1α can
   bind to the MH2 domain of phosphorylated SMAD3 and switch the function
   of TGF-β to promoting glycolysis under hypoxic conditions by altering
   its SMAD binding partner [[57]10]. In addition, in tumor cells,
   increasing evidence has revealed the connections among HIF-1α,
   TGF-β/SMAD3 signaling and glycolysis: TGF-β signaling inhibits
   glycolysis under normoxic conditions but significantly promotes
   glycolysis under hypoxic conditions in tumor cells in vitro and in
   vivo, and SMAD3 plays a key role in the metabolic reprogramming
   associated with tumor progression, including tumor glycolysis [[58]46],
   glutaminolysis [[59]2], and lipid metabolism [[60]20]. These findings
   imply that HIF-1α plays a role in SMAD3-mediated energy metabolism
   reprogramming during SSC transformation.

   In this study, we aim to delineate the cooperative roles of SMAD3 and
   HIF-1α in driving metabolic reprogramming toward aerobic glycolysis and
   to establish this transition as a prerequisite for the spontaneous
   pluripotency acquisition in SSCs. By integrating transcriptomic,
   metabolomic, and functional analyses, we further elucidate how
   metabolic-epigenetic crosstalk orchestrates SSC fate determination. Our
   study provides new insights into the mechanism of spontaneous
   reprogramming of SSCs, offering valuable information for the treatment
   of reproductive diseases and for future advancements in stem cell
   research and animal cloning technology.

Results

Induction of SSC pluripotency via an optimized culture system

   To induce spontaneous reprogramming, a long-term cultured SSCs cell
   line was established. SSCs, isolated from the testes of neonatal mice,
   were purified and cultured on MEFs for 4–5 days (Fig. [61]1A). The
   resulting colonies exhibited a grape-like cluster morphology,
   characteristic of SSCs, and could be stably maintained on a MEF feeder
   layer for more than 30 passages (Fig. [62]1B), exhibiting a robust
   proliferation rate throughout the culture process (Fig. [63]1C). To
   confirm that the obtained cells were SSCs, total RNA was extracted and
   subjected to RT‒PCR analysis, which revealed the expression of key
   germline markers, including Id4, Plzf, Mvh, Gfrα1, Itgα6, Itgβ1, and
   c-Kit (Fig. [64]1D). Immunofluorescence (IF) assays further confirmed
   the expression of germ cell marker MVH, undifferentiated spermatogonia
   marker PLZF, and SSCs marker ID4 (Figure [65]S1). The percentages of
   PLZF-positive and GFRA1-positive cells were 92.38 ± 1.46% and
   88.53 ± 1.86%, respectively (Fig. [66]1E), suggesting that SSCs were
   efficiently enriched in the obtained cells.

Fig. 1.

   [67]Fig. 1
   [68]Open in a new tab

   Characterization and Pluripotent Transition of Spermatogonial Stem
   Cells (SSCs). (A) Schematic of SSC isolation and culture on mouse
   embryonic fibroblast (MEF) feeder layer. (B) Clonal morphology across
   different passages of SSCs. (C) Growth kinetics of SSCs. (D)
   Identification of SSC-specific markers by RT-PCR. (E) Quantitative
   analysis of the proportion of PLZF + and GFRA1 + cells. (F) Observation
   of a transforming state in SSCs during prolonged culture. (G)
   Representative colonies of transformed GSPCs from long-term SSC
   cultures. (H) Assessment of germline and pluripotency marker expression
   by Western blot. (I, J) IF assays revealing pluripotency marker
   expression in GSPCs (I: NANOG, PLZF, DAPI, merged; J: SOX2, SSEA1,
   DAPI, merged). Data are presented as mean ± SD (*p < 0.05; **p < 0.01).
   Scale bars, 20 μm

   As described in a previous study [[69]44], SSCs cultured on fresh MEF
   feeder layers for approximately 5–6 generations could be transformed
   into pluripotent state cells after 2–3 generations exposed to 10 ng/mL
   LIF and 20 ng/mL EGF (Fig. [70]1F). These cells were identified as
   being in an intermediate state of transformation [[71]44]. After
   sustained culture for an additional 5 generations, all colonies adopted
   an epithelium-like phenotype (Fig. [72]1G). These transformed cells
   highly expressed the pluripotency markers SOX2, OCT4, and NANOG but
   expressed very low levels of the germ cell marker MVH and the SSC
   marker PLZF (Fig. [73]1H). IF assays corroborated these findings,
   revealing nearly ubiquitous expression of NANOG, SOX2 and SSEA1 but no
   expression of PLZF (Fig. [74]1I and J). In summary, these data indicate
   that SSCs undergo a transition to a pluripotent state upon the addition
   of LIF and EGF; these cells are termed germline stem cell-derived
   pluripotent stem cells (GSPCs) [[75]44]. The proportion of GSPCs
   progressively increased with continued culture and completed
   transformation after approximately 5 passages, accompanying with
   metabolic changes [[76]44], while the mechanism remained unclear.

Metabolic reprogramming and changes in energy flux during the transformation
of SSCs to GSPCs

   To explore the role of metabolic changes during SSC transformation,
   especially the connection of transcriptomic alterations such as
   Ras/Rac1/p38, TGF-β and JAK-STAT signaling pathways and energy
   metabolisms, we harvested stable cultured SSCs (p10 in culture medium),
   intermediate-state cells (p10 SSCs cultured in transformation medium
   for 2 passages, referred to as In state cells), and stable GSPCs (p10
   SSCs cultured in transformation medium for 12 passages) for
   metabolomics analysis (Figure [77]S2A). In total, 42 metabolites
   involved in energy metabolism were detected by using LC‒MS/MS and a
   self-built metabolic database from the Maiwei Company. The results were
   subjected to principal component analysis (PCA) and cluster analysis
   (Figure [78]S2B), which revealed the pattern of differences in
   metabolite abundance among the 3 groups (Fig. [79]2A, Table [80]S3).

Fig. 2.

   [81]Fig. 2
   [82]Open in a new tab

   Metabolic reprogramming dynamics observed during SSC transformation.
   (A) Sample cluster diagram. (B) Volcanic map of SSC p10 vs. GSPCs p2
   differential metabolites. (C) KEGG enrichment analysis of upregulated
   metabolites. (D) Bubble chart of changes in the content of pyruvate and
   downstream metabolites. (E) Radar chart of changes in the content of
   some amino acids and nucleotides (The values in the figure represent
   Fold Change). (F) Comparison of L-Glutamic acid and glutamine content
   with D-Glucose-6-phosphate content. (G) Cluster analysis of metabolites
   in SSC p10 and GSPCs p12. (H) KEGG enrichment analysis of up-regulated
   metabolites in GSPCs p12. (I-K) Overall trends of different metabolic
   pathways during transformation: change trend of metabolites involved in
   glycolysis pathway (I), change trend of metabolites involved in PPP
   (J), and change trend of metabolites involved in TCA (K). The
   metabolites displayed in this figure were specifically selected based
   on their significantly elevated expression profiles observed in GSPCs
   through targeted metabolomics screening

