A Signal Peptide That Directs Non-Sec Transport in Bacteria Also Directs Efficient and Exclusive Transport on the Thylakoid Delta pH Pathway

Abstract

Signal peptides that specifically direct precursor proteins to the thylakoid Delta pH pathway possess an N domain RR motif. Signal peptides that direct transport of bacterial proteins across a non-Sec export pathway possess an N domain RRXFLK consensus motif. Recent genetic studies suggest an evolutionary link between these two protein translocation pathways. To further explore this relationship, we examined the thylakoid targeting capability of the signal peptide for Escherichia colihydrogenase 1 small subunit (HyaA) by linking it to plastocyanin and assaying the chimeric protein in an in vitro thylakoid transport assay. The chimeric precursor was transported across thylakoids with high efficiency. Transport was characteristic of the Delta pH but not the Sec pathway, i.e. it was eliminated by ionophores that dissipate the ΔpH but occurred in the absence of stromal extract or ATP. This result was confirmed by competition with chemical quantities of a Delta pH pathway precursor. This indicates that the HyaA signal peptide has the necessary elements for efficient and exclusive targeting to the Delta pH pathway and further supports the notion that the alternate targeting pathways in prokaryotes and plant thylakoids are analogous. Signal peptides that specifically direct precursor proteins to the thylakoid Delta pH pathway possess an N domain RR motif. Signal peptides that direct transport of bacterial proteins across a non-Sec export pathway possess an N domain RRXFLK consensus motif. Recent genetic studies suggest an evolutionary link between these two protein translocation pathways. To further explore this relationship, we examined the thylakoid targeting capability of the signal peptide for Escherichia colihydrogenase 1 small subunit (HyaA) by linking it to plastocyanin and assaying the chimeric protein in an in vitro thylakoid transport assay. The chimeric precursor was transported across thylakoids with high efficiency. Transport was characteristic of the Delta pH but not the Sec pathway, i.e. it was eliminated by ionophores that dissipate the ΔpH but occurred in the absence of stromal extract or ATP. This result was confirmed by competition with chemical quantities of a Delta pH pathway precursor. This indicates that the HyaA signal peptide has the necessary elements for efficient and exclusive targeting to the Delta pH pathway and further supports the notion that the alternate targeting pathways in prokaryotes and plant thylakoids are analogous. Many thylakoid lumen-resident proteins of plant chloroplasts are synthesized in the cytosol as larger precursors with bipartite amino-terminal extensions called transit peptides (see Ref. 1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar for review). The stroma-targeting domain of the transit peptide governs import into the chloroplast stroma; the lumen-targeting domain directs subsequent transport into the thylakoid lumen. Two precursor-specific pathways for protein transport into the thylakoid lumen have been identified by in vitro and genetic studies (1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar). The thylakoid Sec pathway requires a chloroplast SecA protein (cpSecA) and ATP (1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar) and appears analogous to the bacterial Sec system. The Delta pH pathway operates independently of ATP and soluble factors, requiring only a thylakoidal pH gradient (1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar). Targeting specificity for the two pathways is determined primarily by the lumen-targeting domains, which contain motifs of bacterial signal peptides, i.e. an amino-terminal charged N domain, a hydrophobic H domain, and a carboxyl-terminal cleavage C domain (2von Heijne G. Steppuhn J. Herrmann R.G. Eur. J. Biochem. 