JB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Other Versions of this Article:
JB.01366-07v1
190/7/2458    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Google Scholar
Right arrow Articles by Dong, Y.
Right arrow Articles by Brady, L. J.
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dong, Y.
Right arrow Articles by Brady, L. J.

 Previous Article  |  Next Article 

Journal of Bacteriology, April 2008, p. 2458-2469, Vol. 190, No. 7
0021-9193/08/$08.00+0     doi:10.1128/JB.01366-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Functional Overlap but Lack of Complete Cross-Complementation of Streptococcus mutans and Escherichia coli YidC Orthologs{triangledown}

Yuxia Dong,2,{dagger} Sara R. Palmer,1,{dagger} Adnan Hasona,1 Shushi Nagamori,3 H. Ronald Kaback,3 Ross E. Dalbey,2* and L. Jeannine Brady1*

Department of Oral Biology, University of Florida, P.O. Box 100424, Gainesville, Florida 32610,1 Department of Chemistry, The Ohio State University, Columbus, Ohio 43210,2 Department of Physiology, Department of Microbiology, Immunology and Medical Genetics, Molecular Biology Institute, University of California, Los Angeles, California 900953

Received 21 August 2007/ Accepted 23 December 2007


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Oxa/YidC/Alb family proteins are chaperones involved in membrane protein insertion and assembly. Streptococcus mutans has two YidC paralogs. Elimination of yidC2, but not yidC1, results in stress sensitivity with decreased membrane-associated F1Fo ATPase activity and an inability to initiate growth at low pH or high salt concentrations (A. Hasona, P. J. Crowley, C. M. Levesque, R. W. Mair, D. G. Cvitkovitch, A. S. Bleiweis, and L. J. Brady, Proc. Natl. Acad. Sci. USA 102:17466-17471, 2005). We now show that Escherichia coli YidC complements for acid tolerance, and partially for salt tolerance, in S. mutans lacking yidC2 and that S. mutans YidC1 or YidC2 complements growth in liquid medium, restores the proton motive force, and functions to assemble the F1Fo ATPase in a previously engineered E. coli YidC depletion strain (J. C. Samuelson, M. Chen, F. Jiang, I. Moller, M. Wiedmann, A. Kuhn, G. J. Phillips, and R. E. Dalbey, Nature 406:637-641, 2000). Both YidC1 and YidC2 also promote membrane insertion of known YidC substrates in E. coli; however, complete membrane integrity is not fully replicated, as evidenced by induction of phage shock protein A. While both function to rescue E. coli growth in broth, a different result is observed on agar plates: growth of the YidC depletion strain is largely restored by 247YidC2, a hybrid S. mutans YidC2 fused to the YidC targeting region, but not by a similar chimera, 247YidC1, nor by YidC1 or YidC2. Simultaneous expression of YidC1 and YidC2 improves complementation on plates. This study demonstrates functional redundancy between YidC orthologs in gram-negative and gram-positive organisms but also highlights differences in their activity depending on growth conditions and species background, suggesting that the complete functional spectrum of each is optimized for the specific bacteria and environment in which they reside.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The gram-positive oral pathogen Streptococcus mutans contains two paralogs of the evolutionarily conserved Oxa/YidC/Alb family of membrane-associated chaperones that mediate and contribute to the insertion and folding of membrane proteins and multisubunit complexes. Both are expressed in S. mutans, and removal of yidC2, but not yidC1, results in a stress-sensitive phenotype (acid, osmotic, and oxidative stresses) similar to that observed following the nonlethal disruption of the signal recognition particle (SRP) cotranslational protein translocation pathway in this organism (14). Elimination of yidC1 has a far less readily apparent phenotype and does not affect S. mutans growth nearly to the extent of yidC2 removal under stress or nonstress conditions, although like those without yidC2, cells lacking yidC1 are impaired in biofilm formation (36). S. mutans is a major causative agent of human dental caries, which results from its ability to metabolize a wide range of dietary carbohydrates, leading to the production of lactic acid and erosion of the tooth enamel. Acid tolerance in S. mutans is mediated in large part by an F1Fo ATPase proton pump (4), and removal of genes encoding either SRP pathway components or YidC2 results in a significant decrease in but not complete elimination of membrane-associated ATPase activity (14), partially explaining the acid sensitivity of these mutants.

In addition to bacteria, YidC family members are found in mitochondria (Oxa) and chloroplasts (Alb) (26, 49). Indeed, the initial discovery of YidC members and their role in membrane biogenesis came from studies in mitochondria (for a review, see reference 42). The mitochondrial YidC ortholog Oxa1 was shown to be required for the incorporation of the F1Fo ATPase and cytochrome bo oxidase into the inner membranes of mitochondria (1, 2, 6, 7, 23). Later studies revealed that Oxa1 was critical for the insertion of subunit II of cytochrome c oxidase and an artificial F1Fo ATPase Su9 derivative (Su9 is homologous to subunit c of the bacterial F1Fo ATPase) into the mitochondrial inner membrane (16-18). In chloroplasts, Alb3 was shown to be critical for the membrane integration of light-harvesting chlorophyll-binding protein into thylakoidal membranes (29). In Escherichia coli, the sole YidC ortholog is known to participate in two membrane protein integration pathways. In the YidC-only pathway, YidC catalyzes the insertion of the Sec-independent proteins, such as M13 procoat, Pf3 coat protein and subunit c of the F1Fo ATP synthase (10, 38, 40, 45, 47, 51). In this membrane insertion process, YidC makes contact with the hydrophobic region of the membrane protein substrate (10, 45). For insertion of Sec-dependent proteins, the role of YidC has not been fully defined. However, YidC is required for membrane insertion of several Sec-dependent proteins, including subunit a of the F1Fo ATPase and certain M13 phage procoat derivatives (12, 50). YidC also mediates the membrane insertion/integration process of Sec-dependent proteins, as cross-linking studies have shown that the hydrophobic domain of leader peptidase, FtsQ, and mannitol permease interacts with YidC (20, 38, 39). In addition, YidC has been shown to function as an assembly site for hydrophobic regions of multispanning membrane proteins (3) and to have a chaperone function for the Lac permease (30).

Conservation of function of YidC family members in organelles and E. coli has been illustrated by studies using heterologous systems. The chloroplast protein Alb3 can substitute for YidC in E. coli and promote membrane insertion of several Sec-independent proteins in the bacterial system (22). In addition, mitochondrial Oxa1 can functionally replace YidC in E. coli (33), and YidC variants can substitute for Oxa1 in the membrane biogenesis of mitochondrial inner membrane proteins (46). In light of its participation in the insertion of membrane integral F1Fo ATPase subunits, the current study was undertaken to evaluate the ability of E. coli YidC to restore stress tolerance, particularly acid tolerance, to S. mutans lacking YidC2. Likewise, the functional activities of S. mutans YidC1 and YidC2 were evaluated in E. coli, the bacterium in which YidC has been most extensively characterized and in which its elimination is lethal (38). Not surprisingly, considering the reported examples of Oxa/YidC/Alb heterologous complementation, functional overlap was observed between the family members in gram-positive and gram-negative organisms when they were expressed in S. mutans and E. coli backgrounds. However, several differences in relative complementation abilities among the bacterial proteins were also noted, suggesting that the full range of functional activities of each is optimized within the homologous organism.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Materials. Amino acids, M9 minimal salts and lysozyme were obtained from Sigma. Proteinase K and a plasmid miniprep kit was from Qiagen. Anti-hemagglutinin, anti-GroEL antiserum was from Sigma. Anti-YidCf (against full-length YidC), anti-YidCc (against the C-terminal peptide), anti-leader peptidase (anti-Lep), and anti-OmpA antiserum were from Ross Dalbey's laboratory collection. Anti-YidC1c and anti-YidC2c (against C-terminal peptides) were from Jeannine Brady's laboratory. Anti-PspA was a gift from Jan Tommassen at Utrecht University. BD Talon metal affinity resin was purchased from BD Bioscience. The enhanced chemiluminescence Western blotting kit was from Amersham Bioscience.

