Abstract
The LmrA multidrug resistance (MDR) efflux pump of Lactococcus lactis is a member of the ATP-binding cassette (ABC) superfamily (TC 3.A.1). LmrA catalyses drug export in a process driven by ATP hydrolysis. Since a primary form of energy (ATP) drives transport, such a system is called a primary active transporter.
Recently, Venter et al. (2003) used molecular genetic techniques to cleave the ATP-hydrolysing domain from the integral membrane transporter domain of LmrA and demonstrated that the latter could catalyse drug transport driven by the transmembrane proton electrochemical gradient. Substrate : proton symport or antiport is a general characteristic of simple carriers also called secondary active transporters. Their observations suggested that the primary active transporter, LmrA, was derived from a secondary carrier by superimposition of the ATP-hydrolysing subunit onto the carrier during evolution as suggested previously (Saier, 2000a).
There are four superfamilies of secondary active transporters that include members that catalyse drug export using a drug : cation antiport mechanism. These are the MF (TC 2.A.1), RND (TC 2.A.6), DMT (TC 2.A.7) and MOP (TC 2.A.66) superfamilies (Saier, 2000b; Saier & Paulsen, 2001). We were curious to know if LmrA could be shown to exhibit sufficient sequence similarity to the members of any one of these superfamilies to suggest a functional or an evolutionary connection.
Fig. 1 shows a binary alignment of an extended region of LmrA with a portion of the MexB MDR efflux pump of the RND superfamily (Tseng et al., 1999). These two sequences exhibit 28 % identity and 42 % similarity for this 98 residue position alignment. A shorter segment including the first 63 residue positions gave 32 % identity and 47 % similarity. Using the GAP program with 500 random shuffles (Devereux et al., 1984), comparison scores of 9·5 standard deviations (SD) and 11·6 SD were obtained for the longer and shorter binary comparisons, respectively. The probability of obtaining such a score by chance is less than 1020. These comparison scores are sufficient to establish that the sequence similarity observed for these regions within the two proteins did not arise by chance.