Molecular screening of PROKR2 gene in girls with idiopathic central precocious puberty

Background Prokineticin receptor 2 (PROKR2) loss of function mutations have been described as cause of hypogonadotropic hypogonadism. In 2017, a first case of central precocious puberty (CPP) caused by PROKR2 heterozygous gain of function mutation was described in a 3.5 years-old girl. No other cases have been reported yet. This study performs a molecular screening in girls with early onset CPP (breast budding before 6 years of age) to identify possible alterations in PROKR2. Methods We analysed DNA of 31 girls with idiopathic CPP diagnosed via basal LH levels > 0.3 IU/L or peak-LH > 5 IU/L after stimulation, without any MKRN3 mutations. The Fisher exact test was used to compare polymorphism allele frequency to corresponding ones in genome aggregation database (gnomAD). Results No rare variants were identified. Five polymorphisms were found (rs6076809, rs8116897, rS3746684, rs3746682, rs3746683). All except one (i.e. rs3746682) had a minor allele frequency (MAF) similar to that reported in literature. rs3746682 presented a MAF higher than that described in the gnomAD (0.84 in our cohort vs 0.25 from gnomAD). Conclusions As for other G protein-coupled receptors (i.e. GPR54), mutations in PROKR2 do not seem to be a frequent cause of CPP in girls.


Background
Idiopathic central precocious puberty (CPP) results from premature activation of hypothalamic GnRH secretion in absence of congenital or acquired organic lesions in central nervous system.
The prevalence of CPP is higher in girls than in boys [1]. Although around 70% of variation of pubertal timing seems due to genes, genetic mechanisms leading to CPP are still a field of scientific investigation [2]. So far, MKRN3 loss-of-function mutations are the most frequently identified monogenic cause of CPP [3,4]. Heterozygous activating mutations in KISS1 and KISS1R genes have been reported as causes of few cases of CPP [5,6]. DLK1 loss-of-function mutations determine a more complex, yet uncommon, phenotype characterized by CPP, overweight, early onset Type 2 Diabetes, hyperlipidemia, and Polycystic Ovary Syndrome [7]. Besides that, rare cases of CPP patients primarily related to clinical syndromes or chromosomal abnormalities have been identified [8].
In 2017 Fukami et al. performed an extended molecular analysis by next generation sequencing and identified a new heterozygous frameshift mutation in PROKR2 gene in a 3.5 years-old girl affected by CPP [9]. PROKR2 is a G protein-coupled receptor (GPCR) expressed on the membrane of GnRH neurons, whose activation promotes GnRH secretion. Loss-of-function mutations in this gene accounts for about 5% congenital hypogonadotropic hypogonadism with or without anosmia [10]. In contrast with other mRNA frameshift mutations producing a stop-codon, the mRNA with the mutation identified by Fukami and colleagues, namely p.C242fsX305, showed to escape nonsense-mediated mRNA decay mechanism. In-vitro assay demonstrated that this variant receptor did not exert activity itself whereas boost the activity of wild-type receptor in heterozygous state. The precise mechanism of action still remains unclear. Perhaps, hyperactive mutant-wild-type heterodimers have a greater ligand affinity or a reduced receptor internalization due to the truncated C-terminal domain.
This study aimed to perform a molecular screening of PROKR2 variations in a cohort of girls with CPP. Considering the exceptionally young age of the patient described by Fukami et al., we limited the molecular screening to girls with CPP onset before the age of 6 years.
To date, no other data of mutations in PROKR2 in CPP are available.

Setting
In order to investigate the possible role of PROKR2 in the pathogenesis of CPP, a prospective, observational multicentre study was set up and carried out over a 2year period. Four Italian centres were involved in this study: Department of Woman, Child, General and Specialized Surgery of University of Campania "Luigi Vanvitelli", Naples, Pediatric Section-Department of Translational Medical Sciences, University of Naples Federico II, Naples, Institute for Maternal and Child Health IRCCS, "Burlo Garofolo", Trieste, and the Department of Paediatrics, Bologna University, Bologna. Ethics committee of the University of Campania "Luigi Vanvitelli" approved the protocol then subscribed by ethics committee of the other centres. According to the World Medical Association Declaration of Helsinki, written informed consent from parents and oral consent from all participants was collected.

