ATP binding cassette (ABC) transporters constitute one of the largest families of transport proteins. The occurrence of multidrug resistance (MDR) in human cancer cells has been correlated with the over expression of human ABC, P-glycoprotein (Pgp). Streptomyces peucetius produces two anticancer agents, doxorubicin and daunorubicin, that belong to the anthracycline family of antibiotics. The organism is self-resistant to the potent effects of the antibiotics it produces due to the action of an efflux pump, DrrAB. Both Pgp and DrrAB carry out similar functions, but in two different cell types. An understanding of the bacterial drug transporter DrrAB is thus expected to help in obtaining a better understanding of the function and evolution of the multidrug transporter P-glycoprotein. In DrrAB, the catalytic and membrane domains are present on separate subunits, DrrA and DrrB respectively. How the catalytic ATP-binding domains and the membrane domains in transporters interact with each other, or how energy is transduced between them, is not well understood. We introduced several single cysteine substitutions in DrrB and then by using a cysteine to amine hetero-bifunctional cross-linker showed that DrrA interacts predominantly with the N-terminal cytoplasmic tail of DrrB. Within this region of DrrB, we also identified a sequence with similarities to the EAA motif found in importers of the ABC family of proteins, thus leading to the proposal that the EAA or the EAA-like motif may be involved in forming a generalized interface between the ABC and the TMD of both uptake and export systems. By using a combination of approaches, including point mutations and disulfide cross-linking analysis, we show here that the Q-loop region of DrrA plays an important role in dimerization of DrrA as well as in interactions with DrrB. Furthermore, we also show that the interaction of the Q-loop with the N-terminus of DrrB is involved in transmitting conformational changes between DrrA and DrrB. The scope of the present study further extends into identifying the factors involved in the biogenesis of the DrrAB pump. We have identified two accessory proteins namely, FtsH and GroEL that may be involved in proper folding and assembly of the transporter.

FtsH is an essential AAA family Zn++ metalloprotease of Escherichia coli, possessing ATPase-dependent chaperon activity and ATP-dependent protease activity. Heat shock transcription factor Sigma32, LpxC, SecY, and bacteriophage protein CII are some of the substrates of FtsH. Although FtsH is known to influence several cellular processes, the role of FtsH in bacterial cell division had not been identified. FtsZ is the principal cell division protein that marks the cell division site at mid-cell by forming a ring structure. Using a pair of ftsH-null and isogenic wild type strain of E. coli, earlier studies in the laboratory had demonstrated that proteolytic function of FtsH is required for the presence of FtsZ rings at mid-cell site. It was also shown that FtsZ is not a substrate for FtsH protease in vivo. In view of these observations, using a pair of ftsH-null and isogenic wild type strain of E. coli, experiments were carried out to find out the mechanism behind the requirement for FtsH protease for the presence of FtsZ ring at mid-cell site. Viability of the cells having ftsH-null status was maintained by a suppressor mutation at another locus, and was found to be comparable to that of isogenic wild type cells. Immunostaining for FtsZ showed that only 20% cells of ftsH-null strain of E. coli has FtsZ ring at mid-cell site, On the contrary, more than 90% cells of isogenic wild type cells had FtsZ ring at mid-cell site. Live cell imaging with FtsZ-GFP also showed similar results. Low fraction of ftsH-null cells having FtsZ ring was found to be independent of slow growth rate of the cells. Confocal microscopy revealed that ftsH-null cells lacked the normal helical spiral-type structure of FtsZ, unlike the intact FtsZ helices present in isogenic wild type cells. FtsZ protein levels in the membrane and cytoplasmic fractions of ftsH-null cells were found to be same as those in the isogenic wild type cells. Exogenous expression of wild type FtsH in ftsH-null cells could restore