Novel MKRN3 Missense Mutations Associated With Central Precocious Puberty Reveal Distinct Effects on Ubiquitination

Abstract

Context

Loss-of-function mutations in the maternally imprinted genes, MKRN3 and DLK1, are associated with central precocious puberty (CPP). Mutations in MKRN3 are the most common known genetic etiology of CPP.

Objective

This work aimed to screen patients with CPP for MKRN3 and DLK1 mutations and analyze the effects of identified mutations on protein function in vitro.

Methods

Participants included 84 unrelated children with CPP (79 girls, 5 boys) and, when available, their first-degree relatives. Five academic medical institutions participated. Sanger sequencing of MKRN3 and DLK1 5′ upstream flanking and coding regions was performed on DNA extracted from peripheral blood leukocytes. Western blot analysis was performed to assess protein ubiquitination profiles.

Results

Eight heterozygous MKRN3 mutations were identified in 9 unrelated girls with CPP. Five are novel missense mutations, 2 were previously identified in patients with CPP, and 1 is a frameshift variant not previously associated with CPP. No pathogenic variants were identified in DLK1. Girls with MKRN3 mutations had an earlier age of initial pubertal signs and higher basal serum luteinizing hormone and follicle-stimulating hormone compared to girls with CPP without MRKN3 mutations. Western blot analysis revealed that compared to wild-type MKRN3, mutations within the RING finger domain reduced ubiquitination whereas the mutations outside this domain increased ubiquitination.

Conclusion

MKRN3 mutations were present in 10.7% of our CPP cohort, consistent with previous studies. The novel identified mutations in different domains of MKRN3 revealed different patterns of ubiquitination, suggesting distinct molecular mechanisms by which the loss of MRKN3 results in early pubertal onset.

Puberty is a complex biological process characterized by accelerated linear growth velocity, sexual development, and acquisition of fertility (1). Pubertal onset and maintenance of reproductive function are regulated by the hypothalamic-pituitary-gonadal (HPG) axis (2, 3). The HPG axis is active during fetal development and becomes quiescent toward the end of gestation. Several weeks after birth the HPG axis is reactivated for a short time, which has been labeled mini-puberty (4). The axis is then again suppressed during childhood until reactivation occurs at the onset of puberty (1, 5). The onset of puberty is marked by increased amplitude and frequency of gonadotropin-releasing hormone (GnRH) pulses leading to an increase in the secretion of the pituitary gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), with consequent activation of gonadal function (1, 6). The mechanisms regulating both HPG axis suppression and reactivation are not fully understood.

Central precocious puberty (CPP) results from early reactivation of the HPG axis. Clinically, CPP is marked by premature and progressive sexual development associated with increased linear growth velocity and accelerated bone age maturation (7). The definitive age of early puberty varies across ethnic and socioeconomic backgrounds; for instance, African American girls may have earlier pubertal development compared to White girls (1). Generally, however, precocious puberty is defined as the initiation of progressive pubertal changes before age 8 years in girls and age 9 years in boys (8–10). The incidence of CPP is estimated to be 1 in 5000 to 10 000 children, with a higher frequency in girls than in boys (11–14).

In 2013, our group identified loss-of-function mutations in MKRN3 (Makorin ring finger protein 3) in familial cases of CPP (15). MKRN3 is a paternally expressed and maternally imprinted, intronless gene located in the Prader-Willi syndrome (PWS) critical region (Chr15q11-q13) (15–18). Importantly, patients with CPP and mutations in MKRN3 do not manifest the classic features of PWS (19). After our first report linking MKRN3 to the reproductive axis, other groups have identified MKRN3 mutations associated with sporadic and familial CPP (15, 20–24). To date, MKRN3 mutations are the most common genetic defect associated with CPP. The overall frequency of MKRN3 mutations among patients lacking an identifiable organic etiology for CPP is 10%. Among familial cases, the frequency of MKRN3 mutations is increased to 33% to 46% (20, 25, 26). Thus far, 65 inactivating mutations located in the coding sequence of MKRN3 have been reported (20, 21, 23, 27–33).

