Skip to main content
NCBI home page
NAR Molecular Medicine logoLink to NAR Molecular Medicine
. 2025 Mar 25;2(2):ugaf008. doi: 10.1093/narmme/ugaf008

High-throughput transcriptome analysis reveals a developmental increase in Acvr1c which mediates epigenetic repression of the gene encoding the pubertal brake, Makorin ring finger protein 3

Dor Shalev 1, Gil Golan 2, Lilach Pnueli 3, Anat Kahan 4, Yael Mandel-Gutfreund 5, Philippa Melamed 6,✉
PMCID: PMC12429960  PMID: 41255711

Abstract

Makorin ring finger protein 3 (MKRN3) acts as a “pubertal brake.” MKRN3 loss-of-function mutations are the most common genetic cause of central precocious puberty, and its expression in the brain drops significantly towards puberty, yet the drivers responsible for this reduction remain unclear. We aimed here to identify factors responsible for repressing Mkrn3 expression across the pubertal transition, initially through bioinformatic analysis of developmental RNA-seq datasets from rodent tissues. Genes whose expression correlated significantly with Mkrn3, and are linked to puberty and signaling, were identified. Notably, these included Acvr1c, whose knockout was shown previously to delay puberty in mice. Experimentally, we show that Acvr1c activation suppresses Mkrn3 via Smad2/3 signaling, involving recruitment of Kap1 and repressive histone modifications. These findings provide mechanistic explanation for the reduction in Mkrn3 required for pubertal onset, while underscoring the value of integrating high-throughput gene expression analyses with experimental validation to uncover novel regulatory pathways.

Graphical Abstract

Graphical Abstract.

Graphical Abstract

Open in a new tab

Introduction

Central precocious puberty (CPP) is strongly associated with loss-of-function mutations in the Makorin ring finger protein 3 (MKRN3) gene, and these are currently the most common mutations implicated in CPP [1–4]. In the rodent and monkey hypothalamus, which comprises the control center of the reproductive axis, MKRN3 mRNA levels decrease markedly leading up to puberty and remain low thereafter [3, 5]. These observations, together with the various roles proposed for MKRN3 in hypothalamic neurons [5–9], suggest its crucial role in inhibiting the initiation of puberty [2, 3, 10, 11]. However, the specific factors and pathways responsible for the pre-pubertal drop in MKRN3 levels have remained largely unclear.

Several studies have reported mechanisms that might regulate MKRN3 expression. One of these focused on the 3′ UTR of Mkrn3 and demonstrated a role for mir30b whose levels increase during mouse post-natal development. This microRNA has a binding site in the 3′UTR of Mkrn3 which, when blocked, increased Mkrn3 protein levels in hypothalamic neurons and delayed puberty in female rats [12]. Epigenetic regulation of the promoter has also been suggested to play a role, given that promoter methylation patterns were found to change across puberty in the hypothalamus of female mice [13]. In that study, 29 different transcription factors were predicted to bind the promoter, some perhaps dependent upon these changes in the DNA methylation [13].

Additional insights on possible transcriptional regulation of MKRN3 have been gained from mutations identified in the MKRN3 promoter that are associated with CPP. One group examined three such mutations, and suggested that two of them could lead to impaired binding of the transcription factors, HMX2 and PRDM14, while the third might create a binding site for SOX4 [14]. In a different study, a girl with CPP was found to have a 4 nt deletion in a putative binding site for the downstream responsive element antagonist modulator (DREAM) DNA-binding protein, and deletion of this element reduced reporter gene activity in a GnRH neuronal cell line [10]. Aside from these focused studies, Kap1 (KRAB-associated protein-1 or TRIM28), was found serendipitously to regulate Mkrn3 expression levels, following a screen on the effects of its knockout (KO) in the forebrain of mice. The Kap1 KO resulted in significant increase in Mkrn3 mRNA levels in the hippocampus [15], though this was not the focus of that study, and its role in regulating Mkrn3 has seemingly not been studied further.

Although these findings hint at possible functional sites in the MKRN3 promoter and various putative regulatory factors, their definitive roles in determining MKRN3 expression in the context of pre-pubertal development, and connection to pubertal timing, have not been shown. Moreover, the reduction in MKRN3 levels during late childhood is presumably driven by upstream signals and pathways that modulate transcription factor activity, so identifying these mechanisms is essential for understanding the processes that enable pubertal onset.

We sought here to identify the signaling pathways and factors that regulate Mkrn3 expression during this pre-pubertal period, initially through bioinformatic analysis of bulk and single-cell RNA-seq datasets from brain and hypothalamic tissues across rodent development. We identified several genes whose expression is highly correlated with that of Mkrn3, that are related specifically to both signaling and puberty. Amongst these, genes involved in transforming growth factor beta (TGFβ) signaling pathways were notably enriched. These included Acvr1c, whose mRNA levels were strongly negatively correlated with those of Mkrn3 in all datasets, and whose KO was found previously to delay pubertal onset in female mice [16]. Acvr1c is widely expressed in the hypothalamus, and our experimental work demonstrated that its activation and downstream signaling inhibits Mkrn3 expression, involving recruitment of Kap1 and the induction of repressive histone modifications.

Materials and methods

Bioinformatic analysis

The RNA-seq datasets were from mice forebrain and hindrain [17], rat forebrain and hindbrain [17], and rat MBH [18], processed in R-program. Analysis of variance (ANOVA) followed by Wilcoxon signed-rank test was performed on each dataset to determine significant changes in expression at each timepoint relative to the preceding timepoint. In the bulk RNA sequencing datasets, differential expression analysis between the first timepoint that Mkrn3 levels dropped and the one prior to it, was done with DEseq2 [19] using iDEP [20]. Pearson correlation coefficients between expression levels of Mkrn3 and those of all other genes in the datasets were calculated based on the set of average values at each of the timepoints (n = 4 in all datasets) normalized and transformed (log2(CPM + 4)) counts. After combining the datasets, genes were retained if their differential expression was significant (Padj < 0.05) in all datasets, and absolute Pearson coefficient 0.6 or higher in at least 3 of the 5 datasets. All figures in this section were generated using R.

