Cloning of electron-transferring flavoprotein from Megasphaera elsdenii.

Electron-transferring flavoproteins (ETF) are heterodimeric enzymes that contain FAD and which transfer reducing equivalents between other flavoenzymes. In mitochondria they accept reducing equivalents from reduced forms of fatty acyl CoA dehydrogenases, sarcosine dehydrogenase, and dimethylglycine dehydrogenase and transfer them to a third flavoenzyme that reduces Coenzyme Q. ETF has similar functions in microorganisms such as Paracoccus denitrificans, Methylotroph WA3 1 and the strictly anaerobic rumen bacterium Megasphaera elsdenii 111. In M. elsdenii it couples the oxidation of NADH or the flavoenzyme Dlactate dehydrogenase to the reduction of butyryl CoA dehydrogenase (BCD) leading to the production of short-chain fatty acids which are excreted by the organism [2]. This enzyme differs from similar enzymes purified from other sources in having NADH dehydrogenase activity and in containing two moles of FAD rather than one FAD [1,2]. The gene for the M. elsdenii enzyme is being cloned to further investigate these differences in structure and function. ETF was purified using a published procedure 131. The subunits of the protein were separated by SDS-PAGE and their N-terminal amino acid sequences determined. An homology search of the EMBL sequence library using BLAST 141 showed that the N-terminal of the p-subunit had 10096 identity with a region downstream of BCD [S]. Two oligonucleotides were synthesised for use in PCR, one of which corresponded to a region between the BCD and ETF genes. The other oligonucleotide corresponded to a region in the Nterminus of the a-subunit of ETF. PCR was carried out with M. elsdenii genomic DNA under standard conditions [6]. The 0.9kb product obtained was ligated into the TA cloning vector (Invitrogen) and used to transform Eschen'chia coli INVaF' cells. The recombinant plasmid was purified from a culture grown from a single colony. The base sequence of the insert confirmed that the gene for the a-subunit is downstream of that for the psubunit and showed that the two genes are separated by 19 bases (Fig. 1 ). The p-subunit gene is separated from the BCD gene by 40 bases. Preliminary alignment of the first 116 amino acids in the N-terminus of the p-subunit with the p-ETF's of human liver and Paracoccus denitrifcans showed little overall homology, but several regions of conserved sequence were identified. Interestingly, there is somewhat greater sequence homology with the fLvA gene products of the N2furing bacteria Rhizobium meliloti, Azorhizobium caulinodans, and Azotobacter vinelandii for which a biological function has not yet been determined.