We have shown that MKRN3 acts in the hypothalamic arcuate nucleus, specifically in KNDy (kisspeptin, neurokinin B, and dynorphin expressing) neurons to regulate GnRH secretion and puberty initiation (34). Based on available human data (genetic studies and decreased circulating MKRN3 levels with onset of puberty), in vitro studies, and animal studies, MKRN3 is presumed to act as a brake on GnRH secretion during childhood. Loss-of-function MKRN3 mutations interfere with this braking activity, leading to premature reactivation of the hypothalamic GnRH pulse generator and HPG axis precipitating CPP (34, 35).

Potential mechanisms of MKRN3 action include ubiquitination of other proteins, autoubiquitination, direct actions affecting GnRH1 expression, and interactions with other proteins (36). MKRN3 encodes an E3 ubiquitin ligase and contains 3 zinc finger C3H1 motifs, a Makorin-type Cys-His zinc finger motif, and a RING finger C3HC4 motif (Fig. 1) (37). Missense mutations reported in patients with CPP tend to cluster at the RING finger C3HC4 domain, implicating a pivotal role of MKRN3 ubiquitin ligase activity in the pathophysiology of CPP (21). The enzymatic activity of RING E3 ubiquitin ligases can be assessed by measuring their autoubiquitination in vitro (38). Since our first report linking MKRN3 to pubertal onset, studies by our group and others have shown that missense mutations identified in patients with CPP located in different domains of MKRN3 differentially alter its ubiquitination activity (34, 39, 40). Hence, available data implicate several mechanisms of action for MKRN3.

In 2017, whole-genome sequencing of DNA samples obtained from 4 girls with CPP in 1 family revealed a complex genomic defect in DLK1 (delta-like noncanonical notch ligand 1) (41). Similar to MKRN3, DLK1 is a maternally imprinted, paternally expressed gene that was mapped to a region of clustered imprinted genes located on Chr14q32.2. Maternal uniparental disomy, paternal deletion, or epimutations of this region of Chr14q32 are associated with Temple syndrome, a neurodevelopmental disorder (42). CPP occurs in up to 76% of cases of Temple syndrome (43, 44). After our first report linking DLK1 to familial CPP, 3 frameshift DLK1 mutations were identified in 5 women from 3 unrelated families with members affected with CPP (45). Although the number of patients with DLK1 mutations is small, metabolic abnormalities were more prevalent in women with CPP and DLK1 mutations than in women with idiopathic CPP (41, 45, 46). Curiously, in addition to MKRN3 and DLK1 mutations being associated with CPP, genome-wide association studies have identified variants associated with age at menarche in loci close to MKRN3 and DLK1. These findings highlight the role of these genes in pubertal onset and suggest that parent-of-origin differential gene expression contributes to regulating the time of pubertal initiation (47, 48).

In this study, we screened our cohort of 84 unrelated patients with CPP for mutations in these 2 imprinted genes, MKRN3 and DLK1. Our goal was to identify novel variants and to further investigate the molecular mechanism(s) through which these genes influence the onset of GnRH-dependent puberty.

Materials and Methods

Recruitment and Clinical Phenotyping

In this study, 84 unrelated patients (79 female and 5 male) with CPP were screened for mutations in MKRN3 and DLK1. Nineteen patients (22%) reported a known family history of precocious puberty. Fifty-seven percent of the cohort identified as White, 34% identified as African American, and the remainder identified as Hispanic, mixed ethnicity, or the information was unavailable. CPP was diagnosed based on clinical signs of progressive pubertal development (Tanner stage ≥ II of breast development in girls and testicular volume ≥ 4 mL in boys) before age 8 years in girls and 9 years in boys, pubertal basal and/or GnRH analogue (leuprolide)-stimulated LH levels, and advanced bone age by at least more than 2 SDs between bone and chronological age (Greulich and Pyle method). All patients had normal results on magnetic resonance imaging of the central nervous system, excluding organic causes of CPP. Patient clinical data, family history, and blood samples were collected by the referring physicians. When available, family members of patients with CPP gave consent and provided their pubertal development history and blood samples. Familial CPP was defined as having at least 2 family members affected (49).

Consent to participate was obtained from the parents and overseen by the institutional review board at Nemours Children's Health or Brigham and Women's Hospital. Clinical and hormonal data were compared between female patients with MKRN3 mutations and those with CPP without MKRN3. Using the nonparametric Mann-Whitney U test, statistical significance was set at P less than .05. For statistical analysis, patients and siblings with MKRN3 mutations were compared to patients without identified mutations.

