The genetic background for Prader-Willi syndrome (PWS) is abnormal genomic imprinting during gametogenesis, which causes a loss of paternal gene expression on chromosome 15 at the molecular level, in the 15q11-q13 region. About 65 – 75% of patients with PWS have a de novo paternal deletion within this critical region. Based on three breakpoints (BP) in this area, three classes of deletions may be distinguished – common type 1 deletion (BP1-3), type 2 deletion (BP2-3) and a rare atypical microdeletion 15q11.2 (BP1-2) (1, 2). An additional 20 – 30% of PWS patients exhibit maternal uniparental disomy (mUPD) of chromosome 15, and the remaining 1 – 3% of cases have microdeletion or epimutation in the imprinting center (1, 3, 4). Very rarely, an unbalanced Robertsonian chromosomal translocation may occur (5).

PWS, with an estimated incidence of 1 in 10,000 to 20,000 newborns, is the most common cause of syndromic life-threatening obesity (6, 7). Eating disorder in PWS can be divided into 4 phases: 1. hypotonia, feeding difficulties, and failure to thrive (from birth to 15 months of age); 2. weight gain without significant change in appetite or caloric intake or with only a slight increase in interest in food (from 15 months to 5 years); 3. development of hyperphagia, typically accompanied by food-seeking and lack of satiety (from 5 to 13 years); some adults progress to phase 4, which is the loss of an insatiable appetite and the achievement of a feeling of satiety (8). Several reports have demonstrated oxytocin abnormalities in PWS patients, which may be causal for several symptoms - poor suckling response at birth, hyperphagia with food addiction, poor social skills and emotional dysregulation. The results of animal studies indicate that oxytocin therapy is one of the important therapeutic strategies for children with PWS (9, 10). Excessive eating behavior in PWS appears to result from hypothalamic dysfunction, as do other key clinical features such as short stature, testicular retention due to hypogonadotropic hypogonadism, and, in some patients, central hypothyroidism and/or central hypocorticism (4). Furthermore, due to hypothalamic syndrome, pubertal development may be incomplete, delayed or completely absent and this can also lead to complications in the reproductive sphere. On the contrary, signs of central precocious puberty (CPP) have been reported in a minority (approximately 3.5%) of children with PWS (11–13).

The candidate gene for abnormal pubertal development in PWS is the maternally imprinted MKRN3 that is located in 15q11.2, within the PWS critical region (14). MKRN3 contains a characteristic ring zinc finger motif and encodes a specific protein (macorin ring finger protein 3) (15), which is abundantly expressed in the hypothalamus and is involved in hormonal regulation of puberty by the inhibition of gonadotropin-releasing hormone (GnRH) secretion (16, 17). In murine models, increased expression of MKRN3 in the hypothalamus was reported in the neonatal and juvenile period with a subsequent decrease before pubertal onset, which is suggestive of its role to inhibit activation of hypothalamo-pituitary-gonadal axis (18). The inhibitory role of the MKRN3 protein is further supported by a decrease in its serum level in both girls and boys before the onset of puberty (19, 20), whereas girls with idiopathic CPP have lower serum MKRN3 when compared to prepubertal peers of similar age (21). The loss-of-function pathogenic variants in the MKRN3 gene are the most common genetic cause of CPP (22–24).

Real-life clinical experience shows aberrant pubertal development in Prader-Willi syndrome, but clear data on the timing of the pubertal milestones have not been clearly elucidated yet. Even more, identification of the MKRN3 gene within the Prader-Willi syndrome critical region as the master regulator of pubertal development led to the need of testing the potential link between the MKRN3 gene expression and/or sequence and individual pubertal development in PWS patients. In this study, we aimed to estimate pubertal milestones in 37 PWS patients (18 females) with median age 11.1 (95% CI: 6.4 – 18.8) years on the background of recent normative data, to correlate bone age (BA) and gonadarche onset and evaluate the link between variant pubarche and gonadarche and the MKRN3 gene sequence or expression level in individual genetic subtypes of PWS.

