Role for ribosome-associated complex and stress-seventy subfamily B (RAC-Ssb) in integral membrane protein translation

Abstract

Targeting of most integral membrane proteins to the endoplasmic reticulum is controlled by the signal recognition particle, which recognizes a hydrophobic signal sequence near the protein N terminus. Proper folding of these proteins is monitored by the unfolded protein response and involves protein degradation pathways to ensure quality control. Here, we identify a new pathway for quality control of major facilitator superfamily transporters that occurs before the first transmembrane helix, the signal sequence recognized by the signal recognition particle, is made by the ribosome. Increased rates of translation elongation of the N-terminal sequence of these integral membrane proteins can divert the nascent protein chains to the ribosome-associated complex and stress-seventy subfamily B chaperones. We also show that quality control of integral membrane proteins by ribosome-associated complex-stress-seventy subfamily B couples translation rate to the unfolded protein response, which has implications for understanding mechanisms underlying human disease and protein production in biotechnology.

Keywords: chaperone; major facilitator superfamily (MFS); membrane protein; protein synthesis; ribosome associated complex (RAC); stress-seventy subfamily B (Ssb); translation control; unfolded protein response (UPR).

Figure 1.

Codon-optimized CDT-1 and growth phenotypes. A, schematic of the cellobiose utilization pathway in S. cerevisiae. The CDT-1 transporter imports cellobiose into the cell, and the GH1-1 β-glucosidase hydrolyzes cellobiose into two glucose monomers, which then enter the glycolytic pathway. B, %MinMax profiles for NC using S. cerevisiae and N. crassa codon usage tables and %MinMax profile for OPT using the S. cerevisiae codon usage table. C, growth curves and growth rates for D452-2 cells expressing NC or OPT and GH1-1 in medium containing cellobiose as the sole carbon source, under aerobic conditions. Growth rates were calculated from the exponential growth phase and were 0.07 h⁻¹ for cells expressing NC and 0.05 h⁻¹ for cells expressing OPT. D, growth curves and growth rates for D452-2 cells expressing NC or OPT in glucose-containing medium. Growth rates were 0.16 h⁻¹ for cells transformed with empty plasmid, 0.1 h⁻¹ for cells with pRS316-NC, and 0.065 h⁻¹ for cells with pRS316-OPT. All growth rates shown represent the mean growth rate from three biological replicates, with the corresponding S.E. value (error bars). E, cellular localization of Nc and Opt. Epifluorescent microscope images of S. cerevisiae strain D452-2 cells with Nc-GFP (Nc) or Opt-GFP (Opt) transporters.

Figure 2.

Synonymous codon changes effects on protein stability and the unfolded protein response. A, measured changes in the thermal stability of Nc via FSEC. For each temperature tested, a 4 °C incubation control was included. The first eluted peak corresponds to soluble aggregates of Nc-GFP in the void volume of the column. The second eluted peak corresponds to monomeric Nc-GFP, and the last eluting peak is free GFP. B, thermal melting curves for Nc and Opt. The fraction of folded transporter was calculated as the ratio of the peak area for the monomeric Nc/Opt-GFP peak at a given temperature to the peak area of the same monomeric peak for the control (4 °C) within the same elution range to yield a thermal melt curve. The calculated T_m values were 37.8 ± 0.16 °C for Nc and 38.2 ± 0.14 °C for Opt, with R² values of 0.97 and 0.98, respectively. C, Nc-GFP and Opt-GFP protein concentration measured via GFP fluorescence. Mean values and the corresponding S.E. values (error bars) are shown for three biological replicates. D, NC- or OPT-induced UPR activation in the D452-2 UPR reporter strain. UPR activation is shown as the median of the log₂ GFP/RFP ratio. Each bar represents the mean from four biological replicates with the corresponding S.E. Empty, cells transformed with empty plasmid as a control. The baseline for minimal activation of the UPR response was determined by using just the D452-UPR cells to measure no UPR activation ((−) DTT). The baseline for maximal UPR activation used D452-UPR cells incubated in medium containing 5 mm DTT (UPR inducer) for 2–3 h to fully induce the UPR ((+) DTT).

Figure 3.

NC-OPT chimeras. A, structure of the MFS transporter XylE from E. coli (Protein Data Bank code 4GBZ (90)) showing the four triple-helix motifs. Motifs are colored in succession from the N to the C terminus: the first motif is blue; the second is green; the third is yellow; and the fourth is red. B, growth rates for D452-2 cells expressing triple-helix chimeras. The corresponding region for each triple helix from the NC sequence was swapped into the OPT sequence. C, growth rates for D452-2 cells expressing chimeras within the first triple-helix bundle. The first triple-helix motif was divided into individual secondary structural elements, and the NC sequence was swapped into the corresponding sequence in the OPT ORF. All growth rate measurements represent the mean growth rate and S.E. (error bars) from biological triplicates.