Hormone Assays

Serum levels of LH, FSH, testosterone, and estradiol (E2) were determined using immunochemiluminometric assays. The interassay coefficient of variation was less than or equal to 5% for all assays. GnRH analogue was used at a dose of 50 mcg/kg of aqueous leuprolide (5 mg/mL), maximum dose of 500 mcg during the stimulation test; serum levels of LH and FSH were measured at the time of GnRH analogue administration and at 60 and 180 minutes after administration. In both sexes, basal LH levels higher than 0.15 U/L were considered to be pubertal levels, and peak GnRH-stimulated LH levels higher than 5.0 U/L were considered to be pubertal responses. Basal E2 levels greater than 15 pg/mL and basal testosterone levels greater than 12 ng/dL were considered to be pubertal levels.

Polymerase Chain Reaction Amplification and Sequencing of MKRN3 and DLK1

Genomic DNA was isolated from peripheral blood leukocytes from all patients and, when available, from their family members, using standard procedures. The promoter and coding regions of MKRN3 (ENST00000314520.6) and DLK1 (ENST00000341267.9) were amplified by polymerase chain reaction and sequenced via Sanger method at DF/HCC DNA Resource Core at Harvard Medical School and the CCIB DNA Core facility at Massachusetts General Hospital. Reference sequence assembly GRCh38/hg38 was used. Primer sequences for MKRN3 and DLK1 are detailed in Supplementary Table S1 (50).

Serum DLK1 Measurements

Serum DLK1 levels were measured in patients with CPP and MKRN3 mutations, the patient with the variant in DLK1 promoter region, and patients with CPP without identified variants using available serum and a soluble DLK1 enzyme-linked immunosorbent assay (Tecan [IBL] catalog No. IB99504, RRID:AB_2934107) per the manufacturer's instructions. This enzyme-linked immunosorbent assay (ELISA) is described to be specific for the measurement of DLK1 and recovers an average of 92% of DLK1 compared to serum samples spiked with known concentrations of human DLK1. DLK1 levels range from 0.4 to more than 2.5 ng/mL in healthy control human serum (51, 52). The lower limit of detection for the assay is approximately 0.4 ng/mL, with a mean intra-assay variability of 5% and mean interassay variability of 8.1%. In 2 different studies using this ELISA kit, the mean DLK1 serum levels were found to be 0.69 ± 0.036 and 1.39 ± 1.37 ng/mL in 114 and 30 healthy controls, respectively (51, 52).

Site-directed Mutagenesis and Cell Culture

MKRN3 mutations identified in our cohort were introduced into 3XHA-hMKRN3, obtained from GeneCopeia (EX-F0470-M06), and expressed the full-length human MKRN3 coding sequence. Missense and frameshift mutations were introduced using the QuikChange II XL Site-Directed Mutagenesis Kit (Agilent Technologies) following the manufacturer's instructions and using primers listed in Supplementary Table S2 (53). Sanger sequencing was performed to confirm that the mutations were inserted appropriately.

HEK293 cells (ATCC) were cultured in Dulbecco's Modified Eagle Medium/High Glucose (DMEM; Sigma Aldrich) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 1% penicillin/streptomycin solution (Pen/Strep; Invitrogen). All cells were incubated in a humidified incubator at 37 °C and 5% CO₂. Cells were tested for mycoplasma contamination (Universal Mycoplasma Detection kit, ATCC) monthly.

Transfection and Protein Extraction

Cells were plated at a density of 0.3 × 10⁶ cells/well in a 6-well plate 24 hours before transfection. Wild-type (WT) or mutant 3XHA-hMKRN3 vectors (1 µg/well) were transfected using Lipofectamine 2000 (Invitrogen).

Twenty-four hours post transfection, cells were harvested by lysis in Pierce IP Lysis Buffer (Thermo) supplemented with Protease Inhibitor (Roche Diagnostics Ltd) for 20 minutes on ice followed by centrifugation for 20 minutes at 13 000 rpm at 4 °C. The protein concentration of the cell lysate was measured by the Bradford Protein Assay (Thermo Scientific) according to the manufacturer's instructions.

