Expression and epitope tagging of the membrane anchor subunit

peng$$0209

Protein Engineering vol.10 no.3 pp.285–290, 1997

Expression and epitope tagging of the membrane anchor subunit (DmsC) of Escherichia coli dimethyl sulfoxide reductase

Raymond J.Turner, Jody L.Busaan, John H.Lee, Marek Michalak and Joel H.Weiner1 MRC Group in the Molecular Biology of Membranes, Department of Biochemistry, University of Alberta, Edmonton, Alberta, Canada, T6G 2H7 1To

whom correspondence should be addressed

Escherichia coli dimethyl sulfoxide reductase is a heterotrimer comprising a catalytic subunit (DmsA), an electron transfer subunit (DmsB) and an integral membrane anchor subunit (DmsC). DmsC is not antigenic and the production of antibodies to this subunit has not been successful. We have tagged DmsC at the C-terminus with a dystrophin-specific amino acid sequence (dysp) to which antibodies are readily available. We were able to use this tagging technique to monitor expression and localization of DmsC in E.coli and non-muscle eukaryotic cells. Growth properties of wild-type E.coli, strain HB101, overexpressing DmsC:dysp suggest that the expression of DmsC is lethal to E.coli. The lethality could be overcome by utilizing an E.coli F0F1 ATPase mutant as the host. Growth conditions of culture density, duration of induction, temperature of incubation after induction and media conditions were investigated to optimize DmsC:dysp accumulation levels. In order to alleviate the problem arising from the toxicity of DmsC, expression in eukaryotic tissue culture was also explored. A plasmid expressing DmsC:dysp was transfected into COS-1 or McA-RH777 cells. The presence of expressed DmsC:dysp was confirmed using specific anti-dysp antibodies and immunofluorescence microscopy analysis revealed that the DmsC:dysp was localized to the endoplasmic reticulum. Expression of DmsC:dysp did not appear to be toxic to the eukaryotic cells. These data suggest methodologies to overcome lethality problems associated with the overexpression of integral membrane proteins like DmsC. Keywords: DmsC/epitope tagging/Escherichia coli dimethyl sulfoxide reductase/expression/membrane anchor subunit

Introduction The production of sufficient amounts of purified recombinant membrane protein is a limiting factor in the study of integral membrane protein structure and function. Problems exist in trying to overexpress integral membrane proteins whether in their natural host or in a heterologous host (Grisshammer and Tate, 1995). In addition to the difficulties associated with obtaining expression at high levels, membrane proteins are often plagued by problems of detection resulting from poor stain binding or lack of antigenicity. The molecular biological methodology of fusing the target gene downstream from a highly expressed gene (N-terminal fusion) has proven useful with soluble proteins (Phizicky and Fields, 1995). This methodology has also been incorporated to determine the topology of integral membrane proteins through C-terminal © Oxford University Press

fusions (Traxler et al., 1993). However, tagging a large soluble protein on to a membrane protein may seriously inhibit the membrane insertion, assembly and function. To reduce the problem associated with such large fusions, affinity tags are quickly becoming the method of choice (Grisshammer and Tate, 1995) Escherichia coli dimethyl sulfoxide reductase (DMSO reductase, DmsABC) is a heterotrimer comprising a molybdopterin guanine dinucleotide cofactor containing catalytic subunit (DmsA, 87.4 kDa), a 4[4Fe–4S] cluster containing electron transfer subunit (DmsB, 23.1 kDa) and an integral membrane anchor subunit (DmsC, 30.8 kDa) (Weiner et al., 1992; Rothery et al., 1995) (Figure 1A). DmsABC can be readily overexpressed from a multicopy plasmid containing the dmsABC operon. Although both DmsA and DmsB are easily visualized on Coomassie Blue-stained SDS–PAGE, DmsC is less readily identifiable. DmsC is not very antigenic and the production of antibodies to this subunit has not been successful (Sambasivarao et al., 1990). Furthermore, attempts

Fig. 1. Models of DMSO reductase. (A) Overall topology and mechanism of E.coli DMSO reductase (DmsABC). Abbreviations: MQ, menaquinone; 4Fe–4S, iron–sulfur centre; Mo-MGD, molybdopterin guanine dinucleotide. (B) Topology and sequence of the anchor subunit DmsC containing the dystrophin-specific epitope sequence (in bold circles).

285

R.J.Turner et al.