FtsZ ring status to normalcy, similar to that in the isogenic wild type cells. However, this restoration could not be accomplished by FtsH mutants, which were lacking in ATP binding, ATPase, or protease activities. FtsA anchors FtsZ to the membrane and a specific FtsZ/FtsA ratio is known to be critical for cell division. Further, FtsA and/or ZipA are required for the stabilisation of FtsZ ring at mid-cell site. The levels of FtsA were found to be lower by more than 2.5-fold in all the membrane and soluble fractions of ftsH-null cells. The levels of FtsA were found restored to normalcy upon complementation with exogenous expression of FtsH. Low levels of FtsA were not due to the slow growth of ftsH-null cells. Exogenous expression of FtsA or FtsA-GFP restored FtsZ in more than 90% of ftsH-null cells. Moreover, FtsA mutants, which are defective in the interaction with FtsZ, did not restore FtsZ rings to normalcy. The levels of ZipA were found to be same in ftsH-null and isogenic wild type cells. Expression of ZipA or ZipA-GFP could restore FtsZ rings to normalcy in ftsH-null cells. These data showed that low FtsA levels might be the reason for low percentage of cells having FtsZ ring in ftsH-null cells. It implied that ftsH-null cells might have been managing FtsZ ring stabilisation with ZipA, to facilitate septation. Real time RT-PCR showed that the levels of ftsA mRNA and those of all the other fts genes, except ftsZ, in the 16-gene dcw cluster, were found to be low in ftsH-null cells. Moreover, real time RT-PCR using specific primers designed for multiple promoters of ftsZ and for the RNaseE processing site, just upstream of ftsZ, showed that the levels of transcripts of the genes upstream to RNaseE site were significantly low and that the levels of ftsZ transcripts, which were downstream to RNaseE site, were unaffected. On the contrary, the levels of mRNAs of fts genes, such as ftsE, ftsX, ftsN, and zipA that were located at another part of the genome, were normal in ftsH-null cells. These observations suggested that the reason for the low levels of FtsA protein might be low levels of ftsA mRNA. In addition, the low levels of other fts mRNAs from the dcw cluster, and probably of the respective proteins, might contribute to the slow growth of ftsH-null cells. The ftsH null strains also showed less compact nucleoids and the nucleoids did not look bilobular. This data suggested that there may be some defect in the compaction of nucleoids in ftsH-null cells. On the contrary, isogenic wild type cells, when grown slow like the growth of ftsH-null cells, had no defect in nucleoid compaction and looked bilobular. The proper compaction of nucleoids could be restored only by wild type FtsH, but not by the protease mutant of FtsH. These observations suggest that proteolytic activity of FtsH might be required for the proper compaction of nucleoids, which in turn might have influence on the placement of FtsZ ring at mid-cell site. In parallel, different percentage of silver stained single-dimension SDS-PAGE showed conspicuous difference in the protein profiles of the membrane and soluble fractions of ftsH-null cells, in comparison to those of isogenic wild type cells. FtsZ co-immuno precipitation (CoIP) of total cell lysates of ftsH-null and isogenic wild type cells showed differential interaction of two proteins, the outer membrane protein A (OmpA) and a 50 kDa protein, between the two strains. The level of OmpA was 2.5-fold high in ftsH-null cells, in comparison to that in isogenic wild type cells. However, overexpression of ompA in isogenic wild type cells did not have any effect on FtsZ rings in isogenic wild type cells. Two-dimensional gel electrophoresis for membrane and soluble fractions of ftsH-null cells, in comparison with that of isogenic wild type cells, showed that several proteins in each fraction were either present or absent between these two strains. Most of these proteins were then identified using MALDI-TOF / LC –MS methods. Identification of these proteins, which were present differentially between ftsH-null and isogenic wild type cells, has revealed existence of many more hitherto unidentified potential substrates of FtsH and therefore cell processes, which FtsH may influence.

The PhD thesis is on the studsy of the influence of the ftsZ antisense RNA and FtsH protease on the synthesis and function of the Escherichia coli cytokinetic protein, FtsZ, which mediates septation during cell division. Thus, it involves three molecules, FtsZ, ftsZ antisense RNA, and FtsH protease. While the E. coli ftsZ antisense RNA is being identified and structurally and functionally characterised for the first time, there has been some earlier studies in the laboratory in which the FtsH protease was found to have influence on the presence of the FtsZ rings at the mid-cell site. The Chapter 1 is the Introduction to the thesis presented in 3 parts –Part 1A, 1B, and 1C, introducing FtsZ and bacterial cell division, bacterial antisense RNAs, and FtsH protease, respectively. The Chapter 2 gives the description of the Materials and Methods used in the study. The Chapter 3 presents the identification, structural and functional characterisation of the ftsZ cis-antisense RNA, and its role in the regulation of FtsZ protein levels. Initially, the expression of cis-encoded antisense RNA from E. coli ftsZ loci was demonstrated during the different growth phases of the bacterium (RT-PCR/qPCR data). Antisense RNA is expressed from three promoters (primer extension and promoter probe data) on the complementary strand of the ftsZ coding region and terminates at the singletrand te complementary toftsAthegenethat 3’islocatedregionupstreamof theofftsZ the gene. Induced overexpression of a portion (423 bp) of the antisense RNA, spanning the ftsZ AUG codon and the ribosome binding site of ftsZ mRNA, from pBS(KS) could downregulate the synthesis of FtsZ protein to approximately 30%, leading to cell division arrest and filamentation of the cells at 42°C. This effect was less dramatic at 30ºC, probably due to less melting of the antisense RNA. Immunostaining performed on the induced culture did not show FtsZ ring formation after overnight induction whereas reduction in the proportion of the cells carrying FtsZ rings could be clearly observed after 2 hrs of induction. Real time PCR analysis performed for relative quantitation of ftsZ mRNA and ftsZas RNA from different growth phases (0.2 to 2.5 OD600 nm) showed growth phase dependent expression of the antisense RNA. While the levels of ftsZas RNA were found to be high at lower OD cultures or early growth phase cultures, the levels were found to be low at the late log phase and stationary phase cultures. Thus, when the cells are actively dividing and therefore need more FtsZ, the levels of the ftsZas RNA are high, while the cells are not actively dividing and therefore the FtsZ levels are low, the levels of the ftsZas RNA are low. At any phase of the growth, the ratio of the ftsZ mRNA to the ftsZas RNA was always found to be 6:1. Thus, the physiological role the ftsZas RNA is to maintain the availability of the ftsZ mRNA at a level that is commensurate with the requirement for the FtsZ protein during the different stages of the cell growth and division. The Chapter 4 is on the study of the possible mechanism behind the influence of FtsH protease on the presence of FtsZ rings at the mid-cell site during septation in cell division. Immunostaining for FtsZ in the mid-log phase E. coli cells showed that 82% of the AR3289 (ftsH wild type) cells possessed FtsZ rings, while only 18% of the AR3291 (ftsH-null maintained viable by a suppressor mutation) cells showed Z-rings. While the AR3289 cells showed a cell doubling time of 20 min, the AR3291 cells had a cell doubling time of 45 min. The mass doubling time of AR3289 and AR3291 were 24 min and 54 min, respectively. These distinct differences were found in spite of the suppressor mutation suppressing all the deleterious effects of the lack of the essential protease, FtsH. Complementation of the ftsH-null cells (AR3291) with the wild type FtsH but not with the ATP-binding or ATPase, or protease-defective mutants of FtsH, restored the FtsZ ring status to about 80% of the cells. The growth rate of AR3291 was also partly restored to comparable to that of the wild type cells upon complementation. Western blotting for FtsZ, and the FtsZ-stabilising proteins, FtsA and ZipA, showed that the ftsH-null cells have low levels of FtsA, as compared to those in the isogenic wild type cells (AR3289). The levels of FtsZ and ZipA were comparable in both the cells. Quantitative PCR performed for different cell division genes within the dcw cluster showed no sign of change in the ftsA transcript levels in the ftsH-null cells, suggesting that the low levels of FtsA in the ftsH-null cells were not due to transcriptional downregulation. Further experiments showed that the half-life of FtsA protein in the AR3289 cells was 45 min, while that in the AR3291 cells was 24 min. This experiment showed that the low levels of FtsA in the ftsH-null cells was due to the low half-life of FtsA in the cells. Growth synchronisation of the AR3289 and AR3291 cells showed that the levels of FtsA prior to cell division stage do not increase in the ftsH-null cells as much as in the isogenic wild type cells. Thus, the ftsH-null cells must be somehow managing the division through the partial stabilisation of FtsZ rings by ZipA. Interestingly, immunostaining for FtsH in AR3289 cells showed the presence of FtsH at the mid-cell site, as co-localised with FtsZ, for a brief period prior to cell constriction. These observations suggest the involvement of FtsH in cell division process. The faster degradation of FtsA in the absence of FtsH protease implies that another protein, which may be a protease that directly degrades FtsA or a chaperone that helps the unfolding of FtsA for degradation, might be the substrate of FtsH protease. The absence of FtsH protease brings up the levels of this unknown protein, which in turn facilitates (if it is a chaperone) degradation of or directly degrades (if it is a protease) FtsA. This model for the link among FtsH, FtsA levels, and the presence of FtsZ has been proposed based on the observations. Thus, the present study reveals for the first time an FtsA-linked role for FtsH protease in the presence of FtsZ ring at the mid-cell site and hence in bacterial septal biogenesis. The thesis is concluded with the list of salient findings, publications, and references.