Investigation of the ABC transporter MRP1 in selected patients with presumed defects in vitamin B12 absorption

To the editor:

Recently, Beedholm-Ebsen and colleagues reported that the ABCC1 gene is involved in the efflux of cobalamin (Cbl; vitamin B12) from the enterocytes and other cells to the plasma.¹ ABCC1 encodes the multidrug resistance protein 1 (MRP1), a member of the ATP-binding cassette (ABC) transporters. MRP1 was shown to export unbound Cbl across the membrane.² 

Cbl is usually bound to various carrier proteins: to haptocorrin or gastric intrinsic factor (IF) after release from the food, or to haptocorrin and transcobalamin 2 in the serum.³  For intestinal absorption, the Cbl-IF complex is recognized by cubam, a multiligand endocytic receptor on the enterocyte in the intestine.⁴  After entering the enterocyte, lysosomal enzymes degrade the IF, and then transcobalamin 2 transports Cbl to the tissues via the blood. ABCC1/MRP1 is an alternative pathway for Cbl to exit the enterocyte.¹ 

Hereditary vitamin B12 uptake deficiency is attributed to defects in CUBN, AMN, or GIF.⁵  We have examined more than 150 patients or sibships with recessive hereditary Cbl malabsorption^6-8  (and unpublished results). Approximately 80% were mutated in either AMN or CUBN (causing Imerslund-Gräsbeck syndrome)^6,7  or in GIF (causing intrinsic factor deficiency).^8,9  However, some 15%-20% of cases remain unexplained, and they represent an ideal cohort to screen for potential mutations in the ABCC1 gene. These presumed Cbl malabsorption cases were excluded for defects in CUBN, AMN, and GIF because of incompatible genetics and/or direct sequencing, and we chose 18 unrelated patients and the mothers of 2 additional patients from North America, Europe, the Middle East, and North Africa to study ABCC1. Using exon-by-exon genomic DNA sequencing, we screened the entire gene, including flanking splice junctions, for any sequence changes.

We detected a total of 27 changes in ABCC1 compared with the published sequence (supplemental Table 1, available on the Blood Web site; see the Supplemental Materials link at the top of the online letter). Four intronic insertions/deletions (14.8%) and 2 missense (7.4%), 6 silent (22.2%), and 15 intronic single nucleotide polymorphisms (SNPs) were identified (55.6%). Twenty-six of the 27 changes have a SNP identifier number (supplemental Table 1). One change in the 5′-UTR has not been described before: c.1-79_66del14bp, which was found to be heterozygous in one of the patients. All identified changes were ruled out as potentially causal for Cbl deficiency because they did not fulfill Mendelian rules for recessive inheritance (even when considering compound heterozygosity for 2 different ABCC1 changes in one patient), were intronic, not affecting the open reading frame, or found in the normal population (SNPbase). Two rare missense SNPs were each found once in heterozygosity. G671V appears to be a conservative change, whereas R723Q is potentially functional, but both were seen in the population at large (SNPbase).

MRP1 is the first known eukaryotic membrane efflux transporter capable of transporting unbound Cbl out of cells.¹  Based on the observation that Mrp1^−/− mice present no Cbl-related phenotype,¹  one might conclude that MRP1 plays no role in inherited Cbl deficiency. However, whereas humans with various loss-of-function mutations in AMN develop Imerslund-Gräsbeck syndrome,⁶ Amn^−/− mice (by both random and targeted Amn inactivation) die prenatally with the amnionless phenotype,¹⁰  indicating that the same gene may have different functions during embryogenesis in humans and rodents. Analogously, we surmised that mutational screening of ABCC1 in suspected Cbl deficiency was warranted if all known causal genes were excluded beforehand. Based on screening this highly selected patient cohort, we conclude that mutations in the ABCC1 gene, if they exist, are unlikely to cause overt Cbl deficiency in humans. On the other hand, defects in ABCC1 that result in a milder or different phenotype would probably not prompt a clinical follow-up suspecting Cbl malabsorption. Accordingly, alternative Cbl efflux pathways yet to be identified may have to explain the remaining cases of inherited Cbl deficiency.¹