Strains and plasmids. The E. coli YidC depletion strain JS7131, plasmid pMS119, and pACYC184 were from Ross Dalbey's laboratory stock. pMS119, which contains the tac promoter and lacIq, was used to overproduce substrates of YidC. pMS119-a-P2, pMS119-c-10h-tag, pMS119-CyoA-N-P2, and pMS119-PClep were constructed as described previously (9, 11, 50, 51). The plasmid vector pACYC184 was used to express E. coli yidC or S. mutans yidC1 and yidC2 in E. coli. Previously, yidC along with its upstream region, including the promoter and ribosome-binding site, was cloned into pACYC184 (22). S. mutans strain NG8 was a gift from Ken Knox at the University of Sydney. S. mutans strains AH374 ({Delta}yidC1) and AH398 ({Delta}yidC2) were generated previously in the Brady laboratory (14). The shuttle vector pDL289 (8) was used to express E. coli yidC in S. mutans. MC1060 was obtained from the E. coli Genetic Resources Center (Yale University) and is the wild-type parent strain of JS7131. 1100{Delta}BC was a gift from Brian Cain at the University of Florida and is devoid of functional F1Fo ATP synthase (41).

Construction of chimeric E. coli yidC and introduction into S. mutans. All plasmid preparations and gel extractions were completed by using Qiagen QIAquick kits. Standard molecular cloning was preformed using enzymes and buffers from New England Biolabs. E. coli yidC with a deletion of DNA encoding the N-terminal 58 amino acids was PCR amplified from pACYC184-YidC plasmid DNA. Vent DNA polymerase, the forward primer SP1F (5' AAACTGCCTAGGGGTTAAGACCGACGTG 3'), containing an engineered AvrII site (underlined), and the reverse primer AH41R (5' TTTCCCGGGTTATTTCTTCTCGCGGCTATG 3'), containing an engineered SmaI site, were used in a standard PCR following the enzyme manufacturer's directions. The 1.4-kb product was cloned into the Zero Blunt TOPO PCR cloning vector from Invitrogen and transformed into Invitrogen chemically competent TOPO10 E. coli cells. The correct orientation was confirmed by enzyme digestion, and the 1.5-kb insert was excised using the multiple-cloning HindIII site from the vector and the engineered AvrII site in the PCR product and then gel extracted by using the QIAquick gel extraction kit. DNA encoding the first 50 amino acids of YidC2 and including the promoter region was PCR amplified from S. mutans NG8 genomic DNA using the forward primer AH39F (5' CCCGGGAAATAAATGCCAACCTTCAATCA 3') with an engineered SmaI site (underlined) and the reverse primer SP1R (5' AAAACCTAGGGATAACACTTCCCATTGG 3') with an engineered AvrII site (underlined). This amplified product was restricted and ligated into pACYC184 digested with EcoRV and transformed into chemically competent TOPO10 E. coli cells (Invitrogen). Plasmid preparations were screened for correct orientation by enzyme digestion. Plasmid DNA from the correct clone was digested with AvrII and HindIII and gel extracted. pACYC184 containing DNA encoding the first predicted transmembrane segment of YidC2 was ligated to the 1.5-kb HindIII-AvrII fragment containing E. coli yidC, encoding amino acids 59 to 548. The chimeric gene in which DNA encoding the first 59 amino acids of E. coli YidC was replaced with that encoding amino acids 1 to 50 of S. mutans YidC2 (2.2 kb) was excised from pACYC184 with SmaI and then cloned into the shuttle vector pDL289, to generate p50YidC. The correct construct was confirmed by DNA sequencing, and the plasmid was transformed into S. mutans AH398 ({Delta}yidC2) by electroporation (34).

Construction of pCR2.1-yidC1 and pCR2.1-yidC2. The TOPO TA cloning kit from Invitrogen was used to clone S. mutans yidC1 and yidC2 into pCR2.1-TOPO. yidC1 and yidC2 were PCR amplified from UA159 genomic DNA using the following primers: for yidC1, AH30F (5' GTGAAAAAGAAATATAGAATTATTGGATT 3') and AH30R (5' GAGCCTTCATACGAGAAATACCCA 3'); for yidC2, AH31F (5' GTGAAAAAAATTTACAAGCGTCTT 3') and AH31R (5' AGCTTATTGCTTATGGTGACGC 3'). The PCR products were cleaned by using Qiagen QIAquick kits and then cloned into pCR2.1-TOPO following the manufacturer's directions.

Construction of chimeric S. mutans yidC1 and yidC2 and introduction into E. coli. The pCR2.1 vector has a NotI (compatible with EagI) restriction site shortly upstream of yidC1 or yidC2 and an EagI site shortly downstream of them. To construct pACYC184-247YidC1, the EagI fragment from pCR2.1-YidC1 containing yidC1 was inserted into pACYC184-YidC downstream of the yidC stop codon, giving plasmid pACYC184-YidC-YidC1. Additional DNA sequences encoding amino acids after the 247th of YidC and before the 26th of YidC1 were deleted by oligonucleotide-directed loop-out mutagenesis by the QuickChange method. Looping out the DNA sequences encoding the sequence between the first amino acid of YidC and the second amino acid of YidC1 in pACYC184-YidC-YidC1 generates plasmid pACYC184-YidC1. The same strategy was used to construct pACYC184-247YidC2 and pACYC184-YidC2. To construct pACYC184-YidC1-YidC2, one EagI site was introduced into pACYC184-YidC2 upstream of its yidC promoter region (note that the previous EagI site was looped out during construction of pACYC184-YidC2). Then the EagI fragment from pACYC184-YidC2 containing yidC2 was inserted into pACYC184-YidC1 downstream of yidC1. Therefore, production of 247YidC1, 247YidC2, YidC1, and YidC2 was under the control of the yidC promoter.

S. mutans growth curves and complementation in THYE broth. Growth of the S. mutans strain was as described previously (14) with the following changes: overnight cultures of S. mutans wild-type strain NG8, AH374 ({Delta}yidC1), AH398 ({Delta}yidC2), and AH398, harboring p50YidC, were diluted 1:20 in THYE (pH 7.0) medium without antibiotics and grown to an optical density at 600 nm (OD600) of 0.4. A 100-well Bioscreen C plate (Labsystems, Helsinki, Finland) was filled with 300 µl of prewarmed medium (THYE pH 7.0, THYE pH 5.0, or THYE pH 7.0 with 4% NaCl). Wells were inoculated with 30 µl of culture and grown for 16 h, with absorbance recorded every 15 min.

YidC depletion from E. coli. JS7131 cells were grown in LB medium supplemented with 0.2% arabinose at 37°C overnight. After being washed with plain LB twice, the overnight culture was diluted 1:50 in LB containing 0.2% arabinose to express YidC or 0.2% glucose to deplete YidC and grown for ~3 h, at which point a significant growth defect was observed for the glucose culture compared with the arabinose culture.

Complementation of E. coli growth in LB broth. To determine whether S. mutans YidC1 and YidC2 could complement JS7131 in broth grown cultures, cells were grown overnight at 37°C in LB medium containing 0.2% arabinose. Overnight cultures were diluted 1:10 in fresh LB medium containing 0.2% arabinose and grown to an OD600 of 0.7 to 0.8. Cultures were washed once with plain LB and diluted 1:20 in LB medium with 0.2% glucose. Cultures were grown at 37°C for 2 h in order to deplete YidC. After YidC depletion, cultures were diluted 1:25 into fresh LB with 0.2% glucose or LB with 0.2% arabinose. Then 200 µl of each culture was applied in triplicate to a 100-well Bioscreen C plate that was inserted into a Bioscreen C machine (Labsystems, Helsinki, Finland), set at 37°C to read the OD600 every 15 min for 5 h with shaking for 10 min between readings. Doubling times for each strain were calculated as previously described (14). Spectinomycin (25 µg/ml) and chloramphenicol (50 µg/ml) were used where appropriate.

Membrane vesicle preparation. Strain JS7131 complemented with S. mutans YidC1 and YidC2 constructs was grown in LB medium containing 0.2% arabinose. E. coli 1100{Delta}BC was grown in LB medium containing 0.2% glucose. After overnight incubation at 37°C, all cultures were pelleted by centrifugation at 2,500 x g and washed once in LB medium. Pellets were resuspended in plain LB broth and transferred to 1 liter of LB with 0.2% glucose medium supplemented with chloramphenicol (50 µg/ml), and spectinomycin (25 µg/ml) where appropriate. Cultures were grown at 37°C until they reached an OD600 of 0.5 to 0.55 (2.5 to 5 h) and then immediately chilled on ice. Bacterial pellets were collected by centrifugation at 6,000 rpm at 4°C for 10 min. Inner membrane vesicles were made by the French press method (5), with a few changes. Pellets were stored overnight (9 h) at 4°C. Pellets were suspended in 3 ml (50 mM Tris-HCl, 10 mM MgSO4; pH 7.5) buffer, and 20 units of Ambion DNase I was added. After a final ultracentrifugation step, membranes were suspended in 1 ml TM buffer. Membrane protein concentration was determined with a standard bicinchoninic acid assay with bovine serum albumin as the standard. Membranes were stored at 4°C for up to 36 h until biochemical assays were performed.

Western blot detection of expressed YidC, YidC1, and YidC2 proteins. E. coli membrane preparations or S. mutans whole-cell lysates prepared by glass bead breakage of cells three times for 30 s in a Mini-Bead Beater 8 apparatus (BioSpec Products, Inc., Bartlesville, OK) were electrophoresed on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels. Proteins of interest were detected by immunoblotting using horseradish peroxidase-labeled secondary reagents and development with 4-chloro-naphthol substrate solution. YidC proteins were identified with antisera from rabbits immunized with C-terminal peptides of E. coli YidC or S. mutans YidC1 or YidC2. Western blots of E. coli membranes were also reacted with polyclonal antisera against E. coli full-length YidC or polyclonal antisera to Lep. C-terminal synthetic peptides for S. mutans YidC1 and YidC2 used to immunize rabbits (Proteintech Group, Inc.) were LEDEARELEAKKRRAKKKAHKKRK and NPPKPFKSNARKDITPQANNDKKLITS, respectively. The C-terminal peptide (CLEKRGLHSREKKK) antiserum against E. coli YidC was obtained from Rosemary Stuart at Marquette University.

Protease accessibility assay. JS7131 cells were grown under yidC expression or depletion conditions. After 3 h of YidC depletion, the JS7131 cells were collected and suspended in M9 minimal medium containing glucose and grown for an additional 30 min. For yidC expression studies, the JS7131 cells were grown in arabinose for 3 h, collected, suspended in M9 minimal medium with arabinose, and grown for an additional 30 min. Production of the plasmid-encoded membrane protein of interest was induced by the addition of IPTG (1 mM final concentration) to the culture for 10 min. The cells were labeled with [35S]methionine for 1 min and then converted to spheroplasts using lysozyme (1 µg/ml). Aliquots of the spheroplasts were treated either with or without proteinase K (1 mg/ml) on ice for 1 h (for a-P2 and CyoA-N-P2) or overnight (for c-10h-tag; subunit c of the Fo sector of F1Fo ATPase was fused at its C terminus with an eight-amino-acid peptide [GVQDFTST], and a 10-histidine tag was placed in the cytoplasmic loop, creating c-10h-tag, which can be detected using a cobalt affinity resin). The acid-precipitated samples that contained a-P2 and CyoA-N-P2 were solubilized and immunoprecipitated with antibody against Lep (which precipitates P2 containing proteins), GroEL (cytoplasmic protein control), and outer membrane protein A (OmpA; outer membrane protein control), as described elsewhere (50). The c-10h-tag protein was detected differently by a pull-down assay with BD Talon metal affinity resin. Samples were analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) and visualized by phosphorimaging.

Determination of PMF. The proton motive force (PMF) generated from inner membrane vesicles derived from the YidC depletion strain JS7131 complemented with S. mutans YidC1 or YidC2 constructs was determined. Positive controls included membrane vesicles derived from JS7131 rescued with E. coli YidC and the MC1060 wild-type parent strain of JS7131. Negative controls included membrane vesicles derived from JS7131 harboring the pACYC184 vector only and the F1Fo ATP synthase-negative mutant strain 1100{Delta}BC (41). The assay was performed by adding 500 µg of membrane proteins to 3 ml of MOPS buffer (50 mM MOPS [morpholinepropanesulfonic acid], 10 mM MgCl2; pH 7.3) and adding 15 µl of 0.5 M ATP. The acidification of the membrane vesicles was determined by monitoring the fluorescence of 9-amino-6-chloro-2-methoxyacridine (ACMA) in a spectrofluorimeter (Photon Technologies International QuantaMaster 4). The integrity of membrane vesicles was determined by adding NADH to each sample and monitoring the fluorescence of ACMA (not shown).

Determination of ATP hydrolysis specific activity. The F1Fo ATP hydrolysis specific activity was determined by measurement of inorganic phosphate production by the acid molybdate method, after the addition of ATP to membrane vesicles (31). Membrane proteins (120 µg) were used in 3 ml of reaction buffer (50 mM Tris-HCl, 1 mM MgCl2; pH 9.1). A 150 mM ATP stock solution (80 µl) was added to start the reaction, which was allowed to proceed for 2, 5, or 7 min. Inorganic phosphate produced was measured, and specific activity was calculated as nmol Pi/min/mg protein.

Signal peptide processing assay. JS7131 cells were grown under YidC expression or depletion conditions. After 3 h of YidC depletion, the plasmid-encoded protein PClep was induced by the addition of IPTG (isopropyl-β-D-thiogalactopyranoside; 1 mM final concentration) to the culture for 10 min, pulse-labeled with [35S]methionine (100 µCi/ml) for 30 s, and then chased with nonradioactive methionine for 5 s or 120 s. After trichloroacetic acid (TCA) precipitation and immunoprecipitation with anti-Lep, the sample was analyzed by 15% SDS-PAGE and phosphorimaging.

Detection of induction of phage shock protein A. To evaluate expression of the PspA phage shock protein in JS7131 cells complemented with S. mutans YidC1 of YidC2 or derivatives thereof, immunoblot analyses of inner membrane vesicles were performed. Cells expressing plasmid-based S. mutans yidC1 or yidC2 or E. coli yidC were grown in glucose to deplete the chromosomally encoded YidC until the cultures reached an OD600 of 0.5 to 0.55. A similar growth procedure was used for isolating membranes from MC1060 and E. coli 1100{Delta}BC. Equal amounts of protein in the membrane vesicles were loaded on a 15% SDS-PAGE gel. For the time-dependent PspA induction study, E. coli JS7131 cells producing the various YidC1 and YidC2 constructs were grown in glucose to deplete the chromosomally encoded YidC for various times. Cells were normalized and added directly to SDS-PAGE gel loading buffer. PspA was detected by immunoblotting using an ECL Western blot detection kit (Amersham Biosciences).

Test of growth complementation of JS7131 by S. mutans YidC1 and YidC2 and derivatives on LB agar plates. The YidC depletion strain JS7131 was grown in LB medium supplemented with 0.2% arabinose at 37°C overnight. After being washed with plain LB twice, the overnight culture was streaked onto LB agar plates containing 0.2% arabinose or 0.2% glucose and incubated at 37°C overnight.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
E. coli YidC partially rescues the stress-sensitive phenotype associated with elimination of yidC2 from S. mutans. Elimination of yidC2, but not yidC1, in S. mutans results in substantially slower growth under nonstress conditions and an inability to initiate growth under conditions of acid and osmotic stress (14). This phenotype is similar to that observed when genes encoding any of the minimal conserved elements of the SRP pathway, including Ffh, FtsY, or scRNA, are eliminated in S. mutans. Contrary to what is found in other bacteria (19, 32), deletion of these genes from S. mutans is not lethal. S. mutans mutant strains lacking a functional SRP pathway or YidC2 demonstrate significantly decreased proton ATPase activity levels in membrane fractions of cells grown under nonstress as well as acid (pH 5.0) shock conditions. E. coli YidC has been reported to be required for the proper insertion of the Sec-dependent a subunit and to catalyze the Sec-independent insertion of the c subunit of F1Fo ATPase. For this reason, we tested whether E. coli YidC could restore the growth defect, especially that associated with acid shock, in S. mutans lacking yidC2. S. mutans YidC1 and YidC2 are predicted to be lipoproteins synthesized with a cleavable signal peptide that is processed by lipoprotein signal peptidase (43) and, like mitochondrial and chloroplast Oxa1 and Alb3, are predicted to contain five transmembrane segments (Fig. 1A). The E. coli YidC contains an additional uncleaved transmembrane segment and long periplasmic loop (37). To facilitate its targeting in S. mutans, we constructed a chimeric gene in which DNA encoding the first 50 amino acids of YidC2 was fused to that encoding residues 59 to 548 of E. coli YidC to generate 50YidC (Fig. 1B). The S. mutans region contains the predicted lipoprotein signal peptide and a short extracellular region. The E. coli YidC portion includes most of the large periplasmic domain and the five C-terminal transmembrane domains. These five transmembrane domains are well conserved between the E. coli and S. mutans orthologs (Fig. 1C).


Figure 1
View larger version (36K):
[in this window]
[in a new window]

  		FIG. 1. Predicted membrane topologies and comparison of E. coli YidC and S. mutans YidC1 and YidC2. (A) YidC from E. coli spans the membrane six times, with the first transmembrane segment serving as an uncleavable signal sequence, followed by a large periplasmic loop. S. mutans YidC1 and YidC2 are each predicted to span the membrane five times, with an additional hydrophobic region functioning as a cleavable transmembrane targeting sequence (42). The predicted cleavage sites between amino acids 19 and 20 of YidC1 and amino acids 22 and 23 of YidC2 are indicated. (B) Schematic illustration of the chimeric proteins used in the complementation studies. 50YidC is a fusion of amino acid residues 1 to 50 of YidC2 and 59 to 548 of YidC. The DNA construct contains the yidC2 promoter and encodes the predicted first transmembrane targeting sequence of YidC2 (black box), including the cleavage site. The transmembrane region of E. coli YidC are indicated by dark gray boxes. 247YidC1 is a fusion of residues 1 to 247 of YidC and 26 to 271 of YidC1. 247YidC2 is a fusion of residues 1 to 247 of YidC and 25 to 310 of YidC2. Each contains the uncleavable signal sequence and large periplasmic domain of E. coli YidC appended to the transmembrane region (gray boxes) of S. mutans YidC1 or YidC2. (C) Clustal W alignment of E. coli and S. mutans YidCs. The predicted C-terminal five transmembrane segments of S. mutans YidC1 and YidC2 (boxes) are conserved compared to the known E. coli YidC transmembrane segments.

 
After introduction of the plasmid encoding the chimeric 50YidC into the S. mutans {Delta}yidC2 background, Western immunoblot analysis using rabbit antisera generated against C-terminal peptides of S. mutans YidC1, S. mutans YidC2, and E. coli YidC was performed to confirm the appropriate production of YidC1, YidC2, and 50YidC in the S. mutans wild-type, {Delta}yidC1, {Delta}yidC2, and complemented {Delta}yidC2 strains (Fig. 2A). All proteins migrated according to the expected molecular mass of a five-transmembrane-domain molecule, indicating the expected cleavage of the S. mutans targeting domain. The migration of the 50YidC chimeric molecule is also consistent with further processing and elimination of the residual sequence up to the first C-terminal transmembrane domain. As expected, growth of S. mutans lacking YidC2, but not YidC1, was notably impeded compared to that of the wild-type strain (Fig. 2B). Introduction of 50YidC into the S. mutans {Delta}yidC2 background largely restored the growth defect, although not to the wild-type level. Growth of the {Delta}yidC2 mutant under acid stress conditions (Fig. 2C) was restored to nearly the wild-type level by 50YidC, whereas growth under osmotic stress conditions was rescued to a more limited extent (Fig. 2D). These results therefore demonstrate a degree of functional conservation of the YidC orthologs in these two species but also indicate that they are not fully interchangeable and suggest differences in their relative abilities to mediate insertion of some but not all substrates.


Figure 2
View larger version (28K):
[in this window]
[in a new window]

  		FIG. 2. Confirmation of expression of E. coli yidC in S. mutans and complementation of growth of S. mutans {Delta}yidC2 by 50YidC under nonstress, acid stress, and osmotic stress conditions. (A) Western blots of whole-cell lysates of S. mutans wild type, deletion strains, and the complemented {Delta}yidC2 strain. Proteins were resolved on 10% SDS-polyacrylamide gels and identified by reactivity with rabbit antisera raised against C-terminal peptides of S. mutans YidC1, S. mutans YidC2, and E. coli YidC. The apparent molecular mass of each protein of interest is indicated. Elimination of yidC2, but not yidC1, from S. mutans results in a stress-sensitive phenotype and the inability to initiate growth following exposure to acid or osmotic shock. The ability of the E. coli chimeric 50YidC to complement the S. mutans YidC2 mutant strain was evaluated in Todd-Hewitt broth (THYE) culture. (B to D) Growth curves under nonstress (pH 7.0), acid stress (THYE, pH 5.0), and osmotic stress (THYE, 4% salt) of wild-type S. mutans, the {Delta}yidC1 and {Delta}yidC2 mutant strains, and the {Delta}yidC2 strain complemented with 50YidC.

 
Complementation of the E. coli YidC depletion strain with S. mutans YidC1 and YidC2. Previously, the amino-terminal region of E. coli YidC containing its first transmembrane segment and part of the periplasmic region was appended to chloroplast Alb3 and mitochondrial Oxa1 to ensure their efficient insertion into the inner membrane (22, 46). We followed the same strategy here and made YidC1 and YidC2 chimeras, called 247YidC1 and 247YidC2. These contain E. coli YidC residues 1 to 247 fused to either amino acids 26 to 271 of YidC1 and 25 to 310 of YidC2, respectively (Fig. 1B). To test whether the S. mutans molecules are functional in E. coli, we first investigated whether the YidC1, YidC2, 247YidC1, or 247YidC2 protein could complement the growth defect of the E. coli YidC depletion strain in broth-grown cells. The YidC depletion strain JS7131 has YidC production under the control of the araBAD promoter and operator; the endogenous chromosomal copy of yidC has been deleted and yidC placed at the lambda att site under the control of the arabinose promoter (38). Growth in the presence of 0.2% glucose represses yidC expression and leads to depletion of YidC. As YidC is necessary for cell growth, JS7131 grows well in glucose only when plasmid-encoded YidC is present (38). Complementation was assayed by testing whether S. mutans YidC1 and YidC2 could rescue the growth defect of the JS7131 strain when cells were grown in broth containing glucose (0.2%) (Table 1). All the S. mutans YidC1 and YidC2 constructs were able to substantially restore growth of JS7131, albeit to somewhat varying extents. The positive control, E. coli YidC, restored growth to the greatest extent, compared to JS7131 harboring the pACYC184 empty vector only. S. mutans YidC1 was least effective at complementing the JS7131 growth defect, and appending E. coli amino acids 1 to 247 onto either of the S. mutans proteins slightly improved their ability to support growth. Because both yidC1 and yidC2 are expressed in S. mutans, coexpression of yidC1 and yidC2 in E. coli was also evaluated but showed no greater growth restoration than complementation with yidC2 alone. All the control and test strains had similar doubling times when grown in 0.2% arabinose.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Complementation of growth of the E. coli YidC depletion strain JS7131 with S. mutans YidC1 and YidC2 constructs

 
Confirmation of appropriate expression of E. coli YidC and S. mutans YidC1 and YidC2 in the E. coli YidC depletion strain. Before proceeding to the next series of experiments, in which insertion of known YidC substrates into E. coli membrane vesicles was assessed, we confirmed by Western immunoblotting that expression of the chromosomally located E. coli yidC was repressed and YidC was depleted in glucose-grown JS7131 harboring the pACYC184 vector and that expression of yidC or the S. mutans orthologs was induced and detectable in membrane preparations derived from JS7131 cells complemented with the plasmid-based genes. Figure 3A shows the Western blot reacted with an antiserum directed against an E. coli YidC C-terminal peptide. Reactivity was not detected in JS7131 harboring pACYC184 only but was detected in JS7131 complemented with pACYC184 containing yidC and in strain MC1060, the wild-type parent strain from which JS7131 was derived. In addition, YidC was detected in 1100{Delta}BC, a strain in which the F1Fo ATPase was eliminated (41) and which was used as a negative control in subsequent PMF assays. As expected, the E. coli YidC C-terminal peptide-specific antiserum did not recognize S. mutans YidC1 or YidC2. Production of YidC1 and YidC2 and 247YidC1 and 247YidC2 was confirmed with antisera raised against C-terminal peptides corresponding to YidC1 (Fig. 3B) and YidC2 (Fig. 3C). Each unappended or chimeric protein migrated with an apparent molecular mass within several kDa of that predicted. In addition, a polyclonal antiserum raised against purified full-length E. coli YidC (Fig. 3D) cross-reacted with the 247YidC1 and 247YidC2 chimeric proteins and confirmed comparable expression levels of the endogenous and plasmid-encoded E. coli and the S. mutans molecules. Lastly, an antiserum against a known membrane protein, Lep (Fig. 3E), was used to demonstrate uniformity of protein concentrations of the membrane preparations from the various strains.


Figure 3
View larger version (58K):
[in this window]
[in a new window]

  		FIG. 3. Confirmation of appropriate expression of E. coli YidC and S. mutans YidC1 and YidC2 in the JS7131 YidC depletion strain by Western blot analysis. Membranes were prepared from JS7131 harboring the pACYC184 vector only or the same vector containing genes encoding E. coli YidC or S. mutans 247YidC1, 247YidC2, YidC1, or YidC2. The mutant and complemented strains were grown in 0.2% glucose to repress E. coli chromosomal yidC expression, as described in Materials and Methods. Strains MC1060 (wild-type parent of of JS7131) and 1100{Delta}BC (41), which lacks a functional F1Fo ATPase but has not been manipulated with regard to yidC, were also evaluated. Proteins were resolved on a 10% SDS-polyacrylamide gel, and the YidC homologs were revealed by immunoblot analysis with rabbit antisera raised against C-terminal peptides of E. coli YidC (A), S. mutans YidC1 (B), or S. mutans YidC2 (C), as well as with polyclonal antisera against full-length E. coli YidC (D) or Lep (E), a YidC-independent membrane protein. The apparent molecular mass of each protein of interest is indicated.