For scRNA seq data analysis, the R data serialization (RDS) and uniform manifold approximation and projection (UMAP) data files containing the Seurat object and the cell annotations created in [21] were downloaded from GEO database (GSE132355) and analyzed in R. The data were filtered initially to contain just the cells from the postnatal stages (P4, P8, P14, and P45). The filtering of hypothalamic neurons was based on the original cell annotations, and these cells filtered to contain only those that express Mkrn3 and/or Acvr1c. Plots were generated using the Seurat package in R [22].

GO analysis was performed using gprofiler [23]. The selection of signaling-related genes was based on the GO database (GO:0023052). The list of genes associated with puberty was created from two datasets. The first contains genes in GWAS publications from the HuGE Navigator Gene-Phenotype Associations dataset [24], and was obtained using the search words “Puberty Precocious” from the Harmonizome 3.0 browser [25]. The second dataset was a curated list from Hou et al. [26]. These two datasets were merged for a collection of genes associated with puberty (Supplementary Table S1).

Plasmid constructs

The human CA-ACVR1C (T194D) and DN-ACVR1C (K222R) pcDNA plasmids [27, 28], both gifts from Dan Bernard (McGill University, Montreal), were cut using Nde1 and Sac2. The inserts were cloned into pCMV plasmid (pEGFP-C1, Clontech) cut using the same enzymes (so as not to include GFP) and verified by sequencing. Plasmids were linearized using AseI for stable transfections.

Cell culture and plasmid transfections

GT1-7 mouse hypothalamic GnRH neuronal cell line was cultured, as previously [29] in high glucose Dulbecco’s modified Eagle Medium containing 10% FBS, 1% penicillin–streptomycin and sodium pyruvate (all from Biological Industries, Beit Haemek), maintained at 37°C with 5% CO2 at 50%–90% confluency, passaging 1–2 times a week. Cells were treated for 24 h with Activin A (5 nM, SRP6057 Sigma), Activin B (7.7 nM [100 ng/ml], 8260-AB-010/CF, R&D Systems), or SB431542 (10 μM, #13031 Cayman), doses based on published studies [30–32]. Stable transfections were carried out in cells at 50% confluency, using Lipofectamine 3000 reagent (L3000008, Invitrogen), according to the manufacturer’s instructions, with linearized plasmids, and fresh medium replaced after 24 h, followed by selection using G418 (400 μg/ml, Sigma).

RNA extraction and real-time quantitative PCR

Total RNA was extracted using TRIzol (Ambion), treated with DNase, and purified using RNA Clean & Concentrator-5 Kit (R1014, Zymo Research). The complementary DNA (cDNA) was synthesized using qScript cDNA Synthesis Kit (Quanta Biosciences) according to the manufacturer’s instructions. Real-time quantitative PCR (qPCR) was carried out using the PerfeCTa SYBR Green FastMix (Quanta Biosciences) using primers listed in Supplementary Table S2. Amplicon levels were quantified using standard curves and normalized to the Rplp0 housekeeping gene.

Chromatin Immunoprecipitation

Chromatin immunoprecipitation (ChIP) experiments were carried out in formaldehyde cross-linked GT1-7 cells, as previously [29, 33–35], after extensive sonication. Antibodies (5 μl of antibody per 10 μg of chromatin) were Smad2/3 (D7G7, Cell Signaling), Kap1 (4123S, Cell Signaling), H3 (ab1791, Abcam), H3K27Ac (ab4729, Abcam), H3K9me3 (ab8898, Abcam), and IgG (ab6721, Abcam). After DNA purification, the regions of interest were amplified by qPCR (primers in Supplementary Table S2) from IP and input samples. Levels are shown as IP/input, and for histone modifications, they are normalized to H3 IP/input levels.

Statistical analysis

The experimental data are from multiple biological repeats (n-value) which were assayed individually. F-test determined similarity of variance and data analyzed accordingly by Student’s t-test (two-tailed) to establish significance, defined as P < 0.05. Graphs show mean ± SEM. For multiple group testing, ANOVA followed by Tukey’s HSD post-hoc test was used.

Results

Analysis of RNA sequencing data identifies genes whose expression levels are strongly correlated with those of Mkrn3 across development

To find genes whose expression levels are correlated with those of Mkrn3, we analyzed five publicly available RNA-seq datasets from forebrains and hindbrains of mice and rats, and from rat medio-basal hypothalamus (MBH) across development, including the juvenile to adulthood transition [17, 18] (Fig. 1A). We first identified the time point by which Mkrn3 has dropped in each dataset (Fig. 1B). We then employed differential expression analysis [20] across this time point, to detect genes with changes in expression that correlate with those of Mkrn3. In addition, we calculated the Pearson correlation coefficient between the expression of Mkrn3 and that of all the other genes in each of the datasets, along all time points during development. These analyses identified 1192 genes whose expression levels are correlated positively or negatively (r > 0.6) with those of Mkrn3 in all of the datasets throughout development, and are differentially expressed (Padj < 0.05) at the same time that Mkrn3 drops in at least three of the datasets (Fig. 1C). These genes are enriched for processes related to neuronal differentiation and development, as revealed by GO analysis (Supplementary Fig. S1A).

Figure 1.

Figure 1.

Open in a new tab

Analysis of RNA sequencing data identifies genes whose expression levels are correlated with those of Mkrn3across development. (A) Datasets and the pipeline for identifying genes whose expression is highly correlated with that of Mkrn3, and related to both signaling and puberty. (B) Boxplots show the expression levels of Mkrn3in the different datasets across development. Statistical analysis was conducted using ANOVA for each dataset (all P < 0.001), followed by Wilcoxon signed-rank tests to compare each time point with the one immediately preceding it; *P < 0.05, otherwise nonsignificant (P > 0.05). (C) The 1192 genes whose expression levels are correlated (r > 0.6) positively or negatively with those of Mkrn3in all of the datasets throughout development and are differentially expressed (Padj < 0.05) at the same time that Mkrn3drops in at least three of the datasets, shown as log2 of the fold-change (logFC) between the level at the time point that Mkrn3dropped, and at the immediately preceding timepoint. (D) The intersection between these 1192 genes, and genes associated with puberty and/or signaling, resulting in (E) 19 candidate genes (correlation value: Pearson coefficient). See also Supplementary Fig. S1.

The genes arising from this screen might encode regulators of Mkrn3, be affected by Mkrn3, or perhaps be influenced by the same pathways, explaining their correlated expression. Given that we were looking to identify factors that regulate Mkrn3 transcription, we focused on genes that are associated with signaling pathways, hypothesizing that changes in their activity through pre-pubertal development could regulate Mkrn3expression. Among the initial candidates, 450 genes are associated with the term “signaling” (Fig. 1D). We further filtered this list to pinpoint genes associated with puberty or pubertal onset. This narrowed the dataset to 19 genes whose expression levels are highly correlated with those of Mkrn3, related to signaling and associated with puberty (Fig. 1D and E).

Notably, 6 of these 19 genes are part of, or specifically regulated by TGFβ signaling pathways: activin receptors type 1C and 2A (Acvr1c, which is also known as Alk7, and Acvr2a), the downstream transcription factors Smad1 and Smad7, and Olfm2 and Mapk14 that are regulated by TGFβ [36, 37]. Acvr1c expression was strongly negatively correlated to Mkrn3 expression in all five of the datasets, and significantly increased at the same point that Mkrn3 drops (Fig. 1E and Supplementary Fig. S1B). This was particularly striking because Acvr1c knockout in mice was shown previously to delay puberty onset [16]. Moreover, Smad7 expression is induced by Acvr1c [38, 39] and, accordingly, its mRNA levels also increased throughout puberty in all of the datasets, in significant negative correlation with those of Mkrn3 (Fig. 1E). These findings suggest that the increase in Acvr1c expression across pre-pubertal development might be responsible for the reduction in Mkrn3 levels that allows puberty to occur.

Expression of Mkrn3 and Acvr1c is negatively correlated in mice hypothalamic neurons

Mkrn3 is clearly expressed throughout the brain, and the drop in its expression across puberty was evident in all of the datasets. However, the control center of the reproductive axis resides in the hypothalamic neurons, so we went on to analyze single-cell RNA sequencing (scRNA-seq) data from the hypothalami of mice at various developmental stages [21], focusing on postnatal (PN) days 4–45, in order to identify the cell types expressing Mkrn3, and determine whether the negative correlation with Acvr1c is apparent in these cells.

In the hypothalamus, Mkrn3 was seen to be expressed most highly in oligodendrocytes, tanycytes and neurons, while Acvr1c is expressed in the neurons and astrocytes (Fig. 2A). Thus, the only hypothalamic cell type expressing both Mkrn3 and Acvr1c are the neurons. To establish whether Mkrn3 and Acvr1c are negatively correlated in these cells, we analyzed the postnatal hypothalamic neurons that express either one, or both, of these genes. Strikingly, of the 511 cells identified, only two (from PN14) expressed both Mkrn3 and Acvr1c (Fig. 2B and C). These findings confirm that the dynamic and opposing changes in expression of these genes during development do occur in the hypothalamic neurons.

Figure 2.

Figure 2.

Open in a new tab

Expression of Mkrn3and Acvr1cis negatively correlated in mice hypothalamic neurons. (A) Dot plot showing Mkrn3 and Acvr1c expression in cells from hypothalami of P4–P45 mice (data from [21]). The size of the dot represents the percentage of cells in the cell type expressing the gene, and the brightness represents the average expression level (normalized counts) across all cells in the cluster. (B) UMAP of hypothalamic neurons from four postnatal stages which are color-coded (P4–P45), and Mkrn3 and Acvr1c expression levels (normalized counts) on this UMAP. (C) The normalized expression levels of Mkrn3 and Acvr1c in the cells from these different age groups are shown in violin and box plots.

The activity of Acvr1c and downstream Smad signaling pathway repress Mkrn3expression in a GnRH hypothalamic neuronal cell line

The above findings, together with previous reports of delayed puberty in Acvr1c KO mice [16], led us to surmise that Acvr1c-mediated pathways might inhibit Mkrn3 expression in hypothalamic neurons. To examine this possibility, we over-expressed either a constitutively-active (CA) or dominant-negative (DN) variant of ACVR1C in the murine GnRH hypothalamic neuronal GT1-7 cell line. This resulted in a significant reduction, or increase, in Mkrn3 mRNA levels, respectively, which contrasted very clearly with the responses observed for the Smad7 control gene (Fig. 3A and B).

Figure 3.

Figure 3.

Open in a new tab

The activity of Acvr1c and downstream Smad signaling pathway repress Mkrn3expression in a GnRH hypothalamic neuronal cell line. (A) A constitutively active (CA) or (B) dominant negative (DN) form of the ACVR1C was stably expressed in GT1-7 cells. Mkrn3 and Smad7 mRNA levels were measured by qPCR and are shown after normalization to Rplp0, relative to levels in wild-type GT1-7 cells. Mean ± SEM, n = 4; Student’s t-test was performed between the WT and over-expressing cells. *P< 0.05, **P< 0.01, ***P< 0.001. (CandD) GT1-7 cells were treated for 24 h with (C) activin A (5 nM) or (D) activin B (7.7 nM) in the presence or absence of the inhibitor SB431542 (10 μM). The mRNA levels of Mkrn3 and Smad7 were assessed and are presented as before, relative to levels in non-treated (NT) controls. Mean ± SEM, n = 3 (in C) or 6 (in D). ANOVA followed by Tukey’s HSD post-hoc test was performed and the same letter (in upper or lower case for each gene) denotes means that are not significantly different (P> 0.05). (E) ChIP was performed for Smad2/3 in GT1-7 WT and CA-ACVR1C cells, and IP and input levels measured across the locus by qPCR, using the primer pairs shown in the schematic beneath the graph (SBE: Smad-binding element). Smad7 serves as a positive control (PC) and ACVR1C-responsive gene; NC: negative control region. Results are shown as IP/input; mean ± SEM, n = 4. Student’s t-test compared levels between the two cell types, presented as before (Fig. 3A and B). See also Supplementary Fig. S2.

We next verified whether exposure of the cells to ACVR1C ligands also represses Mkrn3 expression, and treated the cells with activin A or activin B. Exposure to either ligand resulted in a significant drop in Mkrn3 mRNA levels which was abolished by addition of the ACVR1B/C inhibitor SB-431542 [30] (Fig. 3C and D), confirming that activation of this signaling pathway indeed inhibits Mkrn3 expression.

The SMAD2 and SMAD3 transcription factors, which mediate ACVR1C signaling, have been found at the MKRN3 promoter in other cell types (Supplementary Fig. S2). We thus performed ChIP to determine whether these factors bind the Mkrn3 locus also in the hypothalamic neuronal cells. Smad2/3 binding was not evident at the Mkrn3 promoter but was clearly enriched at two regions in the coding sequence, both of which contain Smad-binding element (SBE) motifs (according to JASPAR [40, 41], P = 10e-4). Notably, the Smad2/3 binding at both sites was significantly elevated in the cells over-expressing CA-ACVR1C, reaching similar levels to those at the positive control Smad7 gene promoter (Fig. 3E).

Activation of Acvr1c increases Kap1 binding and induces repressive chromatin at the Mkrn3locus

Acvr1c and its activated Smad2/3 signaling thus clearly repress Mkrn3 expression, while a previous report [15] indicated a role for Kap1 (Trim28) which was found to interact with Smad2 in a different context [42]. Moreover, Kap1 was seen to bind the Mkrn3 proximal promoter and 5′ coding sequence (CDS; CDS1 in Fig. 3E) at some of the same sites as Smad2/3 in various cell types (ChIP-Atlas data [43] and Supplementary Fig. S2). Our own ChIP-qPCR analysis in WT GT1-7 cells revealed that Kap1 is enriched at the Mkrn3 promoter at least to similar levels as at the positive control locus (Fig. 4A). In cells expressing CA-ACVR1C, Kap1 appeared to be recruited to the Mkrn3CDS, its levels increasing significantly at the Smad2/3-binding sites, while its levels at some of the promoter sites dropped (Fig. 4A).

Figure 4.

Figure 4.

Open in a new tab

Activation of Acvr1c increases Kap1 binding and induces repressive chromatin at the Mkrn3locus. (A) ChIP was performed for Kap1 in GT1-7 WT and CA-ACVR1C cells, as in Fig. 3E, analyzed after subtraction of background (IgG) signals, and presented similarly, n = 3. (B–D) ChIP was performed similarly for (B) H3, (C) H3K9me3, and (D) H3K27ac. In (B), H3 values are shown as IP/input, and in (C and D), levels are shown normalized to those of H3 at the same locus. Mean ± SEM, n = 3. Statistical analysis is by Student’s t-test between the cell types, and is presented as before (*P < 0.05, **P < 0.01, ***P < 0.001). See also Supplementary Fig. S2.

Kap1 acts as a KRAB-scaffold protein, inducing heterochromatin via recruitment of Setdb1 which catalyzes H3K9me3, and NuRD which repositions nucleosomes and deacetylates histones [44–46]. Notably both Setdb1 and the NuRD core protein, Chd4, have been found at the Mkrn3 promoter and 5′end of the CDS in other cell types (Supplementary Fig. S2). We therefore looked for any changes in the chromatin landscape at the Mkrn3 locus in the CA-ACVR1C expressing cells. ChIP for H3 indicated that the promoter might be more compact in the CA-ACVR1C compared to WT cells, though the increase in H3 levels was not significant at most of the individual regions tested (Fig. 4B). However, levels of H3K9me3 were significantly increased in the CA-ACVR1C-expressing cells at all sites examined (Fig. 4C). This contrasted strongly with the levels of H3K27ac which were markedly lower at all sites (Fig. 4D), revealing ACVRC1-induced repressive epigenetic regulation of the Mkrn3 gene.

Discussion

Through bioinformatic analysis of developmental RNA-seq datasets spanning puberty, combined with experimental validation, we have identified Acvr1c and its downstream signaling pathway as a repressor of Mkrn3 expression. Such a function for Acvr1c accords with previous reports that its knockout caused a delay in puberty onset in mice [16]. We now provide a mechanistic explanation for this earlier finding, given that the increase in Acvr1c expression in the brain prior to puberty explains the drop in Mkrn3 required for pubertal onset.

The reduction in Mkrn3 levels leading up to puberty was evident in all of the datasets examined, reflecting expression levels in different parts of the brain, although effects of Mkrn3 on pubertal timing most likely occur in the hypothalamus, and have been reported specifically in Kiss1 and GnRH neurons [5, 7, 9]. The fact that Mkrn3 and Acvr1c levels were negatively correlated in all of these datasets was striking and suggests a common regulatory function for this receptor. Although termed an activin receptor, like other TGFβ family receptors, ACVR1C can be activated by several different ligands, including Nodal and various growth differentiation factor (GDF) proteins [47–54], albeit with varying affinities. Several of these ligands are widely and highly expressed, playing keys roles in many aspects of development [55–57]. Such abundance of potential ligands presumably facilitates activation of this receptor in the various regions of the brain, triggering the same signaling pathway that leads to Mkrn3 repression. However, abundant availability of potential ligands also suggests that a limited number of receptors might comprise a limiting factor, supporting the notion that an increase in expression of ACVR1C during childhood could be key in triggering the drop in MKRN3 levels.

ACVR1C signals through SMAD2/3 proteins which can stimulate or repress gene transcription as a result of their interactions with various protein partners [58–62]. Our finding that the repressive ACVR1C-induced Smad2/3 localizes to the Mkrn3 CDS is consistent with previous observations for the position-dependent effects of transcription factors which can activate or inhibit transcription: such factors binding downstream of the transcriptional start-site (TSS) were found to be more likely repressive [63]. This Smad-mediated repression might involve various mechanisms beyond hindering progression of the polymerase, including recruitment of corepressors found to interact with these proteins in other contexts [58, 64–66]. However, previous reports that Smad2 can interact directly with Kap1 [42], and that Kap1 KO in the forebrain increased Mkrn3 levels [15], pointed strongly to a role for this corepressor in mediating the inhibitory effects of Acvr1c on Mkrn3. This hypothesis was substantiated when Kap1 binding at the Smad2/3 binding sites was seen to be elevated in cells expressing the active receptor, and by the changes in H3K9me3 and H3K27ac at the locus, indicating its downstream effects on the chromatin.

These altered histone modifications can be explained by the ability of Kap1 to interact with and recruit the chromatin modifiers Setdb1 and NuRD: Setdb1 is a methyltransferase that catalyzes H3K9me3 [45, 46, 67, 68], and NuRD contains the histone deacetylases HDAC1 or HDAC2 [44, 69]. The increase in H3K9me3 and drop in H3K27ac, seen across the locus, are both associated with repressive chromatin, and likely underlie the reduction in Kap1 binding at the Mkrn3 promoter. Kap1’s role at the promoter, which was not associated with Smad2/3, is not surprising given that it was found at this locus in other cell types, and has been shown to have both facilitating and repressive roles in transcription, via interaction with various proteins and its E3 ubiquitin and sumo ligase activity [42, 45, 70, 71]. These effects are likely position-dependent and, given the effects of its KO on Mkrn3 expression in the brain, appear secondary to its repressive actions.

In summary, our study identifies the Acvr1c pathway as an epigenetic regulator of Mkrn3 expression and elucidates molecular mechanisms driving changes in the chromatin landscape at the Mkrn3 locus. This finding provides a mechanistic explanation for the reduction in Mkrn3 levels required for pubertal onset, particularly pertinent for understanding disorders of sexual maturation and pubertal timing. Furthermore, our approach of integrating high-throughput analyses of gene expression changes across development with targeted experimental validation demonstrates the power of this strategy for uncovering novel regulatory pathways.

Supplementary Material

ugaf008_Supplemental_File
ugaf008_Supplemental_File.pdf (545.3KB, pdf)

Acknowledgements

We thank D Bernard (McGill University, Montreal) for the CA-ACVR1C and DN-ACVR1C plasmids, and Ronni Avrahami (Hebrew University, Rehovot) for technical help.

Author contributions: Conceptualization: D.S., P.M., and Y.M.G. Methodology: D.S., G.G., L.P., and A.K. Formal analysis: D.S. and P.M. Investigation: D.S. Writing original draft: D.S. and P.M. Writing—reviewing and editing. P.M., Y.M.G., D.S., and A.K. Visualization: D.S. Supervision: P.M. and Y.M.G. Project administration and funding acquisition: P.M. and Y.M.G.

Contributor Information

Dor Shalev, Faculty of Biology, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Gil Golan, Faculty of Biology, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Lilach Pnueli, Faculty of Biology, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Anat Kahan, Faculty of Agriculture, Hebrew University of Jerusalem, Rehovot 7610001, Israel.

Yael Mandel-Gutfreund, Faculty of Biology, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Philippa Melamed, Faculty of Biology, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Supplementary data

Supplementary data is available at NAR Molecular Medicine online.

Conflict of interest

None declared.

Funding

The research was supported by the Israel Science Foundation (1850/17 and 334/23 to P.M.).

Data availability

Plasmids generated in this study will be distributed subject to restrictions of the original sources of their components. Datasets used are publicly available as cited in the manuscript. The code has been uploaded and is available https://doi.org/10.6084/m9.figshare.28623272.v2.

References

  • 1.Brito VN, Canton APM, Seraphim CEet al.. The congenital and acquired mechanisms implicated in the etiology of Central precocious puberty. Mol Cell. 2023; 44:39–53. 10.1210/endrev/bnac020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.De Vries L, Gat-Yablonski G, Dror Net al.. A novel MKRN3 missense mutation causing familial precocious puberty. Hum Reprod. 2014; 29:2838–43. 10.1093/humrep/deu256. [DOI] [PubMed] [Google Scholar]
  • 3.Abreu AP, Dauber A, Macedo DBet al.. Central precocious puberty caused by mutations in the imprinted gene MKRN3. N Engl J. Med. 2013; 368:2467–75. 10.1056/NEJMoa1302160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Canton APM, Seraphim CE, Montenegro LRet al.. The genetic etiology is a relevant cause of central precocious puberty. Eur J Endocrino. 2024; 190:479–88. 10.1093/ejendo/lvae063. [DOI] [PubMed] [Google Scholar]
  • 5.Abreu AP, Toro CA, Song YBet al.. MKRN3 inhibits the reproductive axis through actions in kisspeptin-expressing neurons. J Clin Invest. 2020; 140:4486–500. 10.1172/JCI136564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Liu H, Kong X, Chen F. Mkrn3 functions as a novel ubiquitin E3 ligase to inhibit Nptx1 during puberty initiation. Oncotarget. 2017; 8:85102–9. 10.18632/oncotarget.19347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Li C, Lu W, Yang Let al.. MKRN3 regulates the epigenetic switch of mammalian puberty via ubiquitination of MBD3. Natl Sci Rev. 2020; 7:671–85. 10.1093/nsr/nwaa023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Naulé L, Mancini A, Pereira SAet al.. MKRN3 inhibits puberty onset via interaction with IGF2BP1 and regulation of hypothalamic plasticity. JCI Insight. 2023; 8:e164178. 10.1172/jci.insight.164178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Li C, Han T, Li Qet al.. MKRN3-mediated ubiquitination of Poly(A)-binding proteins modulates the stability and translation of GNRH1 mRNA in mammalian puberty. Nucleic Acids Res. 2021; 49:3796–813. 10.1093/nar/gkab155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Macedo DB, França MM, Montenegro LRet al.. Central precocious puberty caused by a heterozygous deletion in the MKRN3 promoter region. Neuroendocrinology. 2018; 107:127–32. 10.1159/000490059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Shalev D, Melamed P. The role of the hypothalamus and pituitary epigenomes in central activation of the reproductive axis at puberty. Mol Cell Endocrinol. 2020; 518:111031. 10.1016/j.mce.2020.111031. [DOI] [PubMed] [Google Scholar]
  • 12.Heras V, Sangiao-Alvarellos S, Manfredi-Lozano Met al.. Hypothalamic miR-30 regulates puberty onset via repression of the puberty-suppressing factor, Mkrn3. PLoS Biol. 2019; 17:e3000532. 10.1371/journal.pbio.3000532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Fanis P, Morrou M, Tomazou Met al.. Methylation status of hypothalamic Mkrn3 promoter across puberty. Front Endocrinol. 2023; 13:1075341. 10.3389/fendo.2022.1075341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Fanis P, Skordis N, Toumba Met al.. Central precocious puberty caused by novel mutations in the promoter and 5′-UTR region of the imprinted MKRN3. Gene Front Endocrinol. 2019; 10:677. 10.3389/fendo.2019.00677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Jakobsson J, Cordero MI, Bisaz Ret al.. KAP1-Mediated epigenetic repression in the forebrain modulates behavioral vulnerability to stress. Neuron. 2008; 60:818–31. 10.1016/j.neuron.2008.09.036. [DOI] [PubMed] [Google Scholar]
  • 16.Sandoval-Guzmán T, Göngrich C, Moliner Aet al.. Neuroendocrine control of female reproductive function by the activin receptor ALK7. FASEB J Off Publ Fed Am Soc Exp Biol. 2012; 26:4966–76. 10.1096/fj.11-199059. [DOI] [PubMed] [Google Scholar]
  • 17.Cardoso-Moreira M, Halbert J, Valloton Det al.. Gene expression across mammalian organ development. Nature. 2019; 571:505–9. 10.1038/s41586-019-1338-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Toro CA, Wright H, Aylwin CFet al.. Trithorax dependent changes in chromatin landscape at enhancer and promoter regions drive female puberty. Nat Commun. 2018; 9:57. 10.1038/s41467-017-02512-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15:550. 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Ge SX, Son EW, Yao R. iDEP: an integrated web application for differential expression and pathway analysis of RNA-seq data. BMC Bioinf. 2018; 19:534. 10.1186/s12859-018-2486-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kim DW, Washington PW, Wang ZQet al.. The cellular and molecular landscape of hypothalamic patterning and differentiation from embryonic to late postnatal development. Nat Commun. 2020; 11:4360. 10.1038/s41467-020-18231-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Satija R, Farrell JA, Gennert Det al.. Spatial reconstruction of single-cell gene expression data. Nat Biotechnol. 2015; 33:495–502. 10.1038/nbt.3192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Kolberg L, Raudvere U, Kuzmin Iet al.. g:profiler—interoperable web service for functional enrichment analysis and gene identifier mapping (2023 update). Nucleic Acids Res. 2023; 51:W207–12. 10.1093/nar/gkad347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Yu W, Gwinn M, Clyne Met al.. A navigator for human genome epidemiology. Nat Genet. 2008; 40:124–5. 10.1038/ng0208-124. [DOI] [PubMed] [Google Scholar]
  • 25.Rouillard AD, Gundersen GW, Fernandez NFet al.. The harmonizome: a collection of processed datasets gathered to serve and mine knowledge about genes and proteins. Database. 2016; 2016:baw100. 10.1093/database/baw100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hou H, Uuskü La-Reimand L, Makarem Met al.. Gene expression profiling of puberty-associated genes reveals abundant tissue and sex-specific changes across postnatal development. Hum Mol Genet. 2017; 26:3585–99. 10.1093/hmg/ddx246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Bernard DJ, Lee KB, Santos MM. Activin B can signal through both ALK4 and ALK7 in gonadotrope cells. Reprod Biol Endocrinol. 2006; 4:52. 10.1186/1477-7827-4-52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Xu G, Zhong Y, Munir Set al.. Nodal induces apoptosis and inhibits proliferation in Human epithelial ovarian cancer cells via activin receptor-like kinase 7. J Clin Endocrinol Metab. 2004; 89:5523–34. 10.1210/jc.2004-0893. [DOI] [PubMed] [Google Scholar]
  • 29.Bar-Sadeh B, Pnueli L, Keestra Set al.. Srd5a1 is differentially regulated and methylated during prepubertal development in the ovary and hypothalamus. J Endocr Soc. 2023; 7:bvad108. 10.1210/jendso/bvad108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Inman GJ, Nicolás FJ, Callahan JFet al.. SB-431542 is a potent and specific inhibitor of transforming growth factor-beta superfamily type I activin receptor-like kinase (ALK) receptors ALK4, ALK5, and ALK7. Mol Pharmacol. 2002; 62:65–74. 10.1124/mol.62.1.65. [DOI] [PubMed] [Google Scholar]
  • 31.Pnueli L, Shalev D, Refael Tet al.. Proliferating primary pituitary cells as a model for studying regulation of gonadotrope chromatin and gene expression. Mol Cell Endocrinol. 2021; 533:111349. 10.1016/j.mce.2021.111349. [DOI] [PubMed] [Google Scholar]
  • 32.Guo T, Marmol P, Moliner Aet al.. Adipocyte ALK7 links nutrient overload to catecholamine resistance in obesity. eLife. 2014; 3:e03245. 10.7554/eLife.03245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Pnueli L, Rudnizky S, Yosefzon Yet al.. RNA transcribed from a distal enhancer is required for activating the chromatin at the promoter of the gonadotropin α-subunit gene. Proc Natl Acad Sci USA. 2015; 112:4369–74. 10.1073/pnas.1414841112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Luo Z, Wijeweera A, Oh Yet al.. Pin1 Facilitates the phosphorylation-dependent ubiquitination of SF-1 to regulate gonadotropin β-subunit gene transcription. Mol Cell Biol. 2010; 30:745–63. 10.1128/mcb.00807-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Wijeweera A, Haj M, Feldman Aet al.. Gonadotropin gene transcription is activated by menin-mediated effects on the chromatin. Biochim Biophys Acta - Gene Regul Mech. 2015; 1849:328–41. 10.1016/j.bbagrm.2015.01.001. [DOI] [PubMed] [Google Scholar]
  • 36.Shi N, Guo X, Chen S-Y. Olfactomedin 2, a novel regulator for transforming growth factor-β-induced smooth muscle differentiation of human embryonic stem cell-derived mesenchymal cells. Mol. Biol. Cell. 2014; 25:4106–14. 10.1091/mbc.E14-08-1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Lafontaine L, Chaudhry P, Lafleur M-Jet al.. Transforming growth factor beta regulates proliferation and invasion of rat placental cell Lines1. Biol Reprod. 2011; 84:553–9. 10.1095/biolreprod.110.086348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ramachandran A, Mehić M, Wasim Let al.. Pathogenic ACVR1R206H activation by activin A-induced receptor clustering and autophosphorylation. EMBO J. 2021; 40:e106317. 10.15252/embj.2020106317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Michael IP, Saghafinia S, Tichet Met al.. ALK7 Signaling manifests a homeostatic tissue barrier that is abrogated during tumorigenesis and metastasis. Dev. Cell. 2019; 49:409–24. 10.1016/j.devcel.2019.04.015. [DOI] [PubMed] [Google Scholar]
  • 40.Rauluseviciute I, Riudavets-Puig R, Blanc-Mathieu Ret al.. JASPAR 2024: 20th anniversary of the open-access database of transcription factor binding profiles. Nucleic Acids Res. 2023; 52:D174. 10.1093/nar/gkad1059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Fornes O, Castro-Mondragon JA, Khan Aet al.. JASPAR 2020: update of the open-access database of transcription factor binding profiles. Nucleic Acids Res. 2020; 48:D87–92. 10.1093/nar/gkz1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Bacon CW, Challa A, Hyder Uet al.. KAP1 is a chromatin reader that couples steps of RNA polymerase II transcription to sustain oncogenic programs. Mol Cell. 2020; 78:1133–51. 10.1016/j.molcel.2020.04.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Oki S, Ohta T, Shioi Get al.. ChIP-Atlas: a data-mining suite powered by full integration of public ChIP-seq data. EMBO Rep. 2018; 19:e46255. 10.15252/embr.201846255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Reynolds N, Salmon-Divon M, Dvinge Het al.. NuRD-mediated deacetylation of H3K27 facilitates recruitment of polycomb repressive complex 2 to direct gene repression. EMBO J. 2011; 31:593. 10.1038/emboj.2011.431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Randolph K, Hyder U, D’Orso I. KAP1/TRIM28: transcriptional activator and/or repressor of viral and cellular programs?. Front Cell Infect Microbiol. 2022; 12:834636. 10.3389/fcimb.2022.834636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Cheng C-T, Kuo C-Y, Ann DK. KAPtain in charge of multiple missions: emerging roles of KAP1. World J Biol Chem. 2014; 5:308. 10.4331/wjbc.v5.i3.308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Martinez-Hackert E, Sundan A, Holien T. Receptor binding competition: a paradigm for regulating TGF-β family action. Cytokine Growth Factor Rev. 2021; 57:39–54. 10.1016/j.cytogfr.2020.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Tsuchida K, Nakatani M, Yamakawa Net al.. Activin isoforms signal through type I receptor serine/threonine kinase ALK7. Mol Cell Endocrinol. 2004; 220:59–65. 10.1016/j.mce.2004.03.009. [DOI] [PubMed] [Google Scholar]
