Structural variants (SVs) that disrupt topologically associating domains can cause disease by rewiring enhancer-promoter interactions. Duplications involving GPR101 are known to cause X-linked acrogigantism (X-LAG) through ectopic GPR101 expression, but not all of these duplications are pathogenic. This presents a diagnostic challenge, especially in the prenatal setting. We evaluated POSTRE, a tool that predicts the regulatory impact of SVs, to distinguish pathogenic from benign GPR101 duplications. We analyzed seven non-pathogenic duplications and 27 known X-LAG-associated duplications. To enable predictions in an X-LAG-relevant tissue, enhancer maps built using H3K27ac ChIP-seq, ATAC-seq, and RNA-seq data derived from human anterior pituitary samples (NIH research protocol 97-CH-0076, Clinicaltrials.gov Identifier NCT00001595, submitted on 11 March 1999) were integrated into POSTRE. POSTRE correctly classified all 34 duplications as benign or pathogenic. In addition, one X-LAG case with mild clinical features (i.e. severe growth hormone hypersecretion without pituitary tumorigenesis) was found to include only 2/5 VGLL1 enhancers, whereas all typical X-LAG cases had ≥4 enhancers duplicated. This suggests that partial enhancer hijacking at VGLL1 could explain the different clinical features in this individual. These findings support the utility of POSTRE to support diagnostic pipelines when interpreting SVs affecting chromatin architecture in pituitary disease and highlight its potential to reduce uncertainty in genetic counseling without requiring chromatin conformation capture assays.

Structural variants (SVs) that disrupt topologically associating domains (TADs) can cause disease by rewiring enhancer-promoter interactions. Duplications involving GPR101 are known to cause X-linked acrogigantism (X-LAG) by enabling aberrant expression of GPR101 through hijacking of enhancers at VGLL1. However, not all GPR101-containing duplications are pathogenic, presenting a diagnostic challenge, especially in the prenatal setting. We evaluated POSTRE, a tool designed to predict the regulatory impact of SVs, to distinguish pathogenic from benign GPR101 duplications. We analyzed six non-pathogenic duplications, and 27 known X-LAG associated pathogenic duplications. Tissue-specific enhancer maps built using H3K27ac ChIP-seq and ATAC-seq data as well as gene expression data derived from human anterior pituitary samples were integrated into POSTRE to enable predictions in a X-LAG relevant tissue context. POSTRE correctly classified all 33 duplications as benign or pathogenic. In addition, one X-LAG case with mild clinical features (e.g., severe GH hypersecretion in the absence of pituitary tumorigenesis) was found to include only 2/5 VGLL1 enhancers (also predicted to be the weakest enhancers), whereas all 26 typical X-LAG cases had ≥4 enhancers duplicated. This suggests that milder enhancer hijacking at VGLL1 could explain the different clinical features of X-LAG in this individual. These findings support the utility of POSTRE to support diagnostic pipelines when interpreting SVs affecting chromatin architecture in pituitary disease. By accurately modelling enhancer adoption in a cell type-specific context, POSTRE could help to reduce uncertainty in genetic counselling and offers a rapid alternative to performing chromatin conformation capture experiments.

The orphan G protein coupled receptor GPR101, which is implicated in X-linked acrogigantism (X-LAG), a rare pituitary disorder characterized by rapid growth a few years after birth, has received significant attention for its expression pattern in adult vertebrate tissues. However, the characterization of GPR101 expression during early embryonic development is poorly characterized. In this study, we investigated the spatiotemporal expression patterns of Gpr101 during early embryonic mouse development (E7.5-E15.5) using a global Gpr101 knock-in mouse model with an inserted LacZ reporter, Gpr101tm1b(KOMP)Mbp. Similar to previously published studies using adult tissues, we found that LacZ reporter expression was largely restricted to regions of the central nervous system. Expression was not detected until E10.5 in a region near the telencephalic vesicle. In contrast to what has been reported in adult tissues, Gpr101 expression was absent in the hypothalamus and pituitary gland during the developmental timepoints assessed. These novel observations provide a more comprehensive characterization of GPR101's expression and may offer insights into its role in growth and development across species.

X-linked acrogigantism (X-LAG; OMIM: 300942) is a rare X-linked dominant, fully penetrant form of infancy-onset pituitary gigantism caused by Xq26.3 tandem duplications involving the GPR101 gene. All previously reported X-LAG-associated duplications disrupt the integrity of the resident topologically associating domain (TAD). This creates a neo-TAD, permitting ectopic chromatin interactions between GPR101 and centromeric pituitary enhancers postulated to lie between RBMX and VGLL1, and culminating in pituitary GPR101 misexpression and growth hormone excess. Conversely, none of the few previously reported cases of Xq26.3 duplications in unaffected individuals include the tissue-invariant TAD border that shields GPR101 from its centromeric enhancers. Preservation of this boundary has thus been considered synonymous with non-penetrance of X-LAG. We examined a series of four family members from the same kindred with an incidentally detected GPR101-containing Xq26.3 duplication involving the invariant TAD border. Chromosome microarray demonstrated an interstitial chromosome Xq26.3 duplication: arr[GRCh37] Xq26.3(135,954,223 - 136,224,319)x2, including GPR101, the TAD invariant border and RBMX, but not VGLL1. None of the relatives with the Xq26.3 duplication exhibited evidence of growth hormone excess, making this the first unaffected family with a GPR101-containing Xq26.3 duplication involving the invariant TAD border. The predicted neo-TAD in this kindred excludes the VGLL1 region, which is present in all previously described X-LAG patients and absent in all previously described unaffected individuals with Xq26.3 duplications. Our clinical findings suggest that TAD border involvement is not sufficient for X-LAG to develop, and implicates the VGLL1 region as likely the sole pituitary enhancer responsible for GPR101 misexpression and the X-LAG phenotype. Pending corroborative studies, this new insight into X-LAG pathogenesis may guide interpretation of future Xq26.3 duplications and counselling of families in whom such duplications are found.

Experimental structures solved through cryo-electron microscopy have recently been published for GPR101, a G protein-coupled receptor (GPCR) implicated in the genetic condition X-linked acrogigantism (X-LAG). Here, we compared these experimental structures with computational models that we previously published, including our internally developed homology models and third-party models generated through the AlphaFold2 and AlphaFold-Multistate artificial intelligence (AI) methods. Our analysis revealed considerable accuracy for both homology models and AI-generated models. However, it also revealed the general superiority of AI methods. Particularly noteworthy is the model generated by AlphaFold2, which captured with high fidelity various structural aspects, including the challenging second extracellular loop. Our previously published homology model of the GPR101-Gs protein complex, based on the β2-adrenergic receptor, accurately predicted the binding mode of the G protein to the receptor. Moreover, this model predicted the structure of the sixth transmembrane domain (TM6) significantly more accurately than the others, including those built through AI methods, suggesting that homology modeling based on templates solved in complex with the G protein of interest might be the most reliable way of modeling this transmembrane domain. Lastly, our analysis revealed that our molecular dynamics simulations did not have a significant and consistent effect on the accuracy of the models, increasing the accuracy for some domains while decreasing it for others. This work provides insights into the relative strengths of different modeling approaches for our case study on GPR101. More broadly, when considered alongside other assessment studies, it contributes to the growing body of knowledge that can guide the modeling of GPCRs for which experimental structures are not yet available.