Selecting an NTS as the soluble carrier protein for the fusion construct has dual interest: increasing the solubility of the protein of interest hence reducing the risk of forming inclusion bodies and improving Tat export. To date, there are 28 known E. coli NTS, of which 19 have been somewhat characterized experimentally (such as TorA and DmsA), whereas little is known of the others, apart from the signal peptide sequence and its specificity, the molecular weight of NTS and information gleaned from analysis of its genetic organization in silico (section 1.3.2.4., Appendix 1) (Tullman-Ercek et al., 2007, Mejean et al., 1994, Sambasivarao et al., 1990). The largest Tat-exported complex described was the FdnGH subcomplex of the formate dehydrogenase-N which is an NTS of 142 kDa (Sargent et al., 2002). Although there is no report defining the maximum protein size limit of the Tat translocation capability, the translocon is thought to have an export size limit (Berks et al., 2000). Therefore, a small NTS would be preferable as a soluble partner, as this would allow a much larger protein of interest to be used. Of the 28 NTS, 12 NTS are within the smallest size range (< 41 kDa), therefore NTS larger than this were eliminated from this panel (Appendix 1).
Another criterium for selection was grounded on whether the NTS with their natural signal peptides were obligates for Tat-export. Signal peptide-MBP reporter fusions in a Tat knock-out strain have shown that some Tat signal peptides (TatSP) led to promiscuous export into the periplasm via Sec
(Tullman-Ercek et al., 2007). Subsequently NTS that failed to meet this specificity according to available published data, were eliminated. Of the 5 NTS remaining, 2 appear to have partners and are therefore not monomeric and the oligomeric state of 2 others are unknown. The potential of expressing additional partners is replete with issues such as complicated vector design and larger
Experimental design strategy
53 plasmid size, increased carrier protein size, subsequent overexpression, folding and association of the carrier complex, and increased metabolic drain. Ideally the NTS candidate should be monomeric, yet of the remaining panel only AmiA (31.4 kDa) is reported to be monomeric (Table1); the oligomeric state of YcbK (20.4 kDa) although small, is not known and TorA (94.5 kDa), although monomeric, is larger in size. TorA has been used in many published studies and proved to be Tat-specific (Santini et al., 1998, Weiner et al., 1998) and its signal peptide (TorASP) is the most widely used in Tat export
studies (section 3.1.); it may prove to be a useful positive control. To increase the candidate shortlist to five, a heterodimer HyaA (40.7 kDa) and a heterotrimer PaoA (24.3 kDa) subunits were selected to investigate the potential for Tat export of these proteins (Table 3-1). Initially, these proteins will be expressed without their respective partners to gauge whether they alone can be used as a fusion partner.
Table 3-1: Selected NTS candidates
The panel of NTS was selected based on the criteria of small size, monomeric state and specific to Tat export. The table indicates the short protein name, molecular weight (kDa), the number of cysteine (#Cys), the number of disulphide bond (#DSB), the cofactor(s), the oligomeric state in which the said protein is transported via the Tat pathway and the required chaperone. N/D non-determined. N/A non- applicable.
Name Size
(kDa) #Cys #DSB Cofactor Chaperone
Oligomeric state during Tat export complex size (kDa) Reference
AmiA 31.4 0 0 Zn GroEL Monomer N/A (Ize et al.,
2003)
YcbK 20.4 0 0 N/D N/D N/D N/D
TorA 94.5 10 0 Mo TorD Monomer N/A
(Pommier et al., 1998)
HyaA 40.7 14 0 [2Fe-2S]3 HyaE
Putative heterodimer with HyaB 102 (Dubini and Sargent, 2003)
PaoA 24.3 9 0 [2Fe-2S]2 PaoD
Heterotrimer with PaoB and
PaoC
135
(Neumann et al., 2009)
AmiA is a periplasmic amidase and functions as a cell wall hydrolase (Ize et al., 2003). This protein is non-essential, although the amiA gene deletion results in a morphology defect where cells form
Experimental design strategy
54 growing chains and highlights incomplete septum cleavage during cell division (Heidrich et al., 2001). The amiA gene is the first gene of the amiA-hemF operon (Troup et al., 1994). The HemF protein is the oxygen-dependent coproporphyrinogen-III oxidase, an enzyme involved in heme biosynthesis in aerobic metabolism. The ypeA gene is located upstream from amiA and is transcribed in the opposite direction meaning that this operon does not contain more upstream genes.
YcbK is an uncharacterised protein where the signal peptide has been experimentally tested as an MBP fusion for Tat export (Tullman-Ercek et al., 2007). YcbK is encoded as a lone gene and is flanked by the downstream gene gloC encoding a cytoplasmic hydroxyacylglutathione hydrolase, presumably involved in methyglyoxal detoxification whereas the upstream gene ycbB, encodes for a transpeptidase involved in cell wall biosynthesis (Magnet et al., 2008).
TorA is the trimethylamine N-oxide (TMAO) reductase involved in the electron chain transport in anaerobia in E. coli by reducing the TMAO as the terminal electron acceptor (Mejean et al., 1994, Jourlin et al., 1996). This periplasmic enzyme receives electron from TorC, a pentahemic c-type cytochrome (Gon et al., 2001) and requires the cytoplasmic chaperone TorD for correct maturation (Pommier et al., 1998). The TorCAD proteins characterised as the active members of the trimethylamine N-oxide reductase system are encoded by the torCAD operon (Mejean et al., 1994). The torT and torS genes are located upstream from this operon and their respective proteins are responsible for the transcriptional regulation of torCAD (Jourlin et al., 1996).
HyaA (40.7 kDa) is the small subunit of the hydrogenase 1 complex which also contains the large subunit HyaB (66 kDa) (Forzi and Sawers, 2007). HyaA is an inner membrane bound protein facing the periplasm (Sawers and Boxer, 1986, Forzi and Sawers, 2007). It therefore contains a TM domain, suggesting that expression may lead to incorporation into the inner membrane on the periplasmic side. Although no direct evidence has been published, the 102 kDa hydrogenase 1 is believed to be Tat exported and uses the TatSP of HyaA to hitchhike the leaderless HyaB (Rodrigue et al., 1999). The
hyaA gene belongs to the hyaABCDEF operon (Menon et al., 1990). The HyaCDEF proteins encoded
Experimental design strategy
55 PaoA is an iron-sulphur binding subunit that forms a heterotrimer with PaoB (34 kDa) and PaoC (78 kDa) (Lee et al., 2014). The 135 kDa complex functions as a molybdenum-dependent aldehyde oxidoreductase and presumably plays a role in detoxification to prevent cell damage (Neumann et al., 2009). The subunits PaoB and PaoC are both devoid of export signals and are translocated into the periplasm in a piggyback fashion using the TatSP of PaoA (Lee et al., 2014). Both PaoB and PaoC
partners are thought to be essential for Tat export. A cofactor protein PaoD (35 kDa) is also thought to be involved, improving complex stability and a requisite for its activity. The complex is genetically encoded by the paoABCD operon (previously known as yagTSRQ) (Neumann et al., 2009).