The most commonly observed chromosome 20 abnormality is a ring-shaped chromosome 20, r (20). Those patients with r (20) are characterized by growth retardation, intellectual problems, facial dysmorphism, seizures and specific electroencephalogram pattern (Porfirio et al., 1987). However, interstitial chromosomal deletions around 20q13.2-13.32, containing GNAS locus, are very rare. Till now, only 7 patients from 5 reports were shown to possess this type of deletions. They all shared similar phenotypes of feeding difficulty, growth retardation, abnormal facial characteristics, abnormal subcutaneous fat distribution and intellectual disability (Aldred et al., 2002; Geneviève et al., 2005; Butler et al., 132013; Balasubramanian et al., 2015; Solomon et al., 2011). Here, we detected a unique 4.36 Mb paternal deletion on chromosome 20q13.2-q13.32 in a Chinese boy via next generation sequencing (NGS)-based CNV analysis and SNP array-based haplotyping. He showed common phenotypes compared with the other patients. Remarkably, he had good response to growth hormone (GH) treatment, and his height was increased by 23.5 cm after about 3 years’ treatment.

In this case report, we presented a Chinese boy showing learning difficulties, growth retardation and endocrine abnormalities, with a rare de novo 4.36 Mb deletion on paternal 20q13.2-q13.32. Via exploring the online database (http://grch37.ensembl.org/index.html), this region was found consisting of 35 protein coding genes and 27 non-coding genes. Among them, there are 9 OMIM genes (MC3R, AURKA, PCK1, VAPB, TUBB1, STX16, ATP5E, GNAS and EDN3). Noteworthy, GNAS is a haploinsufficient gene locus highly associated with his phenotype (Turan and Bastepe, 2013).

GNAS is an important imprinted locus that is responsible for the production of multiple transcripts: paternally expressed GNAS-A/B, GNAS-AS1, GNAS-XLαs and maternally expressed NESP55, while Gαs is biallelically expressed in most tissues, with maternal expression in some specific tissues (eg. neonatal brown tissue, gonads and thyroids) (Bastepe, 2007; Turan and Bastepe, 2015). Gsα is the most well-studied transcript expressed from GNAS locus, encoding the alpha subunit of G-protein involving in guanine nucleotide-binding. G-protein can trigger complex signaling networks via GDP-GTP exchange. The activated form of GTP-bound Gsα can stimulate different effectors, such as adenylyl cyclase, and thereby initiate various cellular responses, including in endocrine glands (Cabrera-Vera et al., 2003; Bastepe and Jüppner, 2005; Turan and Bastepe, 2015). For example, adenylyl cyclase facilitates the production of cAMP, which can bind to the protein kinase A (PKA) and further lead to the activation of its catalytic subunit. Catalytic-active PKA is involved in the iodine uptake and the expression of genes associated with thyroid hormone production (Pereira et al., 2020). As a result, mutation of GNAS can cause different types of endocrine-related diseases, such as McCune-Albright syndrome (OMIM#174800), progressive osseous heteroplasia (OMIM#166350) and pseudohypoparathyroidism (PHP).

It is worth to note that different forms of PHP can be triggered by heterozygous mutation of GNAS, such as PHP-Ia, PHP-1b, and pseudopseudohypoparathyroidism (PPHP) (Mantovani et al., 2020). PHP-Ia and PPHP are caused by inactivating mutations in Gs-a, while PHP-1b is due to an abnormal methylation on GNAS imprinted regions (Turan and Bastepe, 2015; Mantovani et al., 2020). PHP-Ia patients are characterized by short stature, subcutaneous ossification, brachydactyly and variable degrees of intellectual disability, so called Albright’s Hereditary Osteodystrophy (AHO), as well as parathyroid hormone (PTH) resistance. It may also show resistance to other hormones, such as growth hormone releasing hormone (GHRH) and TSH (Aldred et al., 2000; Weinstein et al., 2001; Levine, 2002). PPHP patients show most AHO phenotypes, but not presenting hormone resistance (Aldred et al., 2000; Mantovani et al., 2020). PPHP often develops when the paternal Gs-a is deficient, while PHP-Ia develops when the maternal one is inactivated (Davies and Hughes, 1993). The less severe phenotype of PPHP may be due to the tissue specific paternal imprint of Gs-a. In our case, the phenotype and genotype of our patient is more consistent with the PPHP. However, deletion of other genes may also play roles in the resulting phenotypes.