Coimmunoprecipitation and Western Blot

Approximately 10% of protein was retained as input. The remaining protein lysates were incubated with rabbit anti-HA antibody (Abcam catalog ab9110, RRID:AB_307019) (1:1000) overnight at 4 °C on a rotating wheel for coimmunoprecipitation studies. A total of 30 µL of Protein A Agarose Beads (ab193255, Novex) were then added to the protein:antibody mix and incubated for another 3 hours at room temperature. The samples were spun down, and the immunoprecipitated complexes were washed at least 3 times with TRIS-buffer saline containing Tween-20 0.1% (TBST) supplemented with protease inhibitor. Immunoprecipitants were then eluted with Laemmli buffer and boiled at 95 °C for 5 minutes. Input and immunoprecipitants were analyzed by Western blot. Samples were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) (10% polyacrylamide) and transferred onto nitrocellulose (Promega) membranes. After blocking with 5% nonfat milk in TBST for 1 hour at room temperature, membranes were incubated overnight at 4 °C with primary antibodies (1:1000): rabbit anti-HA (Abcam catalog No. ab9110, RRID:AB_307019) and mouse antiubiquitin (BioLegend catalog No. 646301, RRID:AB_1659270). After washing in TBST, the membranes were probed for 1 hour with peroxidase-conjugated anti-rabbit (Jackson ImmunoResearch Labs catalog No. 211-032-171, RRID:AB_2339149) or -mouse (Jackson ImmunoResearch Labs catalog No. 115-035-146, RRID:AB_2307392) secondary antibodies (1:10 000). Bands were detected by using an electrochemiluminescence detection kit (E2400, Denville Scientific) and membranes imaged using a Biorad ChemiDoc XRS⁺ Imaging System. Membranes were then stripped with stripping buffer (Restore, Thermo Fisher) and reprobed with monoclonal horseradish peroxidase–conjugated β-actin (Sigma-Aldrich catalog No. A3854, RRID:AB_262011) (1:10 000) for input loading control.

Results

MKRN3 Mutation Screening

We identified 8 heterozygous MKRN3 mutations in 9 unrelated girls in our cohort of 79 girls with CPP (Fig. 1; Table 1). No MKRN3 mutations were identified in the 5 boys in our cohort. Of the 19 cases with a clinically reported family history of CPP, 3 had mutations in MKRN3 (15.7%). Among the identified mutations, 5 were novel missense mutations (p.Tyr117Cys, p.Met126Val, p.Arg345Cys, p.Cys364Phe, p.Ile461Phe). One missense mutation had been identified previously in association with CPP (p.Arg345His) (54) and 2 were frameshift mutations (p.Ala162Glyfs*14, p.Ala288Profs*108). The frameshift mutation, p.Ala162Glyfs*14, is the most commonly identified MKRN3 mutation associated with CPP and was detected in 2 unrelated patients. The other frameshift variant, p.Ala288Profs*108, is very rare variant (rs751716250: GnomAD frequency: 8e-06), not previously identified with CPP, and was identified in one girl with CPP and her affected sister (Fig. 2) (20, 31, 32). The p.Arg345Cys mutation was identified in a proband with CPP, her 3 siblings, and her father; DNA was not available from her affected aunt (family 6, Fig. 2). The p.Cys364Phe mutation was identified in a proband and her affected paternal cousin, indicating that the mutation was inherited from her father (family 8, Fig. 2).

Familial segregation analysis, when DNA was available, revealed that the mutation was always inherited from the father (families 2, 3, 5, 6, 8, and 9, Fig. 2). In other words, no de novo mutations were identified in these families, as shown in Fig. 2. DNA from family of the probands in families 4 (p.Ala162Glyfs*14) and 7 (p.Arg345His) and from the father of the proband in family 1 (p.Tyr117Cys) were not available for genetic studies.

Two novel missense mutations, p.Arg345Cys and p.Cys364Phe, and a previously identified missense variant, p.Arg345His, were in the RING finger domain. One novel missense mutation, p.Tyr117Cys, was located in the first zinc finger C3H4 domain. Two novel mutations, p.Met126Val and p.Ile461Phe, were located outside the key functional domains. Three prediction software programs, Polyphen-2, Mutation Taster, and Panther, predicted these missense variants in the finger domains as being deleterious and disease causing. Only p.Met126Val and p.Ile461Phe were predicted to be benign variants (Table 2) (55–57).