Metabolic changes during the initial stage of SSC pluripotent transformation

   Previous study revealed that after two passages in the presence of
   EGF/LIF, SSCs enter the initial stage of the transition to a
   pluripotent state, probably accompanying with metabolic reprogramming,
   especially changes in energy metabolism, as a critical event in
   reprogramming. Therefore, we compared and analyzed the altered
   metabolic patterns in SSCs and In state cells. Metabolomic analysis
   revealed 25 differentially abundant metabolites between SSCs and In
   state cells, of which 20 showed increased abundance (p ≤ 0.05, fold
   change ≥ 1.5) and 5 showed decreased abundance (p ≤ 0.05, fold
   change ≤ 0.5) (Fig. [83]2B). The upregulated metabolites included
   glucose 6-phosphate, phosphoenolpyruvic acid, pyruvate, D-5-ribulose
   phosphate, and 7-sedoheptulose phosphate. We performed metabolic
   pathway enrichment analysis of the 20 metabolites with increased
   abundance by using the SMPDB database and MetaboAnalyst 5.0, and a
   total of 43 metabolic pathways were enriched. There were 6 pathways for
   which P was ≤ 0.05: glycolysis, gluconeogenesis, the pentose phosphate
   pathway, lactose synthesis, carnitine synthesis and ribonucleotide
   metabolism (Fig. [84]2C). This indicates that during the early stage of
   transformation, the glucose metabolism patterns of the cells changed
   significantly, and the activities of the glycolysis and pentose
   phosphate pathways were markedly increased. ATP, Inline graphic
   -D-fructofuranose 1,6-bisphosphate, 6-phospho-D-gluconic acid,
   L-citrulline and adenine exhibited decreased abundance (Fig. [85]2A,
   Table [86]S3). The observed decrease in the concentration of ATP, the
   main intracellular energy carrier, further proves that an increase in
   glycolytic activity leads to a decrease in the efficiency of
   intracellular energy production.

   We noted a substantial increase in pyruvate concentrations, with a fold
   change of 5.28, during the initial phase of SSC pluripotent
   transformation (Table [87]S3). Given the critical role of pyruvate in
   maintaining stem cell pluripotency [[88]32], we examined alterations in
   the downstream metabolic products of pyruvate (Fig. [89]2D). Alanine,
   which can be directly synthesized from pyruvate via a transamination
   reaction, displayed a negligible fold change of 0.97 (Table [90]S3).
   Although pyruvate predominantly enters the mitochondrial TCA cycle
   under aerobic conditions, only the succinate level was significantly
   upregulated, with a fold change of 12.04. Neither α-ketoglutarate (fold
   change = 1.04) nor citrate (fold change = 0.98) exhibited notable
   changes in abundance (Table [91]S3). This observation aligns with prior
   studies indicating a reduction in the protein level of dihydrolipoyl
   transacetylase within the pyruvate dehydrogenase complex in GSPCs.
   Lactate, an anaerobic byproduct of pyruvate metabolism, exhibited a
   fold change of 1.65, but this change was not statistically significant
   (p value = 0.12) (Table [92]S3). Moreover, phosphoenolpyruvate, a key
   intermediate in gluconeogenesis, exhibited a fold change of 2.45 (Table
   [93]S3). Metabolomic analyses revealed augmented nucleotide and amino
   acid biosynthesis activity, necessitating precursor availability and
   reducing flux through both the glycolysis and pentose phosphate
   pathways (Fig. [94]2E). These findings suggest that the increase in
   pyruvate abundance, attributed to enhanced glycolytic flux, does not
   contribute to oxidative metabolism via the TCA cycle but rather
   facilitates glucose homeostasis through gluconeogenesis and lactate
   clearance.

   Subsequently, we checked on the concentration of glutamine during the
   early dedifferentiation stages of SSCs, since it’s essential for amino
   acids, nucleotides, and fatty acids and intracellular protein synthesis
   [[95]28], as well as stemness maintenance [[96]51]. In the In state
   cells, glutamine and glutamate exhibited the highest absolute
   concentrations among all assessed metabolites, which were 146-fold and
   56-fold greater than that of 6-phosphogluconate, respectively. The fold
   changes in glutamine and glutamate relative to the baseline levels were
   1.5 and 2.7, respectively (Fig. [97]2F). These data indicate an
   increased cellular preference for glutamine during the initial stages
   of pluripotent transformation, facilitating increased biosynthetic
   activity within the cell.

Metabolic changes during SSC-to-GSPC transformation

   Following the addition of growth factors and subsequent culture through
   the 12th generation, the SSCs transitioned to pluripotent stem cells.
   Thus, the associated changes in metabolic patterns reflect the
   metabolic preferences of pluripotent cells, providing insights into the