Mice Are Not Small Furry People! A New Hirschsprung Disease Model Lets Us Pretend This Is Not True

In 1888, Harald Hirschsprung¹ described 2 children with intractable constipation, massive abdominal distension, malnutrition, and new-onset explosive diarrhea. These unfortunate children probably died from sepsis as colon bacteria escaped epithelial and immune cell barriers in leaky aganglionic distal colon. Aganglionic means lacking enteric nervous system (ENS) ganglia, the clusters of intestinal neurons and glia that control bowel motility, epithelial, and immune cell function in response to local stimuli. Fortunately, for most people, the ∼600 million enteric neurons, more than 20 neuron types, and specialized enteric glia work so well that we can eat what we want and pursue our passions without consciously controlling approximately 30 feet of amazing bowel. When ENS is missing or defective, misery ensues, a connection first shown in 1949 by Swenson et al² in children with Hirschsprung disease (HSCR). Although distal HSCR bowel looks normal, it lacks enteric ganglia and lacks propagating contractions. Swenson et al² correctly surmised that aganglionic bowel caused functional obstruction and invented the Swenson pull-through surgery to remove aganglionic bowel and reattach normal bowel near the anal verge. Children with HSCR often are dramatically better after pull-through surgery, but some have persistent constipation, stool leakage, or “Hirschsprung-associated enterocolitis” (abdominal distension, explosive diarrhea, lethargy, and sepsis risk). We still need new treatments and prevention strategies.

One major mystery is why outcomes vary so much after pull-through surgery. In part, the answer may lie in differences in bowel physiology between affected children. Even before treatment, some neonates are critically ill with a distended abdomen, bilious vomiting, and fever (±bowel perforation), and need urgent surgery. Other children with HSCR appear well for years with minimal therapy. In fact, a 53-year-old man in Japan was diagnosed recently with HSCR.³ He had chronic constipation (weekly bowel movements) on a magnesium-based laxative. After stopping his medicine he went a month without passing stool and was diagnosed with HSCR. This remarkable range of symptoms suggests genetic or nongenetic disease modifiers exist that could be targeted to improve outcomes.

HSCR occurs when neural crest–derived ENS precursors fail to fully colonize bowel during the first trimester of pregnancy.⁴ These ENS precursors depend on the transmembrane tyrosine kinase receptor ret proto-oncogene (RET) for survival, proliferation, and efficient migration (first shown in 1994).5, 6, 7 RET kinase activity usually is low in people with HSCR. RET is activated in the ENS by glial cell derived neurotrophic factor (GDNF) and neurturin (NRTN) via GDNF family receptor alpha 1 (GFRA1) and GDNF family receptor alpha 2 (GFRA2) respectively. RET transcription depends of paired like homeobox 2B (PHOX2B), SRY-box transcription factor 10 (SOX10), retinoic acid receptor beta (RARB), GATA binding protein 2 (GATA2), and paired box 3 (PAX3). The genes encoding these proteins are linked to HSCR. SOX10 competes with SRY (the male sex–determining gene) for RET regulatory elements, perhaps explaining the 4:1 male/female ratio in short-segment HSCR. RARB is activated by retinoic acid, a vitamin A derivative made by aldehyde dehydrogenase 1 family member A2 (RALDH2). Vitamin A deficiency causes HSCR-like disease in mice (and might increase human HSCR risk). Raldh2-/- mice have total intestinal aganglionosis (similar to Ret-/- mice and RET-/- human beings). SOX10 mutations cause HSCR with deafness, patchy skin depigmentation, and peripheral neuropathy (Waardenburg–Shah syndrome, WS4C). PHOX2B mutations cause HSCR with congenital central hypoventilation syndrome (Haddad syndrome). RET, GDNF, GFRA1, and retinoid signaling partially explain why 20% of children with HSCR have congenital anomalies of the kidneys and urinary tract. Thus, RET is central to HSCR pathogenesis.

In human HSCR, aganglionosis is limited to the distal colon 80% of the time, suggesting that machinery needed to make ENS is present but not working efficiently. Typically, in human beings, combinations of mild risk alleles conspire to prevent full bowel colonization by ENS precursors. More than 30 genetic loci (including trisomy 21) impact HSCR occurrence.^8,9 Maternal medicines, nutrition, and illness all seem likely to impact HSCR incidence. To prevent HSCR and find new cures, we need great model systems.

In this issue of Cellular and Molecular Gastroenterology and Hepatology, Sunardi et al¹⁰ describe the first model based on a human RET mutation (S811F) in which the mouse (RetS812F) closely mimics human disease. The RETS811F human has HSCR, unilateral kidney agenesis, and oligomeganephronia (reduced nephron numbers). Ret^S812F/+ mice have distal colon aganglionosis (50%) or hypoganglionosis (50%), small kidneys, and 10% unilateral renal agenesis. Combining Ret (S812F) with Ret9 (hypomorphic) or Ednrb+/- (HSCR risk alleles) increased aganglionosis to 100%. RETS811F probably prevents adenosine triphosphate binding to the kinase domain, generating dominant-negative RET that homodimerizes without phosphorylating wild-type RET. Ret^S812F/+ ENS precursors have reduced proliferation, reduced migration, and increased apoptosis, with abnormal enteric neuron subtype ratios. It is easy to say, “I knew that would happen,” but it took nearly 30 years to generate Ret-variant mice closely mimicking human HSCR. Thankfully, Ret^S812F/+ live many weeks, facilitating promising studies of enterocolitis, stem cell therapy, and regenerative medicine (GDNF, 5-hydroxytryptamine receptor 4 (5-HT4)–receptor agonist). Although we need to remember that mice are not small furry people, we now can pretend as we seek new cures.