SecA mediates cotranslational targeting and translocation of an inner membrane protein

Abstract

Protein targeting to the bacterial plasma membrane was generally thought to occur via two major pathways: cotranslational targeting by signal recognition particle (SRP) and posttranslational targeting by SecA and SecB. Recently, SecA was found to also bind ribosomes near the nascent polypeptide exit tunnel, but the function of this SecA-ribosome contact remains unclear. In this study, we show that SecA cotranslationally recognizes the nascent chain of an inner membrane protein, RodZ, with high affinity and specificity. In vitro reconstitution and in vivo targeting assays show that SecA is necessary and sufficient to direct the targeting and translocation of RodZ to the bacterial plasma membrane in an obligatorily cotranslational mechanism. Sequence elements upstream and downstream of the RodZ transmembrane domain dictate nascent polypeptide selection by SecA instead of the SRP machinery. These findings identify a new route for the targeting of inner membrane proteins in bacteria and highlight the diversity of targeting pathways that enables an organism to accommodate diverse nascent proteins.

Figure 1.

Fluorescence measurements of SecA–RNC interactions. (A) Scheme of the FRET assay to detect the interaction of SecA with the RodZ nascent chain on the ribosome. (B) Fluorescence emission spectra for indicated samples. Where indicated, reactions contained 20 nM ${\mathrm{RNC}}_{\mathrm{RodZ}}^{\mathrm{Cm}},$ 40 nM SecA^BDP, and 400 nM unlabeled SecA. (C) Representative equilibrium titrations to measure the K_d values of the SecA–RNC_RodZ complex. Reactions contained 20 nM ${\mathrm{RNC}}_{\mathrm{RodZ}}^{\mathrm{Cm}}$ without (black) or with SRP (blue) or TF (green) present. The titration curves before normalization are shown in Fig. S1 B. Lines are fits of the data to Eq. 3. (D and E) Representative equilibrium titrations to measure the K_d values of the SecA–RNC_RodZ (D) and SecA–RNC_FtsQ (E) complexes at increasing concentrations of SRP. Lines are fits of the data to Eq. 3. (F) Summary of the K_d values of SecA–RNC complexes obtained from the data in C–E and their replicates. Values represent mean ± SD; n = 3.

Figure 2.

Contribution of the ribosome to RNC–SecA affinity. (A) Structure of SecA bound to the 70S ribosome (EMD-2565). The crystal structures of SecA (PDB 1m6n; orange) and ribosome (PDB 2aw4; gray) were docked into the EM density. Residues on L23 (cyan) that contact SecA are in spacefill. (B–D) Equilibrium titrations to measure the affinity of SecA for WT and modified RNC_RodZ (B) and RNC_PhoA (C) as well as the affinity of SRP for RNC_FtsQ (D). Lines are fits of the data to Eq. 3. (E) Summary of the K_d values derived from the data in B–D. Values represent mean ± SD; n = 3.

Figure 3.

Defining the sequence elements of RodZ for SecA recognition. (A) Sequences of TMD in WT RodZ and RodZ TMD mutants. Letters in blue indicate positively charged residues. (B) Equilibrium titrations to measure the affinity of SecA for RNC_RodZ bearing WT and mutant TMD sequences. The data were fit to Eq. 2 and gave K_d values of 0.94 ± 0.42 and 25.9 ± 1.1 nM for WT and TMD mut, respectively. (C) Summary of the K_d values at indicated lengths of the RodZ nascent chain (sequences in Table S1) obtained from the data in Fig. S1 E as well as their replicates. Schemes for RNC_RodZ at each chain length are shown below the graph, with ribosome in gray, RodZ TMD in brown, and sequences upstream of TMD depicted as hexagons. (D) Scheme of sequence elements in WT and mutant RodZ nascent chain used for the RNC–SecA binding measurements in E and F. MBD (purple) denotes the maltose-binding protein (residues 1–103), 6KR (blue) denotes the ¹⁰⁴KKRKRR¹⁰⁹ sequence, the RodZ TMD is in brown, and RodZ peri (red) and FtsQ peri (green) denote the early periplasmic regions of RodZ (residues 134–160) and FtsQ (residues 50–74), respectively. Mutations to acidic residues at corresponding positions of the RodZ periplasmic sequence are indicated in the Peri swap acidic construct. All the mutant constructs are derived from RodZ160 in Fig. 3 C. See Table S1 for detailed sequences. (E) Equilibrium titrations to measure the binding of SecA to RNCs bearing the WT and mutant RodZ nascent chain depicted in D. (F) Summary of the K_d values for RNCs bearing WT and mutant RodZ nascent chain obtained from the data in B and E. (G) Scheme for the competition assay to measure the binding of SUMO fusion proteins to SecA. BDP-labeled SecA was allowed to form a complex with RNC^Cm. This binding equilibrium is perturbed if the inhibitor binds SecA^BDP and traps it into a SecA ⋅ SUMO variant complex, generating free RNC^Cm and resulting in loss of FRET (i.e., increase of Cm fluorescence). (H) Competition reactions to measure the binding of SUMO and SUMO variants to SecA. SUMO, SMT3 residues 1–101; SUMO-RodZ(peri), SMT3 fused to the N terminus of RodZ periplasmic region (residues 134–160); SUMO-FtsQ(peri), SMT3 fused to the N terminus of FtsQ periplasmic region (residues 50–74). The data with SUMO-RodZ(peri) were fit to Eq. 8 and gave a K_i value of 1.2 ± 0.7 µM. In contrast, SUMO and SUMO-FtsQ(peri) did not give robust competition. Values represent mean ± SD; n = 2–3. c.p.s., counts per second.

Figure 4.

RodZ is cotranslationally targeted and translocated in vivo. (A) Scheme of the in vivo assay to distinguish between co- and posttranslational modes of targeting and translocation based on NTS-TrxA fusions. All NTS sequences are provided in Table S1. (B, left) Subcellular localization of NTS-TrxA fusion proteins. C, cytosol; M, membrane; PM, periplasm; T, total. (B, right) Assay for translocation of the C terminus of the NTS-TrxA fusion proteins into periplasm based on protection against proteinase K. K, proteinase K; T, Triton X-100. (C) Controls for cell fractionation. Mature AmpC is secreted into the periplasm (asterisk). YidC is an inner membrane protein. TrxA is a cytoplasmic protein. (D) Effects of Ffh depletion on the targeting and translocation of NTS-TrxA fusions. In vivo targeting and insertion were measured and analyzed as in B. Ffh expression is under control of the arabinose promoter. (E) Ffh is depleted in WAM121 cells grown in glucose without significantly affecting SecA abundance. (F) Translocation efficiency of NTS-TrxA constructs derived from the data in D and their replicates. Asterisks in B–D denote mature translocated secretory proteins whose signal sequences have been cleaved. Values represent mean ± SD; n = 2–3 biological replicates.

Figure 5.

Reconstitution of RodZ targeting and translocation in vitro. (A) Effect of SecA and SRP/FtsY on the translocation of indicated substrates into U-IMVs during PURE IVT. Reactions contained 400 nM Ffh, 1 µM FtsY, and 0.94 µM SecA where indicated. 4.5S RNA was included in the tRNA mix (Kuruma et al., 2005). (B) Targeting and translocation of RodZ is strictly cotranslational, whereas that of proOmpA is not. Reactions contained 0.94 µM SecA and 2.5 µM SecB where indicated. Chl, chloramphenicol. Values under each lane are quantifications of percent translocation from these data and their replicates (Fig. S3) and represent mean ± SD; n = 2–3. Asterisks denote the protected fragment after proteinase K digestion.

Figure 6.

The NTE and early periplasmic region of RodZ together dictate the selection of a membrane protein into the SecA versus SRP pathway. (A) Scheme of the sequence elements of the substrate variants tested in this figure. Detailed sequences are in Table S1. (B and C) Summary of the K_d values of RNCs bearing different nascent chains for binding to SecA (B) or SRP (C) derived from the equilibrium titrations in Fig. S4. All titrations contained 20 nM RNCs and 2 µM TF, 400 nM SRP, or 2 µM SecA where indicated. (D and E) In vitro translocation assays of WT RodZ or mutant RodZ^ΔNTE and their dependence on SecA (D) or SRP (E). (F and G) In vitro translocation assays of WT FtsQ and mutants RodZNTE-FtsQ and RodZNTE-peri-FtsQ. The dependence of the reaction on SecA was shown in F, and the dependence on SRP was shown in G. The reactions in D and F contained 3.8 µM TF, 400 nM Ffh, 1 µM FtsY, and indicated concentrations of SecA. The reactions in E and G contained 50 nM SecA, 3.8 µM TF, the indicated concentrations of SRP, and a fivefold excess of FtsY over SRP. Values represent mean ± SD; n = 2–3.

Figure 7.

Diverse targeting pathways deliver nascent proteins to the SecYEG translocon at the inner membrane. Left path, proteins with weakly hydrophobic signal sequences are maintained soluble by SecB and targeted to membrane via interaction with SecA, which translocates the nascent polypeptide across SecYEG. Right path, proteins containing hydrophobic TMDs or signal sequences are cotranslationally recognized by SRP and targeted to SecYEG via the SRP/SR interaction. Middle path, proteins harboring internal TMDs are cotranslationally recognized and targeted by SecA.