Figure 4.

Impact of the N-terminal codons of OPT on growth phenotypes. A, growth rates of cells expressing chimeras within the first 72 codons. The first 72 codons of OPT were divided into 30-codon overlapping windows, which were replaced with the corresponding NC sequences. REG1 corresponds to codons 1–30 from NC, REG2 corresponds to NC codons 21–50, and REG3 corresponds to NC codons 41–72 (black brackets). B, growth rate of cells expressing OPT codons 1–30 replacing the NC coding sequence (REG1 NC). Mean growth rates and S.E. values (error bars) from biological triplicates are shown.

Figure 5.

Impact of transporter expression on UPR activation and growth. A, UPR activation levels in cells expressing NC, OPT, and the REG1 OPT and REG1 NC chimeras. The level of UPR activation is higher in the presence of region 1 (codons 1–30) from the OPT sequence (REG1 NC) compared with region 1 of NC fused to the remainder of OPT (REG1 OPT). Mean levels are shown for five biological replicates, with corresponding S.E. values (error bars). B, growth rates of BY4741 cells expressing HXT1 and HXT5 from plasmids. Growth rates represent the mean growth rate and S.E. from three biological replicates. C, protein concentration of plasmid-expressed Hxt1, Hxt5, and Opt. The Hxt1 levels were 3.55 ± 0.16, Hxt5 levels were 2.45 ± 0.03, and Opt levels were 1.47 ± 0.04 μg of GFP/mg of total protein. HXT1 and HXT5 were expressed from the same expression cassette as NC and OPT.

Figure 6.

NC30-GFP and OPT30-GFP chimeras. A, growth rate of D452-2 cells expressing GFP chimeras. Region 1 (codons 1–30) from NC and OPT were fused to the N terminus of GFP. NC30-GFP and OPT30-GFP were expressed from the same expression cassette as NC and OPT. Mean values of biological triplicates and S.E. values (error bars) are shown. B, protein levels in whole-cell lysates for these chimeras were 3.45 ± 0.13 μg of GFP/mg of total protein for Nc30-GFP and 7.79 ± 0.19 μg of GFP/mg total protein for Opt30-GFP. Growth rates were calculated from biological triplicates.

Figure 7.

Ribosome footprint distributions for NC and OPT mRNAs. A, mRNA RPKM levels for NC and OPT transcripts. The transcript levels for NC were 530.4 ± 18.9 RPKM; for OPT, they were 502.3 ± 41.7 RPKM, determined from biological triplicates. B, rpM counts at the A site in the ribosome for both NC and OPT transcripts (these transcripts include the eGFP fusion). The x axis represents the nt position in the coding sequence. C, rpM counts for nucleotides 1–411 in NC and OPT. Shaded boxes represent the locations of TM1 and TM2. The blue box marks the pause before TM2. D, ribosome footprint distribution for the first 90 nucleotides of the NC and OPT open reading frames. Shown are the mean rpM counts from biological triplicates with their corresponding S.E. values. E, mean OPT rpM counts subtracted from the mean NC rpM counts (ΔrpM) and corrected for the S.E. (error bars).

Figure 8.

Contributions of chaperone activities to transporter expression and growth phenotypes. A, schematic of RAC and NAC interacting with the ribosome. NAC modulates SRP substrate recognition and chaperones a subset of nascent chains. RAC recruits Ssb to interact with nascent chain substrates and links aberrant proteins at the ribosome with the cytosolic network of chaperones (91). B, growth rates for WT, Δegd1, and Δegd2 strains expressing NC and OPT. C, growth rates for WT, Δssz1, and Δzuo1 strains expressing NC and OPT. D, growth rates for WT, Δssb1, Δssb2, and Δssb1/Δssb2 expressing NC and OPT. E, Opt protein concentration in WT, Δegd1, and Δzuo1 whole-cell lysates. Protein levels were quantified via GFP fluorescence. F, growth rates for WT, Δegd1, and Δzuo1 strains expressing plasmid-borne HXT1 and HXT5. All assays represent the mean and corresponding S.E. (error bars) from three biological replicates. G, Nc, Hxt1, and Hxt5 protein concentrations in WT, Δegd1, and Δzuo1 whole-cell lysates. Protein levels were quantified via GFP fluorescence.

Figure 9.

Ribosome footprint enrichment for HXT1 from Ssb1-mediated selective ribosome profiling. A, footprint distribution over the entire HXT1 transcript. The plot shows the enrichment per codon position in footprint counts in the interactome samples over the total translatome samples for HXT1. The top schematic shows primary structure of Hxt1 with the predicted positions for the transmembrane domains (red boxes). The bottom schematic shows the length of the nascent chain in the ribosome at the peaks denoted by dashed lines. B, same plot enlarged. Light blue shading indicates variance between biological duplicates. AU, arbitrary units.