A major step in the biosynthesis of many membrane proteins is their insertion into the membrane of the endoplasmic reticulum (ER). The insertion of a multi-spanning membrane protein is a complex process since several transmembrane (TM) domains have to be correctly integrated in order to enable its correct assembly. At present it is unclear how the integration of multiple TM domains is co-ordinated by the ER translocon. The aim of this study was to analyse the molecular environment of the TM domains of a model seven TM domain protein, opsin, so as to better understand the mechanism by which integration occurs. For this purpose, stable 'integration intermediates' of defined lengths representing distinct stages of opsin biosynthesis were generated by in vitro translation of truncated mRNA in the presence of semi-permeabilised cells. Cysteine-mediated, site-specific cross-linking and immunoprecipitation were employed to examine the environment of these integration intermediates. In addition, cysteine-specific modification reagents with different physical properties were used to investigate the environment of opsin TM3 during its insertion at the ER membrane. Opsin TM domains exhibit unique patterns of adduct formation with the ER translocon components, Sec61α and Sec61β. TM1 associates with the Sec61 complex at two distinct stages during nascent chain extension, and this behaviour is dependent on the presence of subsequent TM domains. The re-association of TM1 with the transloconmay well facilitate the co-ordinated integration of TMs 1-3 into the lipid bilayer. Opsin TM4 exits the Sec61 complex as soon as the subsequent TM domain is synthesised, while TM5, TM6 and TM7 remain associated with the ER translocon throughout protein synthesis, suggesting their concerted release upon chain termination. Evidence is provided that opsin is integrated via a single Sec61 heterotrimer, despite the fact that the ER translocon appears to consist of multiple copies of the Sec61 complex. On the basis of this work, a model is presented describing the complete integration of opsin at the ER membrane.

Approximately 10% of the open reading frames in the genome of the Gram-negative bacterium E. coli encodes secretory proteins, and 20% encodes integral inner membrane proteins (IMPs). These proteins are sorted to their correct cellular compartments (the periplasm and the outer and inner membranes) by specialized targeting and translocation/insertion systems. So far, a very limited set of model proteins have been used to study proteins sorting requirements in E. coli. The main objective of all the papers presented in this thesis was to determine the targeting and translocation/insertion requirements of more E. coli proteins. In papers I and II, this was done using focused approaches. Selected model proteins (lipoproteins and putative outer membrane proteins) were expressed from plasmids and their targeting and translocation were analysed in vitro by crosslinking experiments and/or in vivo by pulse-chase analysis in different E. coli mutant strains. In papers III a comparative sub-proteome analysis was carried out to define the role of the cytoplasmic chaperone SecB in protein targeting. In paper IV, a similar approach was used to study how protein translocation and insertion is affected upon depletion of the essential Sec-translocon component SecE. The ‘global’ approach used in paper III and IV allowed us to study protein targeting and translocation/insertion requirements on a proteome level. This led to the identification of several novel SecB substrates and a large number of potential Sec-translocon independent IMPs.

The DrrAB proteins of Streptomyces peucetius belong to the ABC family of ubiquitous membrane transporters. The DrrA and DrrB proteins together form a drug efflux pump that carries out the transport of the anticancer drug doxorubicin by carrying out ATP hydrolysis. The present study is the first where the intrinsic factors involved in the assembly of the DrrAB functional complex have been elucidated. The drrA and drrB genes in the wild type operon have overlapping stop and start codons (ATGA) which indicates translational coupling between the two genes. On insertion of a fortuitous stop codon in DrrA it was shown that the expression of DrrB is coupled to that of the upstream gene drrA. Furthermore, it was observed that a functional complex could be achieved only when the genes were maintained in cis in a translationally coupled manner. Translational regulation in DrrA was found to be involved in the control of optimal levels of DrrB. Inhibitory interactions within drrA sequence were speculated to cause translational arrest at the C terminus of DrrA. A novel assembly domain that forms the interface between DrrA containing the Nucleotide Binding Domain (NBD) and DrrB comprising the TransMembrane Domain (TMD) was found. Based on the data presented in this study a model is proposed for the biogenesis of the DrrAB drug pump. The model suggests that translational coupling between DrrA and DrrB is crucial for functional complex formation. Further, there is evidence of regulation of translation by attenuation in the intergenic region of drrA and drrB. The regulation seems to involve the last 30 nucleotides of the mRNA of drrA and some upstream sequences within drrA that cause translational arrest within the C terminus of DrrA. Since DrrB is translationally coupled to drrA, this translational arrest in conjunction with coupling causes lowering in the levels of DrrB. Finally, since the DrrA-DrrB interaction domain lies in the C terminus of DrrA, only the fully translated DrrA product will be competent to form a complex with DrrB. This interaction between the C terminus of DrrA and the N terminus of DrrB may be crucial for initial targeting of the complex to the membrane. The model is expected to serve as primer and open up an interesting yet insufficiently understood subject of membrane protein biogenesis.

alpha-Helical membrane proteins are important for many biological functions. Due to physicochemical constraints, the structures of membrane proteins differ from the structure of soluble proteins. Historically, membrane protein structures were assumed to be more or less two-dimensional, consisting of long, straight, membrane-spanning parallel helices packed against each other. However, during the past decade, a number of the new membrane protein structures cast doubt on this notion. Today, it is evident that the structures of many membrane proteins are equally complex as for many soluble proteins. Here, we review this development and discuss the consequences for our understanding of membrane protein biogenesis, folding, evolution, and bioinformatics. / <p>authorCount :3</p>

Yes / The ribosome exit site is a focal point for the interaction of protein-biogenesis factors that guide the fate of nascent polypeptides. These factors include chaperones such as NAC, N-terminal-modifying enzymes like Methionine aminopeptidase (MetAP), and the signal recognition particle (SRP), which targets secretory and membrane proteins to the ER. These factors potentially compete with one another in the short time-window when the nascent chain first emerges at the exit site, suggesting a need for regulation. Here, we show that MetAP contacts the ribosome at the universal adaptor site where it is adjacent to the α subunit of NAC. SRP is also known to contact the ribosome at this site. In the absence of NAC, MetAP and SRP antagonize each other, indicating a novel role for NAC in regulating the access of MetAP and SRP to the ribosome. NAC also functions in SRP-dependent targeting and helps to protect substrates from aggregation before translocation. / This work was supported by grants from the BBSRC [H007202/1] and Wellcome Trust [097820/Z/11/A].