Central Precocious Puberty: Clinical Observations

The detailed clinical and hormonal data for the 79 girls with CPP are presented in Tables 3 and 4. We restricted our analyses comparing clinical and biochemical data to girls because no pathogenic MKRN3 variants were identified among the boys in our cohort. The mean Tanner stage for girls with MKRN3 mutations was 4, and for girls without MKRN3 mutations was 3. Compared to girls with idiopathic CPP (ie, without identified organic causes and without identified MKRN3 mutations), girls with MKRN3 mutations had an earlier age of initial pubertal signs (6.13 ± 0.24 years vs 6.83 ± 0.11 years; P = .0052), higher basal LH (2.36 ± 0.73 IU/L vs 1.22 ± 0.27 IU/L; P = .0174), and higher basal FSH (4.74 ± 0.56 IU/L vs 3.23 ± 0.27 IU/L; P = .0084) (Fig. 3A-3C).

Patients with MKRN3 missense and frameshift mutations were compared to those without MKRN3 mutations in Supplementary Fig. S1 (58). Girls harboring MKRN3 missense mutations had significantly earlier presentation with initial pubertal signs and also had significantly higher basal LH levels than those without MKRN3 mutations. Girls harboring MKRN3 frameshift mutations had significantly higher FSH levels compared to girls without mutations. (Supplementary Fig. S1C) (58).

DLK1 Screening

We sequenced the promoter and coding region of the DLK1 gene in 79 girls of our cohort; sufficient DNA was not available to sequence the entire region in 5 girls. A rare homozygous G > T missense variant in the DLK1 promoter region, c.-73-144G > T (rs547367494), was identified in a female patient without a family history of CPP. The highest population minor allele frequency was less than 0.01 in 1000Genomes. The girl with DLK1 variant c.-73-144G > T presented with thelarche, pubarche, and elevated basal LH levels (0.2 IU/L) at age 7.25 years, with a bone age of 10.5 years. She did not report any family history of precocious puberty.

DLK1 Enzyme-linked Immunosorbent Assay

To interrogate the consequences of the DLK1 variant, c.-73-144G > T (rs547367494), on protein levels, we measured circulating DLK1 levels in the serum of the girl harboring this variant; her DLK1 levels were not reduced, suggesting that this specific variant does not alter DLK1 expression and is unlikely to be associated with CPP (Fig. 4).

A link between the 2 imprinted regions on chromosome 14 and 15 was demonstrated with the use of induced pluripotent stem cells with genetic alterations in the PWS critical region, showing that the long noncoding RNA located in chromosome 15, IPW, is a regulator of the DLK1 locus on chromosome 14 (59). To investigate if the loss of MKRN3 function alters DLK1 levels, we measured serum DLK1 levels in patients with CPP with and those without MKRN3 variants for whom serum was available. No significant differences in serum DLK1 levels were observed (Fig. 4).

Ubiquitination of Novel MKRN3 Missense Mutations

Western blot analysis was performed to identify the effect of the identified variants on MKRN3 ubiquitination patterns. WT HA-MKRN3 showed 3 specific bands, at approximately 70, approximately 65, and approximately 50 kDa, by Western blot analysis (Fig. 5A). As the molecular weight of MKRN3 is predicted to be 55 kDa (https://www.genecards.org/cgi-bin/carddisp.pl?gene=MKRN3), we presume that higher-molecular-weight complexes likely reflect posttranslational modifications such as glycosylation or ubiquitination. The p.Ala162Glyfs*14 frameshift variant showed a predominant band at approximately 30 kDa, while the p.Ala288Profs*108 mutation had a predominant band at approximately 45 kDa, corresponding with the loss of 331 and 111 amino acids, respectively. The mutations located in the RING finger domain, p.Cys364Phe and p.Arg345Cys, showed reduced intensity of the approximately 65-kDa band relative to WT HA-MKRN3, and a more intense 70-kDa band (Fig. 5A).