 
S. mutans YidC1 and YidC2 promote membrane insertion of ATPase subunits a and c in E. coli. Because E. coli YidC could restore the acid-sensitive growth phenotype of the ATPase-impaired S. mutans {Delta}yidC2 mutant strain, and to begin to determine whether the S. mutans orthologs can function in membrane protein biogenesis in E. coli, we investigated the membrane insertion of the YidC-dependent subunits a and c of the F1Fo ATPase (45, 47, 50, 51). While subunit a inserts into the membrane by the Sec pathway, subunit c inserts by the YidC pathway only. To monitor insertion of subunit c we used a subunit c derivative, c-10h-tag, with a short GVQDFTST tag introduced at the C terminus of the protein and a histidine tag introduced into the short cytoplasmic loop. This allows us to pull down the protein with cobalt chelate resin and monitor the insertion of the protein by digestion with proteinase K (50). JS7131 cells bearing plasmids containing genes encoding 247YidC1, 247YidC2, or both YidC1 and YidC2 were pulsed with [35S]methionine for 1 min and converted to spheroplasts. One aliquot was left untreated and another was treated with proteinase K for 1 h to monitor membrane insertion of the YidC constructs. Figure 4A shows that c-10h-tag accumulated in JS7131 cells depleted of YidC, while c-10h-tag inserted efficiently upon induction of YidC expression and was converted to a shorter protected band. Similar to E. coli YidC, S. mutans 247YidC1, 247YidC2, and YidC1/YidC2 each also promoted insertion of c-10h-tag into the membrane. Similar results were observed with the insertion of subunit a. Insertion of subunit a was monitored using a derivative with a P2 domain (derived from leader peptidase) added to the C terminus so that the protein could be immunoprecipitated with leader peptidase antiserum. JS7131 cells expressing the S. mutans YidC1 and YidC2 derivatives were grown for 3 h in glucose medium to deplete chromosomally encoded E. coli YidC. As shown in Fig. 4B, subunit a was not digested by externally added protease under the YidC depletion conditions in the absence of an S. mutans ortholog. However, in the presence of 247YidC1, 247YidC2, or YidC1/YidC2, subunit a was inserted as efficiently as when E. coli YidC was present. In these experiments, OmpA served as a positive control; its degradation measures the efficiency of spheroplast formation. GroEL, a cytoplasmic protein, was a negative control. We used GroEL to confirm that the spheroplasts are intact. These membrane studies show that, similar to E. coli YidC, the S. mutans orthologs function to insert integral membrane components of the F1Fo ATPase in E. coli.


Figure 4
View larger version (39K):
[in this window]
[in a new window]

  		FIG. 4. Membrane insertion of F1Fo ATPase subunits a and c. (A) Membrane insertion of subunit c of the F1Fo ATPase. Proteinase K digestion of c-10h-tag at the periplasmic side generates a slightly smaller protein. JS7131 expressing 247YidC1, 247YidC2, or YidC1/YidC2 was grown under YidC depletion conditions in 0.2% glucose. JS7131 was also studied under YidC expression (Ara) or depletion (Glc) conditions. These strains also contained pMS119-c-10h-tag. After 3 h of YidC depletion, c-10h-tag was induced by the addition of 1 mM IPTG to the culture for 10 min and labeled with [35S]methionine for 1 min. After TCA precipitation and pull-down with cobalt affinity resin, the sample was analyzed by SDS-PAGE and phosphorimaging. (B) Membrane insertion of subunit a of the F1Fo ATPase. Subunit a of the F1Fo ATPase was fused at its C terminus with the P2 domain of leader peptidase. a-P2 can be detected by immunoprecipitation using antiserum against leader peptidase. Proteinase K digestion (PK) of a-P2 at the periplasmic side generates smaller protein fragments. JS7131 expressing 247YidC1, 247YidC2, or YidC1/YidC2 was grown under YidC depletion conditions (Glc). JS7131 without the homologs was grown under YidC expression (Ara) or depletion (Glc) conditions. These strains also contained pMS119-a-P2. After 3 h of YidC depletion, a-P2 was induced by the addition of 1 mM IPTG to the culture for 10 min and labeled with [35S]methionine for 1 min. After TCA precipitation and immunoprecipitation with anti-Lep, the sample was analyzed by 15% SDS-PAGE and phosphorimaging (top panel). As a control, another aliquot of the [35S]methionine-labeled cells was immunoprecipitated with GroEL and OmpA antisera. The sample was analyzed by SDS-PAGE and phosphorimaging (bottom panel).

 
Rescue of the impaired PMF and F1Fo ATPase activity in the YidC depletion strain by S. mutans YidC1 and YidC2. Previously, it was shown that YidC depletion in E. coli leads to a defect in the assembly and activity of the F1Fo ATPase (48). Because the S. mutans YidC orthologs demonstrated functional activity with regard to insertion of the ATPase a and c subunits, we next wished to determine whether their expression restored the PMF and F1Fo ATPase enzymatic activity in the E. coli YidC depletion strain. First, the integrity of the F1Fo ATPase was evaluated by measuring the PMF with membranes energized by the addition of ATP. Under these conditions, the PMF, or ability to pump protons through membranes, is mediated by the membrane integral Fo component of the F1Fo proton ATPase. The PMF was measured by the fluorescence quenching of ACMA, which was monitored with a spectrofluorimeter. ATP hydrolysis by the F1Fo ATPase generates a PMF by pumping protons into the lumens of the vesicles. Inverted membrane vesicles were isolated from JS7131 cells containing the pACYC184 vector alone or the vector containing genes encoding YidC, 247YidC1, 247YidC2, YidC1, or YidC2. In addition, membranes were isolated from E. coli 1100{Delta}BC, which lacks the F1Fo ATPase (8), as a negative control. Figure 5A shows that, similar to E. coli YidC, each of the S. mutans constructs restored a PMF and rescued the defect in the JS7131 YidC depletion strain. Membrane vesicles derived from the negative control strain 1100{Delta}BC, which is completely devoid of the F1Fo ATPase, did not generate a detectable PMF upon addition of ATP to the membranes. Therefore, the low PMF measured in JS7131 harboring the pACYC184 vector only likely reflects the low level of residual YidC associated with the bacterial membranes following YidC repression by growth in glucose and is the relevant background control by which to compare complementation by plasmid-encoded YidC orthologs. The results of this experiment indicate that both S. mutans YidC1 and YidC2 can support the assembly of the F1Fo ATPase into a functional state in E. coli.


Figure 5
View larger version (33K):
[in this window]
[in a new window]

  		FIG. 5. Rescue of the PMF and F1Fo ATPase enzymatic activity in the YidC depletion strain JS7131 by S. mutans YidC1 and YidC2. (A) The F1Fo ATPase activity was measured by examining the PMF generated upon addition of ATP to membrane vesicles. Inner membrane vesicles were prepared as described in Materials and Methods from E. coli 1100{Delta}BC (curve 1) engineered to lack a functional F1Fo ATPase, from the YidC depletion strain JS7131 harboring the pACYC184 vector alone (curve 2), and from JS7131 harboring pACYC184 containing genes encoding E. coli YidC (curve 3) or S. mutans 247YidC1 (curve 4), 247YidC2 (curve 5), YidC1 (curve 6), or YidC2 (curve 7). Five hundred micrograms of membrane proteins was used to analyze PMF, through the quenching of ACMA monitored with a spectrofluorimeter. The integrity of membrane vesicles was determined by adding NADH to each sample and monitoring the fluorescence of ACMA (not shown). (B) The ATP specific activity associated with the membrane fractions was determined by the acid molybdate method (31). The inner membrane vesicle preparations (120 µg) that were used for the PMF assay were also used for the ATP hydrolysis assay. Specific activities were calculated as nmol Pi/min/mg protein.

 
We also measured directly the ATPase activity of membranes derived from glucose grown JS7131 cells expressing the S. mutans orthologs with E. coli YidC as a positive control. The specific activity was evaluated by the acid molybdate method, which measures the production of inorganic phosphate after the addition of ATP to the membranes (31). As can be seen in Fig. 5B, the YidC depletion strain has reduced ATPase activity compared to the same strain expressing yidC from the pACYC184-based plasmid. Likewise, the S. mutans YidC1 and YidC2 proteins and their derivatives restore the ATPase activity to a level comparable to that achieved when E. coli YidC is present, indicating that each of the S. mutans orthologs can function to confer F1Fo ATPase activity when present in E. coli.