  • 49.Andersson O, Bertolino P, Ibáñez CF. Distinct and cooperative roles of mammalian Vg1 homologs GDF1 and GDF3 during early embryonic development. Dev Biol. 2007; 311:500–11. 10.1016/j.ydbio.2007.08.060. [DOI] [PubMed] [Google Scholar]
  • 50.Andersson O, Reissmann E, Ibáñez CF. Growth differentiation factor 11 signals through the transforming growth factor-beta receptor ALK5 to regionalize the anterior-posterior axis. EMBO Rep. 2006; 7:831–7. 10.1038/sj.embor.7400752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Andersson O, Reissmann E, Jörnvall Het al.. Synergistic interaction between Gdf1 and Nodal during anterior axis development. Dev Biol. 2006; 293:370–81. 10.1016/j.ydbio.2006.02.002. [DOI] [PubMed] [Google Scholar]
  • 52.Andersson O, Korach-Andre M, Reissmann Eet al.. Growth/differentiation factor 3 signals through ALK7 and regulates accumulation of adipose tissue and diet-induced obesity. Proc Natl Acad Sci USA. 2008; 105:7252–6. 10.1073/pnas.0800272105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Bertolino P, Holmberg R, Reissmann Eet al.. Activin B receptor ALK7 is a negative regulator of pancreatic β-cell function. Proc Natl Acad Sci USA. 2008; 105:7246–51. 10.1073/pnas.0801285105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Reissmann E, Jörnvall H, Blokzijl Aet al.. The orphan receptor ALK7 and the activin receptor ALK4 mediate signaling by Nodal proteins during vertebrate development. Genes Dev. 2001; 15:2010–22. 10.1101/gad.201801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Namwanje M, Brown CW. Activins and inhibins: roles in development. Physiol Dis Cold Spring Harb Perspect Biol. 2016; 8:a021881. 10.1101/cshperspect.a021881. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Kitisin K, Saha T, Blake Tet al.. TGF-β signaling in development. Sci STKE. 2007; 2007:cm1. 10.1126/stke.3992007cm1. [DOI] [PubMed] [Google Scholar]
  • 57.Wijayarathna R, De Kretser DM. Activins in reproductive biology and beyond. Hum Reprod Update. 2016; 22:342–57. 10.1093/humupd/dmv058. [DOI] [PubMed] [Google Scholar]
  • 58.Bertero A, Brown S, Madrigal Pet al.. The SMAD2/3 interactome reveals that tgfβ controls m 6 A mRNA methylation in pluripotency. Nature. 2018; 555:256–9. 10.1038/nature25784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Cabezas F, Farfán P, Marzolo M-P. Participation of the SMAD2/3 signalling pathway in the down regulation of megalin/LRP2 by transforming growth factor beta (TGF-ß1). PLoS One. 2019; 14:e0213127. 10.1371/journal.pone.0213127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Chen C-R, Kang Y, Massagué J. Defective repression of c-myc in breast cancer cells: a loss at the core of the transforming growth factor β growth arrest program. Proc Natl Acad Sci USA. 2001; 98:992–9. 10.1073/pnas.98.3.992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Xi Q, Wang Z, Zaromytidou A-Iet al.. A poised chromatin platform for TGF-β access to master regulators. Cell. 2011; 147:1511–24. 10.1016/j.cell.2011.11.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Massagué J, Seoane J, Wotton D. Smad transcription factors. Genes Dev. 2005; 19:2783–810. 10.1101/gad.1350705. [DOI] [PubMed] [Google Scholar]
  • 63.Duttke SH, Guzman C, Chang Met al.. Position-dependent function of human sequence-specific transcription factors. Nature. 2024; 631:891–8. 10.1038/s41586-024-07662-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Xu W, Angelis K, Danielpour Det al.. Ski acts as a co-repressor with Smad2 and Smad3 to regulate the response to type β transforming growth factor. Proc Natl Acad Sci USA. 2000; 97:5924–9. 10.1073/pnas.090097797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Deheuninck J, Luo K. Ski and SnoN, potent negative regulators of TGF-beta signaling. Cell Res. 2009; 19:47–57. 10.1038/cr.2008.324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Tecalco-Cruz AC, Ríos-López DG, Vázquez-Victorio Get al.. Transcriptional cofactors Ski and SnoN are major regulators of the TGF-β/smad signaling pathway in health and disease. Signal Transduct Target Ther. 2018; 3:15. 10.1038/s41392-018-0015-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Markouli M, Strepkos D, Piperi C. Structure, activity and function of the SETDB1 protein methyltransferase. Life. 2021; 11:817. 10.3390/life11080817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Luo H, Wu X, Zhu X-Het al.. The functions of SET domain bifurcated histone lysine methyltransferase 1 (SETDB1) in biological process and disease. Epigenetics Chromatin. 2023; 16:47. 10.1186/s13072-023-00519-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Zhang Y, Remillard D, Onubogu Uet al.. Collateral lethality between HDAC1 and HDAC2 exploits cancer-specific NuRD complex vulnerabilities. Nat Struct Mol Biol. 2023; 30:1160–71. 10.1038/s41594-023-01041-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Hyder U, Challa A, Thornton Met al.. KAP1 negatively regulates RNA polymerase II elongation kinetics to activate signal-induced transcription. Nat Commun. 2024; 15:5859. 10.1038/s41467-024-49905-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.McNamara RP, Reeder JE, McMillan EAet al.. KAP1 recruitment of the 7SK snRNP complex to promoters enables transcription elongation by RNA polymerase II. Mol Cell. 2016; 61:39–53. 10.1016/j.molcel.2015.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ugaf008_Supplemental_File
ugaf008_Supplemental_File.pdf (545.3KB, pdf)

Data Availability Statement

Plasmids generated in this study will be distributed subject to restrictions of the original sources of their components. Datasets used are publicly available as cited in the manuscript. The code has been uploaded and is available https://doi.org/10.6084/m9.figshare.28623272.v2.

Articles from NAR Molecular Medicine are provided here courtesy of Oxford University Press

RESOURCES