Diarrhea 1, Secretory Chloride, Congenital

A number sign (#) is used with this entry because congenital secretory chloride diarrhea, referred to here as DIAR1, is caused by homozygous mutation in the SLC26A3 gene (126650) on chromosome 7q31.

Description

Congenital secretory chloride diarrhea is an autosomal recessive form of severe chronic diarrhea characterized by excretion of large amounts of watery stool containing high levels of chloride, resulting in dehydration, hypokalemia, and metabolic alkalosis. The electrolyte disorder resembles the renal disorder Bartter syndrome (see 607364), except that chloride diarrhea is not associated with calcium level abnormalities (summary by Choi et al., 2009).

Genetic Heterogeneity of Diarrhea

Other forms of diarrhea include DIAR2 (251850), caused by mutation in the MYO5B gene (606540) on 18q21; DIAR3 (270420), caused by mutation in the SPINT2 gene (605124) on 19q13; DIAR4 (610370), caused by mutation in the NEUROG3 gene (604882) on 10q21; DIAR5 (613217), caused by mutation in the EPCAM gene (185535) on 2p21; DIAR6 (614616), caused by mutation in the GUCY2C gene (601330) on 12p12; DIAR7 (615863) caused by mutation in the DGAT1 gene (604900) on 8q24; DIAR8 (616868), caused by mutation in the SLC9A3 gene (182307) on 5p15; DIAR9 (618168), caused by mutation in the WNT2B gene (601968) on 1p13; and DIAR10 (618183), caused by mutation in the PLVAP gene (607647) on 19p13.

Clinical Features

This disorder was first described by Gamble et al. (1945) and Darrow (1945). Voluminous watery stools containing an excess of chloride are present from a few weeks of age. The children are often premature. Hydramnios, presumably due to intrauterine diarrhea (Holmberg et al., 1975), may complicate pregnancy. Indeed, polyhydramnios is probably an invariant feature.

Pasternack and Perheentupa (1966) described vascular changes resembling those of hypertensive angiopathy in 7 children, aged 1 to 42 months at the time of biopsy. All were normotensive. Kidney and muscle were biopsied.

Lubani et al. (1989) identified 16 affected Kuwaiti children over a 7-year period. All children had a shortened gestational period, abdominal distention, and chronic diarrhea. The serum electrolytes in all patients before treatment showed hyponatremia, hypokalemia, hypochloremia, and metabolic alkalosis. Diagnosis was confirmed by a stool chloride content that exceeded the sum of fecal sodium and potassium. In chloride diarrhea, juxtaglomerular hyperplasia, hyperreninemia and hyperaldosteronism, leading to hyperkaluria and hypokalemia, simulate the Bartter syndrome (see 241200). As in the latter disorder, inhibitors of prostaglandin synthetase have beneficial effects (Minford and Barr, 1980). In the intestinal brush border there is both an Na+/H+ and a chloride/bicarbonate exchange mechanism. A defect in either can impede NaCl absorption and lead to secretory diarrhea. The latter exchange mechanism is defective in chloride diarrhea; the former is deranged in sodium diarrhea (270420).

Inheritance

Both sexes have been affected and 2 sibs appear to have been affected in several families (Kelsey, 1954; Perheentupa et al., 1965), suggesting autosomal recessive inheritance.

Hoglund et al. (1994) reported paternal isodisomy for chromosome 7 in a female with congenital chloride diarrhea. She had inherited only paternal alleles at 10 loci and was homozygous for another 10 chromosome 7 loci studied. Physical status and laboratory tests were normal except for mild high-frequency sensorineural hearing loss. Most remarkable was the fact that she was of normal stature. Maternal uniparental disomy for chromosome 7 has been identified in 3 patients with recessive diseases: 2 patients with cystic fibrosis (Spence et al., 1988 and Voss et al., 1989) and 1 patient with osteogenesis imperfecta (Spotila et al., 1992). Both patients with CF had uniparental isodisomy for the complete chromosome, whereas the third patient displayed heterozygosity at the IGFBP1 locus (146730), consistent with uniparental heterodisomy. Severe growth retardation, present in all 3 patients with maternal uniparental disomy for chromosome 7, suggested that there may be imprinted growth-related genes on the maternal chromosome 7. Alternatively, although less likely, short stature could be due to homozygosity for a recessive mutation that occurs at sufficient frequency to be seen in all 3 patients. The lack of growth retardation in the patient reported by Hoglund et al. (1994) suggests that the paternal chromosome 7 lacks the suggested maternal imprinting effect on growth. The patient reported by Hoglund et al. (1994) was a 23-year-old female attending business school. The mother and father were 39 and 44 years of age, respectively, at the time of her birth. CLD was diagnosed at the age of 8 days on the basis of a high fecal chloride content.