S. mutans YidC1 and YidC2 promote membrane insertion of M13 procoat protein and the amino-terminal portion of pre-CyoA. To continue exploring the degree of functional redundancy of the YidC orthologs, the S. mutans proteins were tested for their ability to mediate membrane insertion of other known E. coli YidC-dependent substrates. Previous studies have shown that the M13 procoat and PClep require YidC for membrane insertion (11, 38). PClep is a small M13 phage coat protein fused at its C terminus with the P2 domain of leader peptidase (11). To examine membrane insertion, JS7131 cells expressing PClep and the S. mutans YidC orthologs were labeled with [35S]methionine for 30 s and chased with cold methionine for 5 or 120 s. Figure 6A shows that when 247YidC1, 247YidC2, or YidC1/YidC2 are present with PClep in the YidC depletion strain, signal peptide processing is similar to that observed when E. coli yidC expression is induced with arabinose. In JS7131 grown in glucose (a negative control), processing of PClep to Clep is inhibited.


Figure 6
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 6. Membrane insertion of PClep and CyoA-N-P2, a derivative of bacterial subunit II of cytochrome bo oxidase. (A) Membrane insertion of PClep. PClep is synthesized as a precursor with a cleavable signal sequence. Signal peptide processing by leader peptidase generates the mature protein coat Lep. JS7131 expressing 247YidC1 or 247YidC2 or YidC1/YidC2 were grown under YidC depletion conditions. In addition, JS7131 was grown under YidC expression (Ara) and depletion (Glc) conditions. These strains also contained the pMS119-PClep plasmid. After 3 h of YidC depletion, PClep was induced by the addition of 1 mM IPTG for 10 min, pulse-labeled with [35S]methionine for 30 s, and then chased with nonradioactive methionine for 5 s or 120 s. After TCA precipitation and immunoprecipitation with anti-Lep, the sample was analyzed by 15% SDS-PAGE and phosphorimaging. (B) Protease accessibility assay of CyoA-N-P2. JS7131 expressing 247YidC1, 247YidC2, or YidC1/YidC2 was grown under YidC depletion conditions (Glc). JS7131 was grown under YidC expression (Ara) or depletion (Glc) conditions. These strains also contained the pMS119- CyoA-N-P2 plasmid. After 3 h of YidC depletion, CyoA-N-P2 was induced by the addition of 1 mM IPTG for 10 min and labeled with [35S]methionine for 1 min. After TCA precipitation and immunoprecipitation with anti-Lep serum, the sample was analyzed by SDS-PAGE and phosphorimaging.

 
We next tested membrane protein insertion of a key subunit of the respiratory complex cytochrome bo oxidase. CyoA, subunit II of cytochrome bo oxidase, is known to require E. coli YidC for membrane insertion (9, 13, 44). CyoA is a lipoprotein and spans the membrane twice, with both its N and C termini localized in the periplasmic space (27). It is made in a precursor form termed pre-CyoA with a cleavable signal peptide. After membrane insertion, pre-CyoA is cleaved by signal peptidase 2 to the mature CyoA protein (44). Interestingly, the amino-terminal portion of pre-CyoA inserts by the YidC-only pathway, while the large C-terminal domain translocates across the membrane by the Sec pathway (9, 44). Here, we examined the membrane insertion of the Sec-independent CyoA amino-terminal domain using the preCyoA-N-P2 derivative. The P2 domain (derived from leader peptidase) is added in the cytoplasmic loop so that we can immunoprecipitate the protein using anti-leader peptidase serum. JS7131 producing S. mutans 247YidC1, 247YidC2, or YidC1/YidC2 was grown in LB for 3 h in glucose to deplete the chromosomally encoded E. coli YidC. The cells were pulsed with [35S]methionine for 1 min, converted to spheroplasts, and subjected to a protease accessibility study. When YidC is depleted (Fig. 6B), preCyoA-N-P2 accumulates and is resistant to proteinase K digestion. When YidC is produced or when a S. mutans ortholog is present under YidC depletion conditions, CyoA-N-P2 is processed by signal peptidase 2 and digested by proteinase K to a shorter protected fragment (Fig. 6B). These results indicate that when they are produced in E. coli, the S. mutans YidC1 and YidC2 orthologs can functionally replace YidC to insert its known substrates, M13 procoat protein and the N-terminal portion of pre-CyoA.

Induction of the phage shock PspA protein in the E. coli depletion strain complemented with S. mutans YidC1 and YidC2. As a final step to investigate whether the S. mutans YidC1 and YidC2 homologs can completely substitute for YidC in E. coli, we tested whether the stress phage shock PspA protein is induced in the YidC depletion strain producing the S. mutans proteins. Previous studies showed that PspA is induced upon depletion of YidC (48) stemming from a defect in maintenance of the membrane PMF (24, 28). The membrane preparations used for Fig. 3 (with Lep as the invariant internal control) of JS7131 with pACYC184 expressing E. coli yidC or the S. mutans orthologs, E. coli MC1060 (parent strain of JS7131), and E. coli 1100{Delta}BC were used in this experiment. Figure 7A shows that as expected, when grown in glucose, negative control JS7131 cells harboring the empty pACYC194 vector demonstrated a pronounced induction of the PspA protein. There was also a notable yet less pronounced induction of the PspA protein in 1100{Delta}BC, which lacks a functional F1Fo ATPase and is unable to generate a PMF with membrane vesicles using ATP hydrolysis. Compared to MC1060 (wild type) or JS7131 rescued with plasmid-encoded YidC, there was a slight but observable increase in the detection of PspA in JS7131 complemented with 247YidC1, 247YidC2, YidC1, or YidC2. This indicates that despite considerable functional overlap of each of the S. mutans proteins in an E. coli background, membrane integrity is not fully regained, and suggests the existence of unique features of E. coli YidC that cannot completely be replicated by substitution of the heterologous proteins.


Figure 7
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 7. Evaluation of the phage shock response and PspA induction by Western blot analysis. (A) Membranes were prepared from JS7131 strain with pACYC184 with no insert or genes encoding E. coli YidC or S. mutans YidC1, YidC2, 247YidC1, or 247YidC2. Also, membranes were prepared from the JS7131 parent strain MC1060 and 1100{Delta}BC lacking a functional F1Fo ATPase. Bacteria were grown in LB supplemented with 0.2% glucose, 40 µg of each membrane preparation was resolved on a 10% SDS-polyacrylamide gel, and the PspA protein was revealed by Western blotting. (B) JS7131 harboring plasmids encoding YidC orthologs was grown overnight in LB containing 0.2% arabinose. After being washed twice with plain LB medium, the overnight culture was diluted 1:50 in LB containing 0.2% glucose to deplete the chromosomally encoded YidC during cell growth. After 3, 4, and 5 h of YidC depletion, cells were collected and lysed with SDS buffer. Equal amounts of each cell lysate were analyzed by Western blotting with antiserum against PspA.

 
To examine in more detail the PspA induction profile in the presence of the S. mutans orthologs, we next performed a time course experiment. After 3 h of growth in glucose, JS7131 expressing 247YidC1, 247YidC2, or YidC1/YidC2 did not show an induction of PspA (Fig. 7B). However, this stress protein was induced after 4 h of growth in the depletion strain producing 247YidC1 or 247YidC2. Interestingly, JS7131 cells containing both YidC1 and YidC2 show induction of the PspA protein at the 5-h, but not the 4-h, time point. These results reiterate that 247YidC1 and 247YidC2 only partially complement the PspA shock response in JS7131 and suggest a potential cooperative activity of S. mutans YidC1 and YidC2 that can further alleviate the phage shock response, as evidenced by an additional delay in PspA induction.

Complementation of E. coli growth by S. mutans YidC1 and YidC2 on solid medium. Surprisingly, we observed different results when the E. coli YidC depletion strain JS7131 was complemented with S. mutans 247YidC1, 247YidC2, YidC1, YidC2, or YidC1/YidC2 and streaked on LB agar plates compared to growth in LB broth (Fig. 8). Under YidC repression conditions on 0.2% glucose, JS7131 was unable to grow, while complementation following introduction of the positive control plasmid encoding E. coli YidC was readily apparent. On the control plate, all strains were viable and grew equally well in the presence of 0.2% arabinose. Poor complementation with unappended S. mutans YidC2 and none with unappended YidC1 were observed on 0.2% glucose. However, complementation was notably improved for YidC2 when the chimeric version of the protein, 247YidC2, was used. Slight growth of JS3171 was consistently observed in the presence of 247YidC1. Interestingly, coexpression of yidC1 and yidC2 resulted in improved complementation compared to expression of either alone. Despite clearly observable growth of JS7131 complemented with 247YidC2 or YidC1/YidC2, this growth was slower than when JS7131 was complemented with E. coli YidC. Collectively, these results indicate a functional difference between S. mutans YidC1 and YidC2 when they are expressed in E. coli that is more readily detectable when the cells are grown on solid but not in liquid medium. In addition, similar to the results described above for the delay in induction of the phage shock response, S. mutans YidC1 and YidC2 appear to act in a cooperative manner in E. coli that is more readily observed under certain growth conditions.


Figure 8
View larger version (79K):
[in this window]
[in a new window]

  		FIG. 8. Complementation of growth of the E. coli YidC depletion strain by S. mutans YidC1, YidC2, 247YidC1, 247YidC2, and YidC1/YidC2 on solid medium. The YidC depletion strain JS7131 bearing the empty pACYC184 vector only or plasmids encoding the various YidC constructs was grown in LB medium supplemented with 0.2% arabinose, washed in plain LB broth, streaked onto LB plates containing 0.2% arabinose or 0.2% glucose, and incubated at 37°C overnight.

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this report we showed that in many instances YidC orthologs from gram-positive and gram-negative organisms behave as functional equivalents, but this is not the case in every experiment. Results vary depending not only on the YidC variant itself but also on their context, including the host species and the environmental conditions under which they are grown. Although the molecular basis of salt sensitivity of the yidC2 mutant is not yet understood, our results suggest that YidC2 contributes to membrane biogenesis to confer osmotic tolerance in a way that cannot be fully replicated by E. coli YidC. Acid tolerance in S. mutans has been more widely studied and is mediated in large part by the proton pump of the F1Fo ATPase (4). The known participation of E. coli YidC in the membrane insertion of the a and c subunits of the Fo component of the F1Fo multisubunit ATPase enzyme complex (51) is thus consistent with its ability to restore acid tolerance to the S. mutans {Delta}yidC2 mutant. However, while both S. mutans YidC1 and YidC2 function when introduced into the E. coli YidC depletion strain to insert the a and c subunits and to restore the PMF and ATPase activity (Fig. 3 and 4), indicating proper assembly of the enzyme, elimination of yidC1 from S. mutans does not confer an acid-sensitive phenotype (14). This interesting dichotomy suggests that while E. coli YidC and both of the S. mutans orthologs are equally effective with regard to the YidC-dependent steps involved in conferring the proton pumping capacity of the F1Fo ATPase in E. coli, in the S. mutans background, YidC2 contributes a function that is not shared by YidC1.

Results presented in this paper confirm an expected conservation of function between YidC orthologs of gram-negative and gram-positive bacteria, although at least some of this functional activity appears to be vestigial, as both of the S. mutans YidC orthologs substituted for the E. coli protein in the insertion of the CyoA-N-P2 derivative of subunit II of cytochrome bo oxidase despite the lack of this respiratory complex in S. mutans. Also, while one would not expect such an encounter in nature, both S. mutans orthologs functioned to insert the PClep derivative of the M13 bacteriophage procoat protein. Therefore, depending on the host species or organelle and the substrate, functional substitution of YidC/Oxa1/Alb3 family members may be independent of or dependent on unique structural features of the proteins. For example, YidC orthologs in gram-negative organisms lack a charged carboxyl-terminal domain that is located at the end of mitochondrial Oxa1 and chloroplast Alb3 (25, 49). E. coli YidC can functionally replace mitochondrial Oxa1 only if the carboxyl-terminal ribosomal binding domain of Oxa1 is appended to YidC (33). Recently it was shown that mitochondrial Oxa1 promotes assembly of Atp9 (the mitochondrial homolog of the bacterial subunit c) into the mitochondrial F1Fo-ATP synthase complex in a posttranslational manner that is not dependent on the C-terminal region of Oxa1 (21). Oxa1 is not required for insertion of Atp9, in contrast to E. coli YidC, which is essential for membrane insertion of subunit c both in vivo and in vitro (45, 47, 51). Sequence analysis predicts that S. mutans YidC2, but not YidC1, of S. mutans contains a charged cytoplasmic C-terminal tail similar to that of Oxa1. Hence, E. coli YidC is more YidC1-like in this regard. Yet our results demonstrate that YidC can serve to restore acid tolerance, which is YidC2 dependent but not YidC1 dependent in S. mutans. In the context of a single gene deletion of yidC2, therefore, unique features of the C terminus of the YidC orthologs do not appear to be essential for the generation of the membrane machinery necessary to contend with acid stress. This may not be the case for other phenotypic consequences of yidC2 deletion, e.g., sensitivity to osmotic shock, or when multiple components of the membrane targeting and translocation components are disrupted simultaneously. At least partially overlapping functions appear to exist between the SRP cotranslational protein translocation pathway and YidC2 in S. mutans. Elimination of yidC2 or disruption of the SRP pathway results in almost identical stress-sensitive phenotypes, and mutants with a double deletion of yidC2 and genes encoding SRP pathway components are not viable (14). Likewise, at least some degree of cooperation and/or overlap of YidC1 and YidC2 function is implied, since double mutants of yidC1 and yidC2 also are not viable. Membrane composition changes and the physiological adaptation associated with perturbations in protein translocation in the gram-positive organism S. mutans have only recently begun to be defined (15).

S. mutans YidC1 and YidC2 and their derivatives 247YidC1 and 247YidC2 complemented growth in broth of the YidC depletion strain JS7131 (Table 1) and in all cases tested also functioned in the membrane insertion of known E. coli YidC substrates (Fig. 4 and 6). Despite this demonstrated interchangeability between the bacterial YidC orthologs, these proteins apparently also are optimized for the specific systems in which they reside. For example, a residual perturbation in membrane biogenesis is indicated by induction of the phage shock stress protein PspA when S. mutans YidC1 and YidC2 derivatives are present in the JS7131 YidC depletion strain (Fig. 7). In E. coli, pspA expression is induced by several environmental stress conditions, including infection with filamentous phage, disruption of the PMF, high temperature, osmotic stress, ethanol or organic solvent exposure, and stationary-phase alkaline pH (reviewed in reference 35). The S. mutans YidC orthologs do not completely complement the phage shock protein response in E. coli, as PspA induction is delayed but not totally eliminated, suggesting that the cells still sustain some alteration in membrane composition. However, the detectable phage shock response is not due to an inability of the F1Fo ATPase to generate a PMF using ATP as the energy source (Fig. 5). A cooperative interaction between S. mutans YidC1 and YidC2 is also implied, as PspA induction was further delayed when both proteins were produced simultaneously in E. coli. This result would be consistent with at least some circumstances in which the two S. mutans paralogs work in concert and the inability to delete both of them simultaneously from this organism.

Lastly, we found that S. mutans 247YidC1, YidC1, and YidC2 cannot substitute to an appreciable extent for E. coli YidC in JS7131 during growth on solid agar plates (Fig. 8), possibly because they are not fully operational in supporting membrane insertion of a key protein or proteins needed under these conditions. Interestingly, replacing the N terminus of the S. mutans orthologs with the N-terminal 247 amino acids of E. coli YidC, encompassing the first transmembrane domain and periplasmic loop (Fig. 1), improved complementation of JS3171 growth on plates compared to broth cultures. This was particularly evident for the YidC2 construct. The lack of a universal requirement for this sequence was underscored in experiments in which each of the S. mutans proteins restored the PMF in the E. coli YidC depletion strain regardless of whether the E. coli N-terminal region was appended. Again, similar to the potential cooperative activity observed between YidC1 and YidC2 in the delay and partial alleviation of the phage shock response in JS7131, the degree of rescue of E. coli growth on plates was also visibly improved by simultaneous expression of YidC1 and YidC2 compared to either protein alone.

Taken together, our results add to the growing body of evidence indicating functional overlap between the evolutionarily conserved YidC/Oxa1/Alb3 family of bacterial and organelle proteins but also highlight the underlying complexity and existence of unique features dictating their full range of functions in the systems under which they operate.


   ACKNOWLEDGMENTS
 
This work was supported by National Institutes of Health grants GM63862 to RED, DK51131 to H.R.K., and DE08007 to L.J.B. S.R.P. was supported by National Institute of Dental and Craniofacial Research Training Grant DE07200.