to express or subclone DmsC alone without DmsA and DmsB have also met with little success. Upon determining the topology of DmsC, an alkaline phosphatase (PhoA) fusion was identified at the C-terminus of DmsC (Weiner et al., 1993). This DmsABC:PhoA complex had both DMSO reductase and alkaline phosphatase activities, implying that a fusion at the C-terminus of DmsC is non-disruptive to the folding of DmsC or to the assembly of DmsABC. The topology of DmsC (Figure 1B) dictates that the C-terminus is on the periplasmic side, opposite that of the side in which DmsA and DmsB interact with DmsC (Rothery and Weiner, 1993; Weiner et al., 1993) (Figure 1A). In this study, we generated a fusion to the C-terminus of DmsC by polymerase chain reaction (PCR) mutagenesis. For this fusion we used a small peptide to which an antibody was readily available, thereby increasing the ease of identification. We chose to tag DmsC with a dystrophin-specific amino acid sequence (dysp). Since large quantities of dystrophin are found in muscle tissues only, we were able to use this tagging technique to monitor expression and localization of DmsC via a dysp-specific antibody. We examined DmsC:dysp in wildtype E.coli and an F0F1 ATPase mutant under a variety of growth and induction conditions in order to optimize DmsC expression and accumulation. We also looked at expression in eukaryotic tissue cultures. A preliminary version of this work has been published as an abstract (Turner et al., 1994a). Materials and methods Bacterial strains, plasmids and growth conditions The E.coli strains used in this study were as follows: JM109 (recA1, supE44, endA1, hsdR17, gyrA96, relA1, thi, ∆(lacproAB) F9[traD36, proAB1, laqIq, lacZ∆M15]) (YanischPerron et al., 1985); HB101 (F–, hsdS20 (r-Bm-B), leu, supE44, ara14, galK2, lacY1, proA2, rpsL20, xyl-5, mtl-1, recA13, mcrB) (Boyer and Roulland-Dussoix, 1969); DSS301 (∆(lacpro), supE, thi, hsdD5, ∆dmsABC (Kmr)/F9 traD36, proA1B1, lacIq, lacZ∆M15) (Sambasivarao and Weiner, 1991); LE392 (F–, supE44, supF58, hsdR514, galK2, galT22, metB1, trpR55, lacY1); LE392∆unc (LE392, ∆uncIC). Both LE392 and LE392∆unc were gifts from Dr B.Rosen, Wayne State University, USA. The phagemid vector used for cloning of PCR products was pTZ18R (Mead et al., 1986). The expression vector for bacterial studies was pMS119HE (Strack et al., 1992) which is a pBR322 based plasmid. The expression vectors for eukaryotic tissue work were pSVL (Pharmacia Biotech) and pRc/CMV (Invitrogen). Bacteria cultures were grown in either Luria-Bertani (LB) broth medium or Terrific broth (TB) (Sambrook et al., 1989). Tissue culture experiments were performed using green monkey kidney cells (COS-1) and mouse liver hepatocytes (McA-RH777). Both cell lines have no detectable dystrophin and showed little background in Western blots or immunofluorescence microscopy to the dysp antibodies. Cell lines were maintained in 75 cm2 tissue culture flasks at 37°C in a humidified incubator. Cells were grown in Dulbecco’s modified Eagles medium (Gibco, BRL) supplemented with 10% fetal calf serum, 5 mM glutamate, 100 µg/ml streptomycin and 100 IU/ml penicillin. 286

Polymerase chain reaction mutagenesis and cloning of dmsC The technique of PCR mutagenesis was utilized both to extract the dmsC gene from the dmsABC operon and to generate the fusion. The primers used were as follows: 59 primer TWPCR6: ATATTCTAGAAGGAGAAATAATATGGGAAGTGGATGGCATG 39 primer TWPCR7: ATATGAGCTCTTACGGTTTGCCCGGGGTGTTGCGGCCGCGGCTTGGCGGGCTTGCGACGGCCATCCC The 39 primer (TWPCR7) provided the nucleotide sequence for the peptide PPSRGRNTPGKP (dysp, underlined) inserted before the stop codon as well as the sequence for a SacI site (in bold). The 59 primer (TWPCR6) provided sequence for an XbaI site (in bold), an E.coli consensus ribosome binding site (underlined) and an optimized intervening sequence before the initiating ATG. This technique has been found to work well for cloning and increasing the expression (Bishop and Weiner, 1993; Turner et al., 1994b,c). The restriction sites allowed for cloning into the phagemid pTZ18R and screened in the E.coli strain JM109. Only clones in which the reading frame of dmsC was in the opposite orientation to the lac-promoter could be obtained. This was likely a result of induction of the lac operon as a clone screening tool (blue/white selection), thereby producing expression of the lethal DmsC (see Results). The dmsC:dysp was then subcloned 39 to the tac-promoter into the expression vector pMS119HE creating pTW14 for expression in E.coli. This tac-promoter in this plasmid is under expression control by the lacIq gene which can be induced with isopropyl-β-Dthiogalactopyranoside (IPTG). By screening the clones using minimal media little leakiness of the tac-promoter was expected and expression of lethal genes would not interfere with the ability to isolate recombinants. The dysp tag was added to dmsC within the wild-type dmsABC operon cloned in the plasmid pDMS160 by excising out the C-terminus of dmsC:dysp from pTW14 by digesting with EcoRI (Klenow polymerized to blunt the ends) and then with BstEII. pDMS160 was digested with SalI followed by Klenow polymerase treatment to blunt the end and then digested with BstEII. The fragments were isolated by agarose gel electrophoresis and ligated together as BstEII (59)–blunt (39) fragments, generating dmsABC:dysp (pTW33). Growth curves and expression studies using dmsABC:dysp were performed using wild-type E.coli strain HB101 and the dms– strain DSS301. The dmsC:dysp was also cloned into the eukaryotic expression vector pSVL which uses the SV40 VP1 processing signals, generating pTW20. Plasmid pTW25 was constructed by subcloning dmsC:dysp into the eukaryotic expression vector pRc/ CMV which contains a neomycin selectable marker. This vector uses the human cytomegalovirus promoter and bovine growth hormone polyadenylation and termination sequences. Growth curves of E.coli strains Growth curves in E.coli were performed by inoculating a 250 ml aerobic side-arm Klett flask containing 25 ml LB and appropriate antibiotics for plasmid maintenance with 250 µl of an overnight culture in LