Phosphomannomutase 2 (PMM2) variants leading to hyperinsulinism-polycystic kidney disease are associated with early-onset inflammatory bowel disease and gastric antral foveolar hyperplasia

Abstract

Phosphomannomutase 2 (PMM2) deficiency causes Congenital Disorder of Glycosylation (PMM2-CDG), but does not have a recognised association with Inflammatory Bowel Disease (IBD). A distinct clinical syndrome of hyperinsulinism and autosomal recessive polycystic kidney disease (HIPKD) arises in the context of a specific variant in the PMM2 promotor, either in homozygosity, or compound heterozygous with a deleterious PMM2 variant. Here, we describe the development of IBD in three patients with PMM2-HIPKD, with onset of IBD at 0, 6, and 10 years of age. In each case, intestinal inflammation coincided with the unusual finding of gastric antral foveolar hyperplasia. IBD disease was of variable severity at onset but well controlled with conventional and first-line biologic treatment approaches. The organ-level pattern of disease manifestations in PMM2-HIPKD-IBD may reflect a loss of cis-acting regulatory control by hepatocyte nuclear factor 4 alpha (HNF4A). Analysis of published transcriptomic data suggests that IBD most likely arises due to an impact on epithelial cellular function. We identify a specific pattern of variation in PMM2 as a novel association of early-onset IBD with distinctive gastric pathology.

Fig. 1

a Endoscopic view of the gastric antrum for P1 demonstrating a polypoid appearance at the pylorus with minor visible surface erosion. b, c Representative gastric antral polypoid pathology from P1 demonstrating a polypoid mucosal appearance with elongated and tortuous, corkscrew-like foveolae, and c smooth muscle wisps extending to the surface. d Endoscopic view of the gastric antrum for P2 demonstrating mucosal surface irregularity, patchy redness, and the suggestion of a minor degree of polypoid change at the pylorus with more significant surface erosions and sub-mucosal haemorrhage. e Representative gastric antral pathology from P2 demonstrating focal elongation, branching, and dilatation of the gastric pits with additional moderately chronic active inflammatory changes. f Endoscopic view of the gastric antrum for P3 demonstrating patchy redness

Fig. 2

a Simplified cartoon illustrating proposed mechanism of organ/tissue specificity. ZNF143 binds the WT promotor (left) and CTCF-binding sites, altering the 3-dimensional structure of the CFTF delimited chromatin loop and bringing the PMM2 promotor into proximity with HNF4A binding sites. In cells where HNF4A is expressed, it is thereby able to interact with the PMM2 promotor and function as a cis-acting regulatory element. The c.-167G > T mutant promotor (right) has reduced affinity for ZNF143, disrupting the approximation of promotor and HNF4A binding sites and reducing the HNF4A-dependent transcriptional regulation (adapted from Rubio Cabezas et al. 2017). b Comparison of tissue level expression of HNF4A (upper bar chart), and PMM2 (lower), (images available from The Human Protein Atlas v21.1 https://www.proteinatlas.org/ENSG00000140650-PMM2/tissue, https://www.proteinatlas.org/ENSG00000101076-HNF4A/tissue) (Uhlen et al. 2015): PMM2 is broadly expressed across tissues, whereas the tissue-specific expression of HNF4A closely matches the disease manifestations of patients with PMM2-HIPKD. c Single cell transcriptomic data from the Gut Cell Survey (www.gutcellatlas.org) (Elmentaite et al. 2021) illustrates that intestinal HNF4A expression is restricted to the epithelium (image from https://www.gutcellatlas.org/spacetime/full/). d Single cell transcriptomic data from human gastric epithelium (Busslinger et al. 2021) identifies isthmus cells as having the highest conjoint HNF4A/PMM2 expression. In ileum and colon, HNF4A and PMM2 are broadly expressed across cell types (Wang et al. 2020). Similar to the stomach, expression by stem- and progenitor cells is prominent