1989; 180: 535-545Crossref PubMed Scopus (910) Google Scholar). Precursors targeted to the Delta pH pathway invariably contain an essential N domain twin arginine that provides access to the Delta pH pathway (Fig.1) (3Chaddock A.M. Mant A. Karnauchov I. Brink S. Herrmann R.G. Klösgen R.B. Robinson C. EMBO J. 1995; 14: 2715-2722Crossref PubMed Scopus (226) Google Scholar). In addition, Delta pH pathway precursors have H and/or C domains that are nonfunctional for Sec pathway transport (4Bogsch E. Brink S. Robinson C. EMBO J. 1997; 16: 3851-3859Crossref PubMed Scopus (121) Google Scholar, 5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar). These latter elements have been termed “Sec-avoidance” elements (4Bogsch E. Brink S. Robinson C. EMBO J. 1997; 16: 3851-3859Crossref PubMed Scopus (121) Google Scholar). A non-Sec protein export pathway also appears to operate in bacteria. Nivière et al. (6Nivière V. Wong S.-L. Voordouw G. J. Gen. Microbiol. 1992; 138: 2173-2183Crossref PubMed Scopus (71) Google Scholar) showed that a chimeric precursor containing the Desulfovibrio vulgaris hydrogenase small subunit signal peptide fused to β-lactamase was exported efficiently only under anaerobic conditions and this export depended upon a critical N domain twin arginine motif. Export of Pseudomonas stutzeri nitrous oxide reductase also depends upon an N domain RR (7Dreusch A. Bürgisser D.M. Heizmann C.W. Zumft W.G. Biochim. Biophys. Acta. 1997; 1319: 311-318Crossref PubMed Scopus (47) Google Scholar). Recently, it was shown that trimethylamine N-oxide reductase, which bears an N domain twin arginine, is exported by a mechanism independent of SecA, SecY, or SecE, but dependent on the transmembrane ΔμH+ (8Santini C.-L Ize B. Chanal A. Müller M. Giordano G. Wu L.-F. EMBO J. 1998; 17: 101-112Crossref PubMed Scopus (290) Google Scholar). Berks (9Berks B.C. Mol. Microbiol. 1996; 22: 393-404Crossref PubMed Scopus (561) Google Scholar) pointed out that many precursors for bacterial proteins that bind redox cofactors share a conserved N domain (S/T)RRXFLK motif (Fig.1) and suggested that the export system for such proteins may be related to the thylakoid Delta pH pathway. Strong support for such a notion was recently provided by Settles et al. (10Settles A.M. Yonetani A. Baron A. Bush D.R. Cline K. Martinssen R. Science. 1997; 278: 1467-1470Crossref PubMed Scopus (233) Google Scholar). Maize Hcf106 mutant chloroplasts are selectively defective in the Delta pH pathway. The Hcf106 protein has striking homology to several open reading frames from bacterial genomes. In the case of Azotobacter chroococcum, mutation of the Hcf106 homologue results in mislocalization of hydrogenase (10Settles A.M. Yonetani A. Baron A. Bush D.R. Cline K. Martinssen R. Science. 1997; 278: 1467-1470Crossref PubMed Scopus (233) Google Scholar). Here we show that the signal peptide of Escherichia colihydrogenase 1 small subunit (HyaA) is functionally equivalent to a lumen-targeting domain for chloroplast Delta pH pathway precursors. This indicates that the HyaA signal peptide has the essential elements that both engage the Delta pH pathway and avoid the Sec pathway. All reagents, enzymes, and standards were purchased commercially. In vitro transcription plasmids for the stromal intermediate of OE33 1The abbreviations used are: OE23 and OE33, the 23- and 33-kDa subunits of the photosystem II oxygen evolving complex, respectively; p, i, and t, full-length precursor, intermediate precursor, and truncated precursor form, respectively; PC, plastocyanin; HyaA, E. coli hydrogenase 1 small subunit; DT, dual-targeting; LHCP, the light-harvesting chlorophylla/b protein; SOE, splicing by overlap extension; PCR, polymerase chain reaction. (iOE33) from wheat (11Hulford A. Hazell L. Mould R.M. Robinson C. J. Biol. Chem. 1994; 269: 3251-3256Abstract Full Text PDF PubMed Google Scholar), precursors to plastocyanin (pPC) from Arabidopsis and LHCP (12Cline K. Henry R. Li C. Yuan J. EMBO J. 1993; 12: 4105-4114Crossref PubMed Scopus (172) Google Scholar), and the chimeric precursors t23-PC and DT-PC (5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar) were previously described. For tOE23 expression in E. coli, the coding sequence was amplified from the transcription plasmid with a forward primer that contained an NdeI site encompassing the initiator methionine codon and a reverse primer that also contained aHindIII restriction site. The PCR product was cloned into the NdeI/HindIII sites of pETH3c (12Cline K. Henry R. Li C. Yuan J. EMBO J. 1993; 12: 4105-4114Crossref PubMed Scopus (172) Google Scholar). Expression of tOE23 in E. coli strain BL21 (DE3) and isolation of inclusion bodies were as described (12Cline K. Henry R. Li C. Yuan J. EMBO J. 1993; 12: 4105-4114Crossref PubMed Scopus (172) Google Scholar). The HyaA coding sequence (13Menon N.K. Robbins J. Wendt J.C. Shanmugam K.T. Przybyla A.E. J. Bacteriol. 