Subjects
We recruited 31 females with diagnosis of CPP that met these following criteria: thelarche occurred before age of 6 years, defined as Tanner stage B ≥ 2 at physical examination, a diagnosis of central hypothalamic-pituitarygonadal activation identified by pubertal basal luteinizing hormone (LH) levels (> 0.3 IU/l) or a positive Gn-RH stimulation test (peak-LH > 5 IU/mL) and normal brain MRI.
Exclusion criteria included: congenital defects or abnormalities possibly related to syndromic features, central nervous system pathology (i.e., tumours or nonspecific cerebral anomalies associated with CPP). All the patients involved in this study were Italian. All subjects were unrelated.

Protocol
All girls underwent a clinical examination with special regard to auxological parameters and staging of breast development according to Tanner's classification. Right hand and wrist X-ray were performed for bone age evaluation by TW2 method. Peripheral blood samples were collected for hormonal dosage and genetic analysis.
Chemiluminescence assay (LIAISON, Diasorin) was used to measure Follicle Stimulating Hormone (FSH) and LH concentrations, with detection limits of 0.06 and 0.05 U/L, respectively, and intra-and inter-assay CV less than 5%. Radioimmunoassay was used to measure serum estradiol (CisBio International). The analytical and functional detection limits for plasma estradiol were 4 and 8 pg/mL, respectively.
GnRH stimulation test was provided for patients in which basal hormone level did not meet diagnostic criteria for CPP. Peak-LH > 5UI/L after administration of 0.1 mg of Relefact LH-RH (Sanofi-Aventis, Frankfurt am Main, Germany) was considered positive.
Thin-section, contrast-enhanced MRI examination of sellar region with T1-weighted and T2 weighted sagittal, coronal sequences, 3DT2 thin section axial sequence and FLAIR and EPI DWI on axial sequence was acquired for all patients.

Genetic analysis
Genomic DNA was extracted from peripheral whole blood using a DNA extraction kit (Promega, Madison WI, USA) following the manufacturer's instructions. Each of the two coding exons and the intron-exon boundaries of the PROKR2 gene was amplified by polymerase chain reaction (PCR) using two couples of primers each and subsequently analysed by direct sequencing (ABI PRISM 3100, Perkin Elmer, USA) under standard conditions. Primer sequences utilized in this study will be provided by authors upon request. In the attempt to reduce possible bias of selection in the sample due to other mutations, an analogous procedure was used to screen MKRN3 gene sequence to rule out the most frequent genetic cause of CPP nowadays individuated. All genetic analyses were performed at Department of Woman, Child, General and Specialized Surgery of University of Campania "L. Vanvitelli", Naples, Italy.

Data analysis
Epidemiological data were expressed as medians (interquartile ranges). Fisher exact test was used to compare our polymorphisms allele frequency to those reported in gnomAD. Mann-Whitney U test or Kruskal-Wallis test were used as appropriate to compare genotype subgroups for each single-nucleotide polymorphism (SNP). Results reached statistical significance at a p-value less than .05. All statistical analyses were performed using Stat-Graph Centurion XVII software for Windows.

Results
Clinical and laboratory characteristics of our cohort are shown in Table 1. Median age at first occurrence of thelarche was 5.6 (min-max: 1.1-5.9 years). Family history of precocious sexual development was identified in 26.9% of them. No mutation was found in MKRN3 locus in the whole sample.
All patients were treated by GnRH analogues (triptorelin) with a good response. In particular, two patients with a breast budding occurred before 3 years of age, included in our cohort, had a very rapidly progressive form of precocious puberty, with accelerated growth, thelarche progression and in one case uterine bleeding, both requiring blocking treatment. All these clinical features excluded the alternative diagnosis of minipuberty in these patients.
No rare variants in the coding region of PROKR2 were identified. Five polymorphisms were found as listened in Table 2. All except one patient had more than one SNPs in different combination. Four SNPs had a MAF similar to that reported in literature (Table 2). A statistically significant difference in MAF was found for SNP rs3746682: 0.84 in our cohort vs 0.25 from gnomAD (pvalue <.0001).
No differences were found for median age at puberty onset among different genotypes subgroups for each SNP (Table 3).

