Divalent metal transporter 1 (DMT1) was shown to be important for iron metabolism through studies of the mk/mk mouse, which carries a spontaneous mutation (G185R) resulting in defective intestinal iron absorption and anemia. To further investigate the importance of this transporter in vivo, we inactivated the DMT1 gene through targeted deletion of transmembrane domains 3, 4 and 5. We obtained mice carrying a universally inactivated DMT1 allele (DMT1−/−) and mice carrying a floxed DMT1 allele that could be selectively inactivated by breeding to mice expressing a Cre recombinase transgene. We crossed the floxed mice to mice expressing Cre under the control of the Villin promoter to inactivate DMT1 exclusively in the intestine. All mutations were made on a 129S6/SvEvTac background. For comparison, we backcrossed the mk mutation onto the same background. We noted that, in this context, the mk/mk phenotype was much less severe.

DMT1−/− mice were born with substantial liver iron stores, indicating that fetal DMT1 is not required for placental iron transfer or hepatic iron loading. However, DMT1−/− mice invariably died of anemia by day 7, suggesting that DMT1 is important for erythroid iron acquisition but not absolutely required. Transfusions improved survival. The DMT1−/− phenotype was much more severe than the mk/mk phenotype, and compound heterozygous DMT1-/mk mice survived to adulthood. These results confirm that the mk mutation does not result in total loss of protein function.

Intestine-specific inactivation of DMT1 caused no abnormalities at birth, but progressive anemia developed thereafter. At 12 weeks, mice lacking intestinal DMT1 were more anemic than mk/mk mice. To investigate the importance of DMT1 in hematopoietic cells, we transferred fetal liver hematopoietic stem cells (HSC) from DMT1−/− animals to irradiated wild type adult mice. Twelve weeks after transfer, DMT1−/− HSC recipients were anemic, but less so than singly transfused DMT1−/− mice or mice lacking intestinal DMT1. DMT1−/− HSC recipients accumulated more liver iron than wild type HSC recipients, suggesting a compensatory increase in intestinal iron absorption and/or altered iron distribution.

Surprisingly, though DMT1−/− mice died as neonates, heterozygosity for a null Hfe mutation allowed some DMT1−/−;Hfe+/−animals to survive without transfusions. The effect was more pronounced in animals homozygous for Hfe null mutations. We took advantage of this to compare untransfused DMT1−/− mice with mk/mk mice on an Hfe−/− background. Liver iron content and hemoglobin levels of both mk/mk;Hfe−/− and DMT1−/−;Hfe−/− mice were significantly lower than DMT1+/+;Hfe−/− mice, as expected. Again, mk/mk mice were less severely affected than DMT1−/− mice.

We have confirmed that DMT1 plays primary roles in intestinal iron absorption and erythroid iron acquisition but it is likely that there are minor, alternative iron uptake mechanisms. The mk mutation allows for considerable residual iron transport activity. Inactivation of Hfe ameliorates the effects of DMT1 deficiency through an as yet unknown mechanism.

Author notes

Corresponding author

Sign in via your Institution