We thank Paula Crowley for editorial assistance with the manuscript.


   FOOTNOTES
 
* Corresponding author. Mailing address for L.J.B.: Department of Oral Biology, University of Florida, P.O. Box 100424, FL 32610. Phone: (352) 846-0785. Fax: (352) 846-0786. E-mail: jbrady{at}dental.ufl.edu. Mailing address for R.E.D.: Department of Chemistry, The Ohio State University, 100 W. 18th Ave., Columbus, OH 43210. Phone: (614) 292-2384. Fax: (614) 292-1532. E-mail: dalbey{at}chemistry.ohio-state.edu Back

{triangledown} Published ahead of print on 4 January 2008. Back

{dagger} Y.D. and S.R.P. contributed equally to this work. Back


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Altamura, N., N. Capitanio, N. Bonnefoy, S. Papa, and G. Dujardin. 1996. The Saccharomyces cerevisiae OXA1 gene is required for the correct assembly of cytochrome c oxidase and oligomycin-sensitive ATP synthase. FEBS Lett. 382:111-115.[CrossRef][Medline]
  2. Bauer, M., M. Behrens, K. Esser, G. Michaelis, and E. Pratje. 1994. PET1402, a nuclear gene required for proteolytic processing of cytochrome oxidase subunit 2 in yeast. Mol. Gen. Genet. 245:272-278.[CrossRef][Medline]
  3. Beck, K., G. Eisner, D. Trescher, R. E. Dalbey, J. Brunner, and M. Muller. 2001. YidC, an assembly site for polytopic Escherichia coli membrane proteins located in immediate proximity to the SecYE translocon and lipids. EMBO Rep. 2:709-714.[CrossRef][Medline]
  4. Bender, G. R., S. V. Sutton, and R. E. Marquis. 1986. Acid tolerance, proton permeabilities, and membrane ATPases of oral streptococci. Infect. Immun. 53:331-338.[Abstract/Free Full Text]
  5. Bhatt, D., S. P. Cole, T. B. Grabar, S. B. Claggett, and B. D. Cain. 2005. Manipulating the length of the b subunit F1 binding domain in F1F0 ATP synthase from Escherichia coli. J. Bioenerg. Biomembr. 37:67-74.[CrossRef][Medline]
  6. Bonnefoy, N., F. Chalvet, P. Hamel, P. P. Slonimski, and G. Dujardin. 1994. OXA1, a Saccharomyces cerevisiae nuclear gene whose sequence is conserved from prokaryotes to eukaryotes controls cytochrome oxidase biogenesis. J. Mol. Biol. 239:201-212.[CrossRef][Medline]
  7. Bonnefoy, N., M. Kermorgant, O. Groudinsky, M. Minet, P. P. Slonimski, and G. Dujardin. 1994. Cloning of a human gene involved in cytochrome oxidase assembly by functional complementation of an oxa1– mutation in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 91:11978-11982.[Abstract/Free Full Text]
  8. Buckley, N. D., L. N. Lee, and D. J. LeBlanc. 1995. M39 construction of a mobilizable vector for genetic analysis of Streptococcus sobrinus. Dev. Biol. Stand. 85:399-401.[Medline]
  9. Celebi, N., L. Yi, S. J. Facey, A. Kuhn, and R. E. Dalbey. 2006. Membrane biogenesis of subunit II of cytochrome bo oxidase: contrasting requirements for insertion of N-terminal and C-terminal domains. J. Mol. Biol. 357:1428-1436.[CrossRef][Medline]
  10. Chen, M., J. C. Samuelson, F. Jiang, M. Muller, A. Kuhn, and R. E. Dalbey. 2002. Direct interaction of YidC with the Sec-independent Pf3 coat protein during its membrane protein insertion. J. Biol. Chem. 277:7670-7675.[Abstract/Free Full Text]
  11. Chen, M., K. Xie, N. Nouwen, A. J. Driessen, and R. E. Dalbey. 2003. Conditional Lethal Mutations Separate the M13 Procoat and Pf3 Coat Functions of YidC: different YidC structural requirements for membrane protein insertion. J. Biol. Chem. 278:23295-23300.[Abstract/Free Full Text]
  12. Chen, M., K. Xie, J. Yuan, L. Yi, S. J. Facey, N. Pradel, L. F. Wu, A. Kuhn, and R. E. Dalbey. 2005. Involvement of SecDF and YidC in the membrane insertion of M13 procoat mutants. Biochemistry 44:10741-10749.[CrossRef][Medline]
  13. du Plessis, D. J., N. Nouwen, and A. J. Driessen. 2006. Subunit a of cytochrome o oxidase requires both YidC and SecYEG for membrane insertion. J. Biol. Chem. 281:12248-12252.[Abstract/Free Full Text]
  14. Hasona, A., P. J. Crowley, C. M. Levesque, R. W. Mair, D. G. Cvitkovitch, A. S. Bleiweis, and L. J. Brady. 2005. Streptococcal viability and diminished stress tolerance in mutants lacking the signal recognition particle pathway or YidC2. Proc. Natl. Acad. Sci. USA 102:17466-17471.[Abstract/Free Full Text]
  15. Hasona, A., K. Zuobi-Hasona, P. J. Crowley, J. Abranches, M. A. Ruelf, A. S. Bleiweis, and L. J. Brady. 2007. Membrane composition changes and physiological adaptation by Streptococcus mutans signal recognition particle pathway mutants. J. Bacteriol. 189:1219-1230.[Abstract/Free Full Text]
  16. He, S., and T. D. Fox. 1997. Membrane translocation of mitochondrially coded Cox2p: distinct requirements for export of N and C termini and dependence on the conserved protein Oxa1p. Mol. Biol. Cell 8:1449-1460.[Abstract]
  17. Hell, K., J. Herrmann, E. Pratje, W. Neupert, and R. A. Stuart. 1997. Oxa1p mediates the export of the N- and C-termini of pCoxII from the mitochondrial matrix to the intermembrane space. FEBS Lett. 418:367-370.[CrossRef][Medline]
  18. Hell, K., W. Neupert, and R. A. Stuart. 2001. Oxa1p acts as a general membrane insertion machinery for proteins encoded by mitochondrial DNA. EMBO J. 20:1281-1288.[CrossRef][Medline]
  19. Honda, K., K. Nakamura, M. Nishiguchi, and K. Yamane. 1993. Cloning and characterization of a Bacillus subtilis gene encoding a homolog of the 54-kilodalton subunit of mammalian signal recognition particle and Escherichia coli Ffh. J. Bacteriol. 175:4885-4894.[Abstract/Free Full Text]
  20. Houben, E. N., P. A. Scotti, Q. A. Valent, J. Brunner, J. L. de Gier, B. Oudega, and J. Luirink. 2000. Nascent Lep inserts into the Escherichia coli inner membrane in the vicinity of YidC, SecY and SecA. FEBS Lett. 476:229-233.[CrossRef][Medline]
  21. Jia, L., M. K. Dienhart, and R. A. Stuart. 2007. Oxa1 directly interacts with Atp9 and mediates its assembly into the mitochondrial F1F0-ATP synthase complex. Mol. Biol. Cell 18:1897-1908.[Abstract/Free Full Text]
  22. Jiang, F., L. Yi, M. Moore, M. Chen, T. Rohl, K. J. Van Wijk, J. W. De Gier, R. Henry, and R. E. Dalbey. 2002. Chloroplast YidC homolog Albino3 can functionally complement the bacterial YidC depletion strain and promote membrane insertion of both bacterial and chloroplast thylakoid proteins. J. Biol. Chem. 277:19281-19288.[Abstract/Free Full Text]
  23. Kermorgant, M., N. Bonnefoy, and G. Dujardin. 1997. Oxa1p, which is required for cytochrome c oxidase and ATP synthase complex formation, is embedded in the mitochondrial inner membrane. Curr. Genet. 31:302-307.[CrossRef][Medline]
  24. Kleerebezem, M., W. Crielaard, and J. Tommassen. 1996. Involvement of stress protein PspA (phage shock protein A) of Escherichia coli in maintenance of the protonmotive force under stress conditions. EMBO J. 15:162-171.[Medline]
  25. Luirink, J., G. v. Heijne, E. Houben, and J.-W. d. Gier. 2005. Biogenesis of inner membrane proteins in Escherichia coli. Annu. Rev. Microbiol. 59:329-355.[CrossRef][Medline]
  26. Luirink, J., T. Samuelsson, and J. W. de Gier. 2001. YidC/Oxa1p/Alb3: evolutionarily conserved mediators of membrane protein assembly. FEBS Lett. 501:1-5.