The HA-MKRN3 proteins were immunoprecipitated using an anti-HA antibody, after which the immunoprecipitants were run on a Western blot and probed with an antiubiquitin antibody (Fig. 5B). The polyubiquitination smear of WT MKRN3 and the missense mutants studied was detected from approximately 90 kDa to approximately 260 kDa, whereas the truncated proteins encoded by p.Ala162Glyfs*14 and p.Ala288Profs*108 showed no detectable ubiquitination. The RING finger domain missense mutants, p.Cys364Phe and p.Arg345Cys, showed a reduced intensity of the polyubiquitination smear compared to WT MKRN3. In contrast, the p.Tyr117Cys, p.Met126Val, and p.Ile461Phe MKRN3 mutants showed a more intense ubiquitination smear compared to WT MKRN3.

Discussion

Since our first report linking MKRN3 to HPG axis function, groups from other countries have identified mutations in MKRN3 associated with CPP; the reported frequency approximates 10% of idiopathic cases (20, 31, 32). Familial cases appear to have a higher frequency of MKRN3 mutations, up to 33% to 46% (15, 24, 25, 31, 32). In this publication, we identified MKRN3 mutations in 9 girls, corresponding to 10.7% of patients in our entire cohort. Fifteen percent with a positive family history of CPP harbored MKRN3 mutations in our cohort.

When family segregation analyses were performed, MKRN3 mutations have always been found to be paternally inherited. No confirmed de novo mutations have been identified in MKRN3 to date. In this study, probands in families 2, 3, 5, 6, and 9 inherited the deleterious variant from their fathers, and the proband in family 8 is also presumed to have inherited the mutation from her father. Family members, especially fathers, may be undetected carriers especially due to inaccurate recollection regarding age at pubertal onset. Here, of the 5 fathers harboring a pathogenic MKRN3 variant, only 1 reported a history of CPP.

Consistent with the low frequency of CPP in boys, only 5 boys were included in our cohort. No MKRN3 mutations were found in these boys. In some studies, the frequency of MKRN3 mutations in the affected boys is high, and it is likely that the small number of affected boys in our cohort is the reason for the absence of MKRN3 mutations (15, 60).

We identified 6 missense mutations, of which 5 are novel mutations (p.Tyr117Cys, p.Met126Val, p.Arg345Cys, p.Cys364Phe, and p.Ile461Phe). The missense variant p.Arg345His, identified in a sporadic case in our cohort without DNA available from the father, has been associated with familial CPP previously, with the mutation confirmed to have been inherited from the father (54).

A rare frameshift variant, p.Ala288Profs*108 (rs751716250—GnomAD frequency 8e-06) was identified in a girl (proband of family 4), her affected sister with CPP, and their father. The hot spot mutation Ala162Glys*14 was identified in 2 patients in our cohort.

Unlike MKRN3, DLK1 is a secreted protein, which provides us with a tool to assess the effect of the variants on protein levels (61). All DLK1 mutations identified to date associated with CPP are also associated with decreased serum DLK1 protein levels (41, 45). We identified a rare variant in the promoter region of DLK1 (rs547367494) in a girl with sporadic CPP. Serum DLK1 levels were not decreased in this patient, indicating that the promoter region single-nucleotide variant likely did not affect protein levels. Based on these findings, we presume that this rare variant is not associated with CPP in this patient. Thus, DLK1 variants appear to be a rare cause of CPP (46).

Previous reports have shown that IPW, the long noncoding RNA located on chromosome 15 in proximity to MKRN3, regulates the expression of imprinted genes on chromosome 14 in the region where DLK1 is located (59). To ascertain whether loss of MRKN3 could affect serum DLK1 levels, we compared serum DLK1 levels between patients with and without MKRN3 variants. No differences were found in DLK1 levels, suggesting that MKRN3 does not influence DLK1 expression.

Girls with CPP associated with MKRN3 mutations presented with initial signs of pubertal onset at an earlier age than those with CPP without mutations in MKRN3 (see Fig. 3). Among larger CPP cohorts, Macedo et al (62) reported comparable ages at pubertal onset between those with MKRN3 variants and those without MKRN3 variants. However, Simon et al (25) found that the median age of puberty onset differed significantly between girls with CPP with and without MKRN3 mutations, similar to our findings. While a difference in family observation and detection of initial signs of puberty may be the reason for the discrepancies among the different studies, the higher number of missense mutations in our study compared to that by Macedo et al (62), where only one mutation identified was missense, may be a reason for this difference. As we can see in Supplementary Fig. S1A, only girls with CPP and MKRN3 missense mutations (not those with frameshift mutations) presented with earlier signs of puberty, compared to girls without mutations in MKRN3 (58). These data need to be interpreted carefully, as the analysis compares 2 distinctly differently sized populations, with only a small sample size of patients with MKRN3 mutations.