1991; 173: 4851-4861Crossref PubMed Scopus (124) Google Scholar) was amplified by PCR with E. coli genomic DNA as template and cloned into pGEM-3z. PCR splicing by overlap extension (SOE) (14Horton R.M. Hunt H.D. Ho S.N. Pullen J.K. Pease L.R. Gene. 1989; 77: 61-68Crossref PubMed Scopus (2641) Google Scholar) was used to construct a chimeric precursor, Hya-PC, which is an exact fusion between coding sequences for the HyaA signal peptide and the mature domain of Arabidopsis PC. DNA fragments corresponding to the signal peptide and to PC were amplified separately and spliced in a second round of PCR. Forward and reverse primers for the SOE reaction contained restrictions sites for HindIII andSstI sites, respectively, and the SOE product was cloned into the HindIII and SstI sites of pGEM-3z. The sequences of all PCR-cloned constructs were confirmed by TaqDyeDeoxy Terminator cycle sequencing. Capped RNA for the various precursors was produced in vitro with SP6 polymerase and uncut plasmid. Precursors were translated in a wheat germ system in the presence of [3H]leucine and adjusted to import buffer (50 mm HEPES-KOH, pH 8.0, 0.33 m sorbitol) containing 30 mm unlabeled leucine prior to use (12Cline K. Henry R. Li C. Yuan J. EMBO J. 1993; 12: 4105-4114Crossref PubMed Scopus (172) Google Scholar). Chloroplasts and thylakoids were prepared from pea seedlings as described (15Cline K. J. Biol. Chem. 1986; 261: 14804-14810Abstract Full Text PDF PubMed Google Scholar). Transport of radiolabeled proteins into thylakoids was conducted with chloroplast lysates or washed thylakoids in 75-μl assays (12Cline K. Henry R. Li C. Yuan J. EMBO J. 1993; 12: 4105-4114Crossref PubMed Scopus (172) Google Scholar). Precursors and recovered thylakoid membranes were analyzed by SDS-polyacrylamide gel electrophoresis and fluorography. Quantification was accomplished by scintillation counting of radiolabeled proteins extracted from excised gel bands (15Cline K. J. Biol. Chem. 1986; 261: 14804-14810Abstract Full Text PDF PubMed Google Scholar). We initially assayed transport of the HyaA precursor protein into isolated thylakoids. A low level of transport was achieved as determined by expected proteolytic maturation and protection from exogenous protease (data not shown). However, this low level of transport was insufficient for a full examination of the targeting properties. Because the focus here was on the targeting capability of the signal peptide of HyaA, we constructed a chimeric precursor protein (Hya-PC) possessing the HyaA signal peptide fused to the mature domain of Arabidopsis PC. PC was chosen as a passenger protein because it can be transported on the Sec pathway and the Delta pH pathway when linked to appropriate signal peptides (4Bogsch E. Brink S. Robinson C. EMBO J. 1997; 16: 3851-3859Crossref PubMed Scopus (121) Google Scholar, 5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar). Fig. 2 shows a thylakoid transport assay with Hya-PC. Incubation of Hya-PC with isolated thylakoids produced a smaller product at the location of mature PC (lane 2) that was resistant to thermolysin treatment of the membranes (lane 3). Mature PC was recovered in the lumen subfraction when the recovered thylakoids were sonicated to release the lumenal contents (lane 7). Mature PC was not produced when assays were conducted in the presence of ionophores (lane 4), and the membrane-associated precursor was degraded by thermolysin (lane 5). Fig. 2 also shows control assays for translocation/integration of authentic pPC and the membrane