Clinical Management

In patients with congenital chloridorrhea, the oral intake of chloride, sodium, and potassium must exceed fecal output (i.e., there must be a positive gastrointestinal balance) so that obligatory losses in sweat can be replaced. A positive balance can best be insured by a high intake of chloride, even though it exacerbates diarrhea. Aichbichler et al. (1997) concluded that suppression of gastric chloride secretion by a proton-pump inhibitor, omeprazole, reduces fecal electrolyte losses in patients with this disorder and thus promotes a positive gastrointestinal balance. However, this treatment does not reduce the need for careful monitoring of dietary intake, serum electrolyte concentrations, and urinary chloride excretion. Their patient was a 34-year-old man who had had severe diarrhea since birth. In early childhood he had been hospitalized many times because of volume depletion and hypokalemia. Early growth and physical development were delayed. His sibs were highly antagonistic toward him, because he would have diarrhea at any time and smelled bad. After the age of 4 years, he was raised by foster parents. As an adult, he had great difficulty obtaining and keeping a job because of diarrhea, fecal incontinence, and the need for frequent hospitalizations. He had an average of 6 stools per day, with as many as 12 on some days. The stools were large in volume and liquid. Most bowel movements were associated with urgency and many with fecal incontinence. During 8 months of follow-up on omeprazole, the number of stools decreased to 2 to 4 per day, with no fecal incontinence. With potassium supplementation he had only occasional episodes of hypokalemia and the patient returned to work.

Mapping

By studies of linkage disequilibrium as well as genetic linkage in a small number of Finnish families, Kere et al. (1993) obtained initial results suggestive of linkage between CLD and the CFTR gene (602421), which is mutant in cystic fibrosis (219700). However, extended analyses in 8 families established close linkage to chromosome 7 markers proximal to CFTR. Multipoint analyses mapped CLD at D7S496 with a maximum lod score of 9.33. Kere et al. (1993) concluded that the CLD gene is close to but distinct from CFTR. Strong allelic association (linkage disequilibrium) with D7S496 in Finnish patients was consistent with the hypothesis of a single founder.

To identify the CLD gene, Hoglund et al. (1996) constructed and refined a physical map based on a 2.7-Mb YAC contig around D7S496 and identified 2 candidate genes. Four known genes were established: SLC26A3 (126650), PRKAR2B (176912), LAMB1 (150240), and DLD (238331). SLC26A3 is expressed in the gut and encodes a protein with sequence homology to anion transporters; PRKAR2B encodes a regulatory subunit for protein kinase A. Both genes map within 450 kb of D7S496 (which is linked to CLD), making them functionally and positionally relevant candidates for the site of the mutation in CLD.

Molecular Genetics

Hoglund et al. (2001) stated that a total of 3 founder and 17 private mutations in the SLC26A3 gene (see, e.g., 126650.0001) underlying congenital chloride diarrhea had been described in various ethnic groups. They screened for mutations in 7 unrelated families with CLD and found 7 novel mutations as well as 2 previously identified ones. They reported for the first time rearrangement mutations in SLC26A3 (see, e.g., 126650.0004). Molecular features predisposing SLC26A3 for the 2 rearrangements may include repetitive elements and palindromic-like sequences.

Makela et al. (2002) noted that the only extraintestinal tissues showing SLC26A3 expression are eccrine sweat glands and seminal vesicles. They presented a summary of published mutations and polymorphisms of the SLC26A3 gene and reported 2 novel mutations of the gene: a 13-bp deletion (126650.0007) and a trp462-to-ter change (W462X; 126650.0008).

Choi et al. (2009) used whole-exome capture and massively parallel DNA sequencing to identify a homozygous pathogenic mutation in the SLC26A3 gene in a Turkish infant with congenital chloride diarrhea who was initially thought to have renal Bartter syndrome. Sequencing this gene in 39 additional patients referred with a suspected diagnosis of Bartter syndrome identified recessive SLC26A3 mutations in 5 patients. All except 1 presented in infancy with watery diarrhea associated with hypokalemia, increased serum bicarbonate, and high aldosterone; the last patient presented at age 6 years. High stool chloride was documented in 2 patients studied. Choi et al. (2009) emphasized the utility of this novel approach for the identification of pathogenic mutations.

Population Genetics

Holmberg and Perheentupa (1980) estimated that 31 cases of congenital chloride diarrhea in 21 families have been identified in Finland as compared with 30 cases in 24 families elsewhere. Lubani et al. (1989) identified 16 affected Kuwaiti children over a 7-year period, giving an estimated incidence of 7.6 per 100,000 live births.

Hoglund et al. (1996) demonstrated that the Finnish form of congenital chloride diarrhea is caused by a homozygous mutation in the SLC26A3 gene. Homozygosity for the same mutation, deletion of GGT(val) of codon 317 (126650.0001), was found in all 32 patients studied. The reverse use of the Luria-Delbruck equation resulted in an estimation of the age of the mutation. The calculation gave an average age of 19 generations, with a range of 13 to 25. This estimate was in agreement with the population history and suggested that the spread of the val317-to-del mutation in the expanding subpopulation in eastern Finland started 400 to 450 years ago.

Makela et al. (2002) described the geographic and population distributions of 3 founder mutations in the SLC26A3 gene: the Finnish V317del mutation (126650.0001), the Polish I675-676ins mutation (126650.0005), and the Arabic gly187-to-ter mutation (G187X; 126650.0006). They also tabulated genetic disorders with congenital or neonatal diarrhea as a main symptom.