The various MKRN3 domains suggest that this protein may have several mechanisms of action. We have previously demonstrated that MKRN3 has ubiquitin ligase activity including autoubiquitination (34, 63). It has been reported that MKRN3 targets methyl-CpG-DNA binding protein 3 (MBD3) and poly(A)-binding proteins (PABPs) for ubiquitination, resulting in the suppression of GnRH production, suggesting that impairment of MKRN3 ubiquitination activity results in failure to suppress the reactivation of the HPG axis (39, 40).

We performed Western blot analyses to assess the effect of the mutations on MKRN3 protein expression and ubiquitination. As shown previously, the MKRN3 protein has 3 main bands likely due to posttranslational modifications (34). MKRN3 frameshift mutants, p.Ala162Glyfs*14 and p.Ala288Profs*108, had the expected reduced molecular sizes (34). We tested the ability of MKRN3 to bind to ubiquitin through immunoprecipitation studies. Ubiquitination was absent for the MKRN3 frameshift mutants, p.Ala162Glyfs*14 and p.Ala288Profs*108, as expected. The markedly reduced ubiquitination in MKRN3 carrying missense mutations in the RING finger domain (p.Arg345Cys and p.Cys364Phe MKRN3) compared to WT MKRN3 suggested decreased ubiquitination ability, consistent with previous findings and further demonstrating the importance of the RING finger domain for ubiquitination (34, 39). The increased intensity of the polyubiquitination pattern observed in immunoprecipitation studies of p.Tyr117Cys, p.Met126Val, and p.Ile461Phe MKRN3, mutations located outside the RING finger domain, suggests increased autoubiquitination and altered protein function. These findings suggest pathogenicity of these variants. Furthermore, these findings suggest that alterations across different MKRN3 domains differentially affect MKRN3 ubiquitination, ultimately influencing its protein structure and function.

Early puberty is associated with adverse long-term health outcomes, including increased risk of type 2 diabetes, heart disease, stroke, estrogen-dependent cancer, and cardiovascular mortality (14, 47). Consequences of untreated CPP include adult short stature, poor self-esteem, and psychological distress (64). This study shows that MKRN3 mutations should be considered in patients presenting with CPP, especially if there is a family history, not only in the father but also in siblings and paternal-related family members. The differences in the clinical and biochemical presentations of patients with CPP harboring missense or frameshift MKRN3 mutations, combined with distinct patterns of MKRN3 ubiquitination induced by mutations in different domains, suggest a sophisticated mechanism of action by which this protein regulates puberty initiation. Although loss-of-function mutations in MKRN3 are the most common genetic defects associated with CPP, the number of patients harboring these mutations is still small, given the rarity of this disorder. As a result, the relevance of the clinical and biochemical differences observed in this study need to be validated; continued detailed studies of patients with CPP will further advance the ability to correlate genetic and molecular findings with clinical phenotypes.

Funding

This work was supported by NIH R00 HD091381 to A.P.A, and NIH R01 HD019938 to A.P.A and U.B.K; NIH R01 HD082314, R21 HD098684, and the Brigham and Women's Hospital Women's Brain Initiative to U.B.K, and NIH K08 100595 to S.A.R.

Disclosures

NM had research grant support from Tolmar and present support from Abbvie. SFW is a site investigator for Neurocrine Biosciences and received research funding unrelated to this study.

Data Availability

Original data generated and analyzed during this study are included in this published article or in the data repositories listed in “References.”

References

Abbreviations

 
• CPP

  central precocious puberty

 
• DLK1

  delta-like noncanonical notch ligand 1

 
• E2

  estradiol

 
• ELISA

  enzyme-linked immunosorbent assay

 
• FSH

  follicle-stimulating hormone

 
• GnRH

  gonadotropin-releasing hormone

 
• HPG

  hypothalamic-pituitary-gonadal

 
• LH

  luteinizing hormone

 
• MKRN3

  Makorin ring finger protein 3

 
• PWS

  Prader-Willi syndrome

 
• WT

  wild-type

Author notes