Characterization of porcine betaine homocysteine methyltransferase (BHMT) and betaine homocysteine methyltransferase-2 (BHMT-2) genes

R.S. Ganu, T.A. Garrow, M. Sodhi, L.B. Schook
Experimental Biology Annual Meeting, April 18-22, 2009, New Orleans, LA


Betaine homocysteine methyltransferase (BHMT) and BHMT-2 methylate homocysteine to form methionine using betaine or S-methylmethionine, respectively. These enzyme activities are only observed in the liver of adult rodents, whereas in adult humans and pigs it is detected in the liver and kidney cortex. Because of this similarity, we have chosen the pig as a model to study the spatial and temporal expression of these enzymes and to determine whether the BHMT and BHMT-2 genes are transcribed into multiple mRNA isoforms. This report describes our progress to date.

Immunohistochemical staining revealed the presence of BHMT in adult liver and kidney cortex, as reported earlier, but we also found immunodetectable levels BHMT in fetal lungs (aged days 30, 60, 84, 90, 105 of gestation). The BHMT and BHMT-2 cDNAs were subsequently cloned and sequenced, and their 5’ and 3’ UTRs were amplified using RLM-RACE. BHMT has a longer 5’ and 3’ UTR, consisting of 77 and 1,142 nucleotides, whereas BHMT-2 UTRs were composed of 17 and 893 nucleotides, respectively. The deduced amino acid sequences of BHMT and BHMT-2 contain 407 and 363 amino acids, respectively, and share 78% amino acid identity. Relative to BHMT-2, BHMT has two additional regions of amino acid sequence, a 9 amino acid sequence (86-94) in the N-terminal region, and a 34 amino acid sequence (373-407) at the carboxy terminus. Eight splice variants of porcine BHMT have been observed and one variant found in the kidney medulla and heart encodes a truncated form of BHMT. Although we do not know if this mRNA is efficiently translated and whether the resulting protein is stable, if it is this protein is predicted to lack BHMT activity because it doesn't have critical determinants for binding the enzyme's catalytic Zn. We have modeled this truncated form of BHMT and the results show a dramatic change in tertiary structure when compared to wild type BHMT. The model predicts the truncated protein to adopt a horseshoe fold, whereas wild type BHMT is a (βα)8 barrel. The function of this hypothetical protein remains unknown.