In this cross-sectional study combined with relevant retrospective data, we investigated 37 patients with PWS with spontaneous pubarche and/or gonadarche. We originally hypothesised that pathogenic variants in the MKRN3 gene might be the cause of this abnormal pubertal pattern. MKRN3 is included in the cluster of genes causative for PWS and is known to act as a repressor of puberty initiation (24, 34). If a loss-of-function variant in the MKRN3 gene causes CPP, all patients with PWS caused by Type 1 and 2 deletions including MKRN3, would have to develop a CPP phenotype. However, we were unable to confirm this hypothesis, the percentage of those PWS adolescents with CPP is similar to the healthy population (5.4% in our cohort vs. 3.5% in healthy population) (13, 35). Interestingly, both PWS girls with CPP in our cohort had mUPD, but not all girls with mUPD have precocious puberty, suggesting that mUPD is not a predictor of CPP in patients with PWS.

We aimed to employ the newest normative data (“pubertograms”), thanks to which the age at onset of pubarche, gonadarche and further progression of puberty in our patients could be objectively evaluated. Most previous studies have concluded that puberty is delayed or incomplete in patients with PWS (36). However, our results have shown that the original concept of delayed puberty onset was modified by our results showing that in PWS patients the gonadarche occurs at a similar age to the healthy population consistently with others studies (36–38), occurring at a typical age of 10.0 years in girls and 11.7 in boys. On the other hand, pubarche occurs in both sexes substantially earlier than in the healthy population. Age at pubarche is correlated neither with age at GH treatment initiation, nor with BMI (39). However, BA in PWS patients was advanced by 0.7 years, and is significantly correlated with pubertal onset (gonadarche). In a previous study, accelerated BA was associated with GH treatment, which is known to accelerate skeletal maturation (40). This may also be causative in our patients. Alternatively, accelerated BA can be caused by the early onset of androgen production, premature pubarche (adrenarche) and obesity as observed by other observational association studies (41, 42), however the natural course of pubarche cannot be evaluated. Another factor influencing the early or late onset of puberty and reproductive physiology in PWS patients may be hyperleptinemia, which is derived from body composition and the number of adipocytes (43, 44).

We have confirmed that pubarche progression in adolescents with PWS is prolonged by over 1 year compared to the healthy population in both sexes. However, we failed to prove that there was no difference in breast development and testicular volume progression compared to the healthy population. The recent study (37) together with our paper, consistently show that these patients do not reach the adult stage of pubertal development without gonadal hormone replacement. Thus, hormone replacement is not needed to initiate puberty but is essential for its progression to later stages and for reaching the adult stage of maturation. Therefore, the early onset of pubarche and abnormal pubertal pattern is suggestive of a dissociation pattern from adrenal maturation leading up to adrenarche, and the functionality of the hypothalamic-pituitary-gonadal axis in these patients.

The relationship between genotype and phenotype in PWS patients is complex and is influenced by specific genetic mechanisms leading to the disorder (4). Understanding these genotype-phenotype correlations can help in managing and providing tailored care for individuals with PWS. However, we have shown that this is not applicable to pubarche, gonadarche, BA and pubertal progression. The onset and progression of puberty in PWS patients cannot be predicted by their genotype.

The clinical assessment of puberty was one of the most significant challenges. Therefore, we conducted a search for recent studies of pubertal development in populations of similar ethnicity and socioeconomic standard to allow comparison for our cohort with patients with PWS. For comparison, we used reference standards from Denmark (28–30). Despite this, due to the fact that it is not the same population, this limitation remains. If we eliminated this limitation, we would not be able to evaluate pubertal development in a standard way.

Another limitation was that only pubarche or only incipient signs of gonadarche were present in some of our patients. Therefore, it was not possible to evaluate the further progression of these signs in the whole cohort and part of these results were influenced by the size of the group. We will have more accurate information about the progression of puberty in these patients in a few years.

Regardless of the genetic subtype and the MKRN3 gene status, children with PWS had early and slowly progressing pubarche. However, the age at gonadarche was normal and further progression of puberty was insufficient. The BA at the onset of gonadarche was advanced and was significantly associated with gonadarche. Genetic background was not suitable predictor for the timing of pubertal onset.