protein LHCP. These assays demonstrate that the HyaA signal peptide directs transport into the thylakoid lumen. Average transport of Hya-PC for three experiments was 32% of the added precursor. This compares very favorably with the 16% of t23-PC transport (Fig. 3, lane 3) for the same three experiments. t23-PC is a fusion protein between the core targeting peptide for the Delta pH substrate OE23 and the mature domain of Arabidopsis PC (5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar).Figure 3Transport requirements suggest that Hya-PC is transported by the Delta pH pathway. Transport of precursors across thylakoid membranes was conducted for 30 min at 25 °C with chloroplast lysate to provide stromal extract (SE) or washed thylakoids. Energy was provided in the form of ATP and/or light to generate a ΔpH. Assay conditions examined the requirement for a ΔpH (lane 4), ATP (lane 5), stromal extract (lane 6), and sensitivity to azide (lane 7). Final concentrations were 5 mm ATP, 10 mmsodium azide, 0.5 μm nigericin (nig), and 1 μm valinomycin (val). Apyrase (0.5 unit per 75 μl) was used to eliminate residual ATP in lysate and translation products for assays conducted in the absence of ATP. These conditions are designated above the panel. Recovered thylakoids were post-treated with thermolysin. Analysis and designations were as in Fig. 2.View Large Image Figure ViewerDownload (PPT) Assays in Fig. 3 were conducted under a variety of conditions designed to assess energy and stroma requirements that are characteristic for thylakoid translocation pathways. As with t23-PC (a Delta pH pathway substrate), transport of Hya-PC was completely abolished by addition of ionophores that dissipate the thylakoidal ΔpH (lane 4), but was unaffected by the addition of sodium azide (lane 7), a SecA inhibitor (16Oliver D.B. Cabelli R.J. Dolan K.M. Jarosik G.P. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 8227-8231Crossref PubMed Scopus (253) Google Scholar), or by removal of the stromal extract (lane 6). Depletion of ATP with apyrase in these experiments diminished, but did not abolish, Hya-PC transport (lane 5). A similar reduction of t23-PC in the presence of apyrase also occurred. Such an effect was not previously recognized for t23-PC (5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar) but appears to be related to the PC mature domain because in parallel assays, tOE23 transport was not reduced by apyrase (data not shown). In contrast, transport of the Sec pathway substrate pPC was only marginally affected by ionophores (lane 4), virtually eliminated by removal of ATP or of stromal extract, the source of ∼90% of the cpSecA (lanes 5 and 6), and inhibited by azide (lane 7). These requirements suggest that Hya-PC is transported on the Delta pH pathway rather than the Sec pathway. In addition, Hya-PC transport requirements rule out the participation of two other pathways that are responsible for insertion of membrane proteins (1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar); i.e. membrane integration by the chloroplast SRP is absolutely dependent on the presence of stroma and NTPs, whereas membrane integration by a spontaneous mechanism occurs even in the absence of a thylakoidal ΔpH. To further clarify the pathway utilized by Hya-PC, competition assays were conducted with bacterially synthesized tOE23 (Fig. 4 A). Increasing concentrations of tOE23 progressively competed transport of t23-PC and Hya-PC. At 2 μm tOE23, transport of Hya-PC was reduced to ∼5% of that achieved in the absence of competitor. In contrast, transport of Sec pathway substrates pPC and iOE33 was unaffected by tOE23 competitor. To determine if the residual Hya-PC transport in the presence of 2 μm tOE23 competitor was mediated by the Sec pathway, the effect of ATP and stromal extract were assessed in the presence of 2 μm tOE23 (Fig. 4 B). DT-PC transport served as a positive control for this experiment. DT-PC is a chimeric precursor protein that contains the N domain of the OE23 precursor and the H/C and mature domains of PC. Importantly, DT-PC is transported by both the Sec and the Delta pH