To date, there were only 7 patients reported to carry this rare deletion around 20q13.32 containing GNAS imprinted locus (Figure 3D), and 4 of them were confirmed to have paternal deletion (Table 1). Almost all of those affected by 20q13.32 deletion containing GNAS showed reduced birth weight, abnormal facial appearance, feeding difficulties, intellectual disabilities and growth retardation (Aldred et al., 2002; Geneviève et al., 2005; Butler et al., 132013; Balasubramanian et al., 2015; Solomon et al., 2011). The two cases reported by Aldred et al. showed obesity, and those reported by Genevieve et al. had uneven subcutaneous fat distribution, while the patient reported by Balasubramanian et al. had reduced adipose tissue (Aldred et al., 2002; Geneviève et al., 2005; Balasubramanian et al., 2015). Compared with those previous reports, the patient in our case showed similar triangular face, sparse-hair eyebrows, low birth weight, feeding problems, reduced adipose tissue, severe growth delay and intellectual abnormalities. However, when compared with UPD(20)mat and Silver Russell Syndrome (SRS)-like phenotypes, the clinical manifestations in our patient (paternal 20q13.2-q13.32 deletion) was a little bit different. As previously reported, patients with UPD(20)mat generally have phenotypic features such as growth retardation, including intrauterine growth retardation (IUGR), feeding difficulties, about 1/5 patients shared phenotypes like: hypotonia, dysmorphic features, developmental delay (DD) and Silver-Russel-syndrome like phenotypes [http://cs-tl.de/DB/CA/UPD/0-Start.html (accessed 02/25/2022)]. And the clinical diagnosis of SRS was based on six features: pre- and postnatal growth failure, relative macrocephaly, prominent forehead, body asymmetry, and feeding difficulties (Netchine–Harbison clinical scoring system (NH-CSS)) (Tannorella et al., 2021). So our patient shared growth retardation and feeding difficulties with UPD(20)mat patients, and shortened left lower limb fits the phenotype of Silver Russell Syndrome, 3/6 met the NH-CSS criteria. Accordingly, phenotypes of this 20q deletions were different from that of UPD(20)mat and SRS, mainly in the triangular face, sparse-hair eyebrows, reduced adipose tissue, and intellectual abnormalities. In mice and humans, loss of function of XLαs was associated with intrauterine and perinatal growth retardation, as well as poor adaption to feeding (He et al., 2015), which was similar to our patient. But XLαs deficiency also caused hypoglycemia and disrupted glucose counter-regulation which did not occur in our patient (He et al., 2015). Considering that the paternal GNAS allele was deleted, it was also predicted that the patient might have some clinical features of Albright hereditary osteodystrophy (AHO). In our case, The patient had part of the characteristics of AHO, such as: short stature, brachydactyly, he also had limb asymmetry, but he did not have other characteristics of AHO such as obesity, round face and subcutaneous ossification.

The dosage of GH was based on the guideline recommendations for the clinical use of recombinant human growth hormone in Chinese children: childhood GHD: 0.075- 0.15 U/(kg d), ISS: 0.125-0.2 U/(kg d) (Subspecialty Group of End, 2013). The amount of GH used was in this range (Figure 2B). Considering the initial age (5 years) and bone age of the child, and the space for subsequent height gain is still relatively large, our initial GH dose is slightly smaller (0.12 U/(kg d) at 5 years old), and then gradually increased to 0.18 U/(kg d) (at 8 years old) because the growth rate is not very satisfactory. A recent study showed that, after 1 year of GH treatment, HtSDS (Height standard deviation score) increased in 13 ISS children with normal IGF1 level (start at -2.06, end at -1.76) and 17 GHD children with normal IGF1 level (start at -2.23, end at -1.87) (Soliman et al., 2021). GH therapy was relatively effective in SRS patients. Meropi Toumba et al. repoted that 26 SRS patients received GH treatment for at least 7 years [median duration of treatment was 9.8 years (range 7.0-15.7)], and the median height of the 26 patients at the commencement of treatment increased significantly by the end of treatment (height SDS: start at -2.74, end at -1.33) (Toumba et al., 2010). As for GH treatment in PPHP patient, R Manfredi et al. repoted a female PPHP patient, who received a 3-yr-6 months GH treatment. As a result, a significant improvement of growth velocity was obtained, while the height SDS failed to improve (-3.30 pretreatment (10 years old); -3.14(11 years old), -3.36(12 years old) and -3.59 (13 years old) after the first, the second and the third year of treatment, respectively) (Manfredi et al., 1993). As for our patient, although he failed to return to normal height with GH treatment, his height SD changed more in the first 2 years (5–7 years old) than that in those reports (Figure 2A).

Although GH treatment did not restore the child’s height to normal, the increase of about 9.25 cm/year in height during GH treatment (GH treatment data from 5–7 years old) was higher than that of normal Chinese male children by 6.35 cm/year (5–7 years old) (Li et al., 2009), so GH treatment was beneficial for this child. If GH treatment was not taken at that time, the child’s height would be shorter, just as he suspended GH treatment between 7–7.5 years old, and the height increased by only 1.5 cm in half a year. The IGF1 level of 3.25 nmol/L (at initial diagnosis) and the change in IGF1 levels during GH therapy afterwards (Figure 2B), although low, were within the normal range (4–6 years: 2.86-27.04 nmol/L; 7–9 years: 5.2-33.15 nmol/L) (Xu et al., 2010).

Moreover, a genetic testing method combining NGS-based CNV analysis and SNP array-based haplotyping could become a powerful tool to precisely identify the location of a chromosomal deletion and trace its parent of origin. There are also some limitations to our research, due to the limitation of sample collection, the relationship between methylation status in this region and phenotype needs further study in the future.

In conclusion, here we presented a Chinese patient with a novel interstitial deletion around 20q13.2-13.32, containing GNAS locus. He exhibited high concordance with previously reported phenotypes. However, none of the previous reports presented any treatment information on the patients. Therefore, this case report was the first time to provide clinical treatment data on a patient with such rare interstitial deletion on chromosome 20, which shows GH treatment did have a positive effect on his growth.