Discussion
PROKR2 plays a critical role in regulating olfactory bulb morphogenesis and sexual maturation [11]. A hyperactive prokineticin system appears to be an obvious pathogenetic mechanism of sexual precocity. Nevertheless, no sequence rare variation was detected in the coding region of PROKR2 in our cohort.
We do realize that our sample size is relatively small, however it is exceptionally homogenous due to strict requirements of selection: all patients had an idiopathic CPP with onset before the age of 6 and genetic analysis negative for MKRN3. This recruitment method allows to exclude the "grey zone" between 7 and 8 years where normal variants of "accelerated puberty/rapid progressive thelarche" can occur without medical intervention needed. In addition, it sets up a cohort with the most similar clinical features to the index case of PROKR2 mutant precocious puberty reported by Fukama et al. Therefore, although negative, the finding of our study remains interesting. Indeed, we can speculate that if mutation in PROKR2 coding region could explain 10% of CPP, we would have had identified at least one mutation within our sample size with a 95% level of confidence and a precision of 90%. Therefore, we can suggest that PROK2R is not at least a common cause for CPP, even in early onset CPP.
Those findings suggest that as for other GPCR (e.g. KISS1R) [6], gain of function variants are a very rare cause of CPP probably because hypersignal in this pathway are barely tolerated due to the critical role exerted by those genes.
Since we evaluated only coding region and intronexon joints, we cannot rule out the presence of possible mutations outside those regions. Promoter regions and miRNAs regulatory elements should be the next target of further investigations as they may also play an important role in transcriptional and post-transcriptional control of gene expression as the experience with MKRN3 gene in CPP has taught us [12].
We found five different SNPs in our cohort: 4 of them showed allele distribution similar to gnomAD, suggesting no individual role in CPP. The current knowledge and results of this study are insufficient to evaluate if particular combinations of polymorphisms might affect gene expression and contribute to determine CPP onset. On one side, no difference in allele frequency distribution between our cohort and gnomAD for 4 out of 5 SNPs identified and their characterization as benign variants would exclude their potential role in CPP. On the other side, epigenetic mechanism of action could be Interestingly, we observed a statistically significant difference in MAF of rs3746682 polymorphism in our cohort compared to literature report. It is highly difficult to distinguish whether this different allele frequency has a clinical meaning or is a stochastic finding due to our small sample size. In fact, as rs3746682 (p.Thr195=) is a synonymous SNP, it probably does not display a pathological role. However, this variant has been associated with methamphetamine dependence in Japanese population [13] and bipolar disorder in Japanese females [14].
Although belonging to a different medical field, these findings suggest that this particular SNP could modulate individual's response to certain drugs and susceptibility to develop mental disease. As CPP is a multifactorial condition, we cannot exclude rs3746682 involvement in the complex interaction between environment and genetic background. However, in absence of functional studies, it remains unclear which would be the molecular mechanism of action of this polymorphism in altering gene expression. It would be interesting to investigate rs3746682 in larger cohorts and with functional studies.
Besides, it is important to remember that murine studies on the prokineticin system demonstrated that PROK2-PROKR2 signalling seems to be implicated in many other hypothalamic functions such as the regulation of suprachiasmatic nucleus circadian clock [15]. Therefore, it is possible that mutations of this gene might cause not isolated CPP in the context of more complex phenotypes. The role of prokineticin system in the pubertal timing regulation and its disruption in humans still warrants further investigation.

Conclusions
This study states PROKR2 gene variants are not a common cause of CPP, also in very young girls. At the best of our knowledge, this is the first study providing a