Central precocious puberty (CPP) is a rare multifactorial disorder caused by an age-inappropriate secretion of the gonadotropin-releasing hormone from the hypothalamus1. CPP can occur as a result of monogenic mutations, although it is frequently associated with brain lesions, such as tumor and injury1. Thus far, a few genes, including KISS1R, KISS1, PROKR2, and NR0B1, have been reported as causative genes for CPP1,2,3. In addition, two imprinted genes, MKRN3 and DLK1, have recently been implicated in the development of CPP4,5. Mutations in MKRN3 and DLK1 cause CPP when they reside on paternally derived alleles. The association between epigenetic defects of MKRN3 or DLK1 and CPP has yet to be determined.

Previous studies in Western countries have identified pathogenic MKRN3 mutations in 9–46% of familial cases and 3–20% of sporadic cases with CPP (Table 1), indicating that these mutations play an important role in the etiology of CPP. In contrast, Lee et al. identified pathogenic MKRN3 mutations only in one of 260 Korean patients with CPP6. Likewise, Jeong et al. reported the lack of pathogenic MKRN3 mutations in 26 Korean patients with familial CPP7. These data indicate that there may be an ethnic difference in the frequency of MKRN3 mutations in CPP patients. However, MKRN3 mutation analyses have rarely been performed in Asian countries other than Korea, except for our previous study in which MKRN3 mutations were identified in one of 15 Japanese patients8. Moreover, since Lee et al.6 and Jeong et al.7 did not examine DNA methylation defects or copy-number alterations of MKRN3, these abnormalities may be hidden in their patients.

Table 1 Previous reports of large-scale MKRN3 mutation screening on patients with central precocious puberty
Full size table

Here we searched for genetic and epigenetic defects of MKRN3 in Japanese and Chinese patients with etiology-unknown CPP. Nucleotide substitutions were analyzed in 24 (22 Japanese and 2 Chinese) patients, whereas DNA methylation defects and copy-number alterations were examined only in 19 Japanese patients for whom we could obtain a sufficient amount of genomic DNA. This study was approved by the Institutional Review Board Committee of the National Research Institute for Child Health and Development and performed after obtaining written informed consent. All patients satisfied the following criteria: (i) early pubertal onset (in boys, testicular enlargement before 9 years of age, pubic hair before 10 years of age, or axillary hair/voice change before 11 years of age; in girls, breast budding before 7.5 years of age, pubic hair before 8 years of age, or menarche before 10.5 years of age); (ii) increased blood levels of gonadotropin and sex hormone; (iii) normal findings in brain magnetic resonance imaging; and (iv) no pathogenic mutations in DLK1, KISS1R, KISS1, PROKR2, or NR0B1. Patients with congenital malformation syndromes and those with chronic disorders that may affect hormone secretion were excluded from this study. Patients had no apparent family history of early puberty.

First, we searched for MKRN3 sequence variations in 24 patients. Thirteen patients were previously subjected to whole-exome sequencing using a Nextera Rapid Capture Exome Kit (HiSeq SBS Kit v4-HS Illumina, San Diego, CA, USA) and a HiSeq 2500 sequencer (Illumina)8. The remaining 11 patients were first examined in the present study; targeted sequencing of their DNA was performed for 148 genes using the HaloPlex HS Target Enrichment System (Design ID 40350-1451214604; Agilent Technologies, Palo Alto, CA, USA) and a MiSeq sequencer (Illumina). Sequence data of the 24 patients were analyzed as described previously3. We focused on nonsynonymous variants in the coding region and intronic substitutions affecting splice sites of MKRN3. Variants whose frequency in the Japanese general population [the ExAC browser (http://exac.broadinstitute.org/) and the Human Genetic Variation Browser (http://www.hgvd.genome.med.kyoto-u.ac.jp/)] is more than 1% were excluded as polymorphisms.

Next, we conducted DNA methylation and copy-number analyses for 19 patients. The DNA methylation status of seven CpG sites at the MKRN3 locus was examined by pyrosequencing using a previously described method9. Copy-number alterations of MKRN3 were analyzed by real-time PCR using a TaqMan Copy Number Assay Kit (MKRN3, Hs02079798; internal control, 440332; ThermoFisher Scientific, Tokyo, Japan) according to the manufacturer’s instructions.

Consequently, rare nucleotide substitutions of MKRN3 were not detected in the 24 patients except for one female patient with c.684dupA (p.Glu229Argfs*3), who has been reported previously8. Moreover, DNA methylation statuses were comparable between the patients and control individuals (Fig. 1). Similarly, copy-number analysis identified no deletions or duplications of MKRN3 in all patients examined.

Fig. 1: DNA methylation analysis of 19 patients.
figure 1

a Genomic structure of MKRN3 and flanking CpG sites. C1–C7 represent cytosines at the CpG sites in the differentially methylated region. F and R primers were used for PCR amplification, and the Seq primer was used for pyrosequencing. This figure is not drawn to scale. b Results of methylation analysis. Black dots represent the methylation status of the patients. Gray shaded areas indicate reference ranges obtained from 49 control individuals

Full size image

The results of this study expand the prior notion of Lee et al.6 and Jeong et al.7 to suggest that genetic and epigenetic defects in MKRN3 are relatively rare in Asian CPP patients. Although underlying factors of the ethnic difference in the frequency of MKRN3 mutations remain to be determined, the relatively high frequency in Western countries possibly reflects the presence of multiple founder mutations. Indeed, c.482delC (p.Pro161Argfs*10), c.482dupC (p.Ala162Glyfs*15), c.802_803delAT (p.Met268Valfs*23), c.982C > T (p.Arg328Cys), and c.1095G > T (p.Arg365Ser) have been repeatedly identified in patients from these countries (Table 1). In this regard, since there have been no reports of de novo MKRN3 mutations in CPP patients, the de novo occurrence of MKRN3 substitutions seems to be an exceptional event. Moreover, our data suggest that DNA methylation defects and copy-number alterations of MKRN3 play only a minor role in the development of CPP, if at all.

In conclusion, the results of this study, together with two reports from Korea6,7, indicate that (epi)genetic defects of MKRN3 represent only a minor cause of CPP in Asian populations. The presence of multiple founder mutations in Western countries as well as the rarity of de novo occurrences of intragenic nucleotide substitutions likely underlies the ethnic differences in the frequency of MKRN3 mutations in CPP cases.