pathways (5Henry R. Carrigan M. McCaffery M. Ma X. Cline K. J. Cell Biol. 1997; 136: 823-832Crossref PubMed Scopus (71) Google Scholar). In the absence of stromal extract, tOE23 competitor substantially reduced transport of Hya-PC and DT-PC (lanes 4–7). As expected, addition of stromal extract greatly stimulated the DT-PC residual transport (lane 8), and this stimulation was largely eliminated by removal of ATP (lane 9) or inclusion of sodium azide (lane 10). These three effects are characteristic of Sec transport as exemplified by the control assays with iOE33. In contrast to these results, the residual Hya-PC transport was unaffected by the addition of stroma extract, the removal of ATP, or the inclusion of azide. This demonstrates that Hya-PC is not transported by the cpSecA-dependent mechanism even when the transport on the Delta pH pathway is virtually eliminated by competition. Furthermore, these results indicate that the reduction of transport of Hya-PC and t23-PC by apyrase seen in Fig. 3 does not reflect the involvement of the cpSecA translocation ATPase. Taken together, these results indicate that Hya-PC is exclusively targeted to the Delta pH pathway and does not utilize the Sec pathway (or other thylakoidal pathways). Thus, the signal peptide for HyaA is functionally equivalent to the Delta pH pathway lumen-targeting domain and contains the two essential elements required for exclusive targeting, an N domain twin arginine (3Chaddock A.M. Mant A. Karnauchov I. Brink S. Herrmann R.G. Klösgen R.B. Robinson C. EMBO J. 1995; 14: 2715-2722Crossref PubMed Scopus (226) Google Scholar) that provides access to the Delta pH pathway and an element that prevents engagement by the Sec pathway. It was uncertain whether the N domain twin arginine in the HyaA signal peptide would be functional for the Delta pH pathway because it is followed by several amino acids, SFLK, prior to the H domain. For most Delta pH pathway precursors, the twin arginine immediately precedes the first amino acid of the H domain. Furthermore, a chimeric construct that placed an asparagine between the twin arginine motif and the H domain was nonfunctional for Delta pH pathway transport (4Bogsch E. Brink S. Robinson C. EMBO J. 1997; 16: 3851-3859Crossref PubMed Scopus (121) Google Scholar). However, the HyaA signal peptide was twice as efficient as the authentic core signal peptide for Delta pH pathway precursor OE23. It is likely, but remains to be demonstrated, that the thylakoid Sec avoidance element in the HyaA signal peptide resides in the H/C domain. Because of the endosymbiotic origin of chloroplasts from an ancestral cyanobacterium, it was speculated that protein transport into plant thylakoids would be evolutionarily related to prokaryotic transport mechanisms. Work during the past several years has shown that three of the four known mechanisms of thylakoid protein translocation can be ascribed to prokaryote mechanisms. These include a SecA-dependent pathway, an SRP-like pathway for insertion of a membrane protein, and a spontaneous protein insertion mechanism (1Cline K. Henry R. Annu. Rev. Cell Dev. Biol. 1996; 12: 1-26Crossref PubMed Scopus (187) Google Scholar). Although characteristics of the thylakoid Delta pH pathway suggest a prokaryote-like mechanism, i.e. a classical signal peptide and initiation via a loop mechanism (17Fincher V. McCaffery M. Cline K. FEBS Lett. 1998; 423: 66-70Crossref PubMed Scopus (34) Google Scholar), no analogous system had been identified in bacteria. The recent report of the identity of the Hcf106 protein involved in Delta pH pathway, and the existence of the bacterial homologues (10Settles A.M. Yonetani A. Baron A. Bush D.R. Cline K. Martinssen R. Science. 1997; 278: 1467-1470Crossref PubMed Scopus (233) Google Scholar) now suggests the identity of the bacterial counterpart. Our results further support the notion that the pathway associated with export of redox cofactor-binding proteins in bacteria is related to the Delta pH pathway of chloroplast thylakoids. We thank Shan Wu for excellent technical assistance.

Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.