Genomic Insights into Growth and Its Disorders: An Update

Abstract

Purpose of review

To provide an update of the most striking new developments in the field of growth genetics over the past 12 months

Recent findings

A number of large genome-wide association studies have identified new genetic loci and pathways associated to human growth and adult height as well as related traits and comorbidities. New genetic etiologies of primordial dwarfism and several short stature syndromes have been elucidated. Moreover, a breakthrough finding of Xq26 microduplications as a cause of pituitary gigantism was made. Several new developments in imprinted growth-related genes (including the first human mutation in IGF-II) and novel insights into the epigenetic regulation of growth have been reported.

Summary

Genomic investigations continue to provide new insights into the genetic basis of human growth as well as its disorders.

Introduction

The field of growth genetics has continued to evolve at a rapid rate over the past year with a number of key insights into the regulation of normal growth as well as new genetic etiologies of severe growth disorders. In this manuscript, we will review the latest findings of large genome-wide association studies of height and related traits as well as discuss their implications for general health. We will then focus on new genetic etiologies of primordial dwarfism followed by recent findings in genes important to growth plate function presenting with short stature, in particular mutations in NPR2 and ACAN. Next, we review a groundbreaking finding in pituitary gigantism. Finally, we will briefly mention new developments in imprinted genes and epigenetics related to growth.

Text of review

1. Population-based studies and height

The continuous search for genetic determinants of height is actively ongoing in an attempt to better understand the fundamentals of human growth. In November 2014, the largest genome wide association study (GWAS) to date investigating the genetic factors underlying adult height was published by the GIANT consortium(1). In this study, authors combined GWAS data from more than 250,000 individuals of European ancestry. They identified 697 common single nucleotide polymorphisms clustered in 423 different loci across the human genome which together explain ~16% of the phenotypic variation in height. Furthermore, they calculated that ~50% of the variation in height in the general population is due to common genetic variants. To understand the biological significance of these loci, they used a number of bioinformatics techniques to identify the likely candidate genes in the significantly associated regions that affect height. They then performed network analyses of these candidate genes. Importantly, in addition to identifying many of the known growth pathways (e.g. growth hormone/IGF-1, TGF-β and CNP signaling), many of these loci were found to be enriched for genes and pathways that were not previously recognized to be associated with growth, such as mTOR, osteoglycin and hyaluronic acid opening up new avenues for investigation into the biology of growth.

In addition to GWAS studies of adult height, birth length and infant height were also evaluated by a recent meta-analysis of 22 different GWAS as part of the Early Growth Genetics consortium(2). In this study seven SNPs were found to be independently associated with birth length, four of which were located in or near loci that were previously known to be associated with adult height. After replication studies, only one of the three novel SNPs (rs905938) was found to be significantly associated with birth length. This specific SNP is located within the DCST2 gene which encodes a member of the DC-STAMP-like protein family involved in regulation of osteoclast fusion(2). Interestingly, while there was some overlap between the genetic loci found to underlie adult height and infant birth length, a genetic prediction score based on 180 of the adult loci only explained 0.13% of the variance in infant length. This suggests that there are different genetic factors influencing fetal and adult growth.

Another GWAS reported genetic determinants of sitting height (SH) in a cohort of 25,000 individuals.(3) Specific analysis of SH is of interest, since SH can be differentially affected in the diverse forms of skeletel dysplasia (e.g. decreased SH in spondylo-epiphyseal dysplasia versus increased SH in achondroplasia). Moreover, SH has been independently linked to a number of indicators of the metabolic syndrome including blood pressure, dyslipdemia and insulin resistance. The GWAS identified six genome-wide significant loci including the regions containing IGFBP3 and TBX2. Furthermore, 130 of the 670 adult height loci were nominally significant with 59 of those leading to an increased SHR and 71 to a decreased SHR. Based on the direction of the association, they were able to identify which loci had a differential effect on leg length versus spine and head length. In the loci affecting leg length, there was a clear enrichment for genes involved in the bone, cartilage and growth plate pathways(3).

It has been well established by epidemiological studies that taller individuals have a decreased risk of coronary artery disease (CAD) when compared to shorter individuals. In groundbreaking work, Nelson et al.(4) investigated the genetic underpinnings of this relationship using a Mendelian randomization technique. They calculated a score for genetically determined height using 180 of the adult height GWAS SNPs and correlated this score with risk of CAD in 65,066 cases and 128,383 controls. They found that a genetically determined lower adult height was associated with a significant increase in coronary artery disease with a relative increase of 14% per 1 SD decrease of height. This association was partly mediated by significant associations of genetically determined height with levels of LDL cholesterol and triglycerides but not with other traditional cardiovascular risk factors. On further analysis, growth associated genes and atherogenic pathways showed considerable overlap(4) suggesting that there are shared biological pathways affecting stature and risk of CAD. These findings were confirmed in a second Mendelian randomization performed by the CARDIoGRAMplusC4D consortium(5). In contrast to the protective effect on cardiovascular disease, taller stature appeared to increase the risk of colorectal cancer in women in a large Mendelian randomization study with an estimated OR of 1.15 (95% C.I: 1.05–1.26) for every 10 cm height increase, but this effect was not observed in men(6). Further work is needed to understand the mechanistic link between growth biology and the risk of CAD and colorectal cancer.

2. Primordial Dwarfism

Microcephalic primordial dwarfism (MPD), defined as a syndrome of severe pre-natal and post-natal growth failure (> −4 SDS) with associated microcephaly, has seen a surge in new etiologic discoveries over the past decade with further progress in the past year. Previously well-known etiologies of MPD included genetic mutations in cell cycle regulation, DNA damage repair, microtubule formation and centrosome function. With respect to the latter category, Martin et al recently demonstrated mutations in MPD patients in PLK4, which plays a key role in centriole duplication and is necessary for adequate ciliary function.(7) Interestingly, in addition to typical findings of MPD, these patients had retinopathy which is often seen in other ciliopathies. Patients with mutations in TUBGCP6, the substrate for PLK4, were also found to present with MPD. A second group identified a homozygous frameshift mutation in PLK4 in 3 individuals with MPD from a consanguineous Saudi family(8). Other mutations in genes affecting centrosomal function, most notably MAP4, have been described in a similar fashion(9).

Further expansion of the genotypic abnormalities in MPD was provided by the demonstration that compound heterozygous frameshift mutations in NSMCE2, a gene involved in stabilization of chromatin structure, led to a clinical phenotype of severe growth failure, insulin resistance and gonadal failure(10). Around the same time, three research groups independently published articles, which reported human mutations in the non-homologous end-joining DNA repair enzyme XRCC4 as a cause of primordial dwarfism (11–13). Interestingly, the autosomal recessive XRCC4 deficiency phenotype is highly comparable to that of the hypomorphic NSMCE2 mutation and also includes hypergonadotropic hypogonadism and early-onset metabolic syndrome with severe insulin resistance in the absence of obesity. In addition, another study demonstrated that autosomal recessive missense mutations in the homologous DNA repair enzyme MCM9, lead to a strikingly similar phenotype of short stature and hypergonadotropic hypogonadism, much like the XRCC4 phenotype(14). Overall, the subcategory of DNA repair syndromes is increasingly being recognized as a cause of severe growth failure. Gonadal failure appears to be a consistent finding in a number of these DNA damage repair syndromes. Insulin resistance is also a common feature, but the precise mechanism linking DNA damage repair with insulin resistance is yet to be determined.

3. Genetics of short stature syndromes

A considerable number of genetic etiologies have been identified for various syndromes associated with short stature over the past year. Although it extends beyond the scope of the current article to describe all of these findings individually, it is important to note that many of these individual gene discoveries were made within the specialized field of skeletal dysplasia with disproportionate short stature. These genetic findings will lead to more accurate patient classification in future years allowing for better prognostication and potentially altering therapeutic approaches in a subset of patients. We focus on new findings in two genes, NPR2 and ACAN, which have important clinical ramifications for the practicing endocrinologist.

NPR2

Cumulative evidence has been generated for the important role of the C-natriuretic peptide (CNP) signaling pathway in endochondral bone formation and linear growth. The NPR2 gene codes for the CNP receptor (NPR-B), and when mutated in either the homozygous or compound heterozygous state, it causes severe growth failure in the form of acromesomelic dysplasia, Maroteaux type.(15) One study last year found that 2% of ~300 patients with non-syndromic idiopathic short stature demonstrated heterozygous inactivating NPR2 mutations (16). Interestingly, in the small subset of 22 patients who had a parent with a recorded height below −2 standard deviations, three of these patients (13.6%) had mutations in NPR2. This suggests that heterozygous mutations in NPR2 may be a relatively common cause of dominantly inherited familial short stature. Reciprocally, an activating heterozygous NPR2 mutation was identified in an individual with tall stature taken from the extremes of a population based study. An additional case report identified an activating mutation in NPR2 in a family with a syndrome of tall stature, macrodactyly of the great toes, scoliosis, coxa valga, and slipped capital femoral epiphysis²⁵. A study of 173 patients with disproportionate short stature, who were initially referred for possible Léri-Weill Dyschondrosteosis (LWD) but in whom no SHOX mutation had been identified, revealed heterozygous NPR2 mutations at a remarkably similar incidence rate of 3% (6/173)(17). In this study, no NPR2 mutations were found in 95 cases of proportionate ISS. Interestingly, none of the disproportionately short patients with a NPR2 mutation in this study had a Madelung deformity, one of the typical features seen in LWD.

The NPR2 pathway inhibits downstream signaling of FGF receptor 3 (FGFR3), in which activating mutations lead to achondroplasia or hypochondroplasia. This past year, Makrythanasis et al. describe the first homozygous inactivating missense mutation in FGFR3 in a patient with tall stature(18). Given the central role that the CNP/NPR2/FGFR3 pathway plays in chondrocyte development and growth, it has been targeted for potential therapeutic intervention. Wendt et al. created a modified CNP variant with increased resistance to proteolytic cleavage thereby increasing its therapeutic half-life. This analog was shown to have enhanced growth-promoting effects in a mouse model of achondroplasia(19). Currently, a Phase 2 clinical trial of this analog is ongoing in patients with achondroplasia (clinicaltrials.gov NCT02055157). In addition, statin treatment was recently shown to rescue the bone dyplasia phenotype in a murine achondroplasia model, as well as correction of cartilage degradation in pluripotent stem cells that were derived from human achondroplasia patients (20). These striking findings may provide a novel avenue for therapy in achondroplasia patients.

Aggrecan

The large majority of children with short stature will either present with normal or delayed bone age at the time of their initial clinical evaluation. Therefore, the combination of proportionate short stature with an advanced bone age is highly remarkable and led to further genetic evaluation of three families in whom several family members displayed this unusual combination. In all the affected individuals, heterozygous mutations were identified in the ACAN gene, which codes for the extracellular matrix proteoglycan aggrecan, a key component of the cartilage matrix(21). Prior to this discovery, there had only been three families reported with ACAN mutations, each with a distinct skeletal dysplasia. Premature growth cessation occurred in a number of the patients with cessation of growth prior to menarche in affected females. Arthritis was variably present in affected individuals. A fourth patient with a similar presentation was subsequently reported(22). The full phenotypic spectrum of this disorder is yet to be elucidated, but patients with dominantly inherited short stature and advanced bone age should certainly be tested for mutations in ACAN.

4. Pituitary gigantism and tall stature

The other side of the spectrum of growth disorders, i.e. tall stature and overgrowth syndromes, has been marked this year by a large study describing an inborn X-linked gene microduplication syndrome as the cause of early-onset (early childhood) gigantism. In this study, the authors found microduplications of Xq26 in a third of patients (13/43) with gigantism, both familial and sporadic cases.(23)** The smallest region of overlapping duplication amongst the patients included 4 genes, one of which encoded GPR101, an orphan G-protein-coupled receptor. GPR101 mRNA was found to be overexpressed in patients’ pituitary tissue. When a cohort of 248 acromegalic patients was screened, a recurrent GPR101 variant (p.E308D) was identified in 11 (3 with constitutive mutations and 8 with mutations in tumor DNA only). Functional studies showed that the p.E308D mutation led to an increase in growth hormone production in vitro as well as increased proliferation of growth hormone producing cells. This variant is present at low frequency in the general population and the exact role that it plays in acromegaly still requires confirmation. A second study of 208 patients with gigantism found 12 patients harboring an Xq26.3 microduplications confirming this as a causal etiology for early-onset gigantism(24). The discovery of a genetic basis for gigantism is both remarkable and important, since until recently, few genetic pathways in acromegaly were known other than familial cases due to AIP mutations. These findings will open up new avenues of investigation into the underlying pathophysiology of growth hormone producing adenomas.

In comparison to short stature, genetic syndromes associated with extreme tall stature are relatively rare. One of these is Primrose syndrome, characterized by tall stature, macrocephaly, developmental delays, insulin-resistant diabetes, and facial dysmorphisms. Exome sequencing in four unrelated individuals suspected to have this condition, showed de novo missense mutations in ZBTB20 in all four individuals with functional validation of the genetic variants in vitro(25). ZBTB20 encodes a specific transcriptional repressor and plays a role in growth, glucose metabolism, and neurogenesis.

5. Imprinted Genes and the Epigenetic regulation of growth

This past year has also seen some significant advances in the understanding of imprinted genes and the epigenetic regulation of growth. Russell Silver Syndrome is a well-known cause of short stature due to genetic imprinting defects, most commonly hypomethylation of the H19/IGF-II domain or maternal uniparental disomy of chromosome 7. The importance of IGF-II in regulating not only pre-natal but also post-natal growth was highlighted by an intriguing study in the NEJM, in which the authors describe the first human heterozygous IGF-II mutation (p.Ser64Ter) in four affected individuals from the same family born severely small for gestational age with progressive post-natal growth failure(26). Segregation analysis was consistent with paternal transmission and maternal imprinting of the IGF-II gene. The patients had some dysmorphic features often found in RSS including a triangular face, broad forehead and clinodactyly suggesting that IGF-II deficiency may be a causal factor for these features in patients with RSS.

In a study examining the role of epigenetics in RSS, Prickett et al. performed a genome-wide methylation study in 18 patients with RSS and identified two additional genomic regions near RB1 and ANKRD11 which were hypermethylated in a considerable percentage of patients (70 and 33%, respectively)(27). This study did not show a consistent pattern of hypermethylation on multiple genetic loci within any of the individual RSS patients, highlighting the genetic heterogeneity of this patient group.

Moreover, in many patients with severe short stature, no genetic sequence defects in the GH/IGF-1 axis can be demonstrated, but epigenetic phenomena may still impair proper functioning of these genes. A series of recent studies by Pierre Bougneres’ group found that the methylation status of a cluster of CGs dinucleotides located within the proximal part of the P2 promoter of the IGF-1 (IGF1) gene was correlated to individual GH sensitivity on the IGF-1 generation test as well as first year response to recombinant GH treatment in ISS patients(28, 29). Methylation status as a single CG dinucleotide (CG-137) explained 25% of the variance of growth response to growth hormone in the study of 136 children²⁸. Another study from the same group indicated elevated methylation patterns in CGs at both the P2 and P1 promoter region of the IGF-1 gene in 94 children with ISS compared to 119 age-matched children of normal stature(30). Overall these findings provide important new insights linking epigenetic phenomena to growth and its related disorders. It is to be expected that this area of genetic epiregulation will make further advancements in the next few years.

Conclusion

This past year has seen tremendous developments in our understanding of both the common and rare genetic factors affecting growth biology. Much work remains to translate these findings into improved care for our patients.

Key points.

• Large GWAS have identified a number of potential novel growth-related pathways.

• NPR2 and ACAN mutations lead to distinct clinical subgroups of patients with dominantly inherited short stature.

• The list of etiologies of primordial dwarfism continues to expand, with DNA damage repair syndromes and various mutations in centrosome function.

• Early-onset childhood gigantism (acromegaly) may be caused by germline Xq.26 microduplications

• Epigenetic regulation of growth, including IGF-II imprinting and methylation status of the IGF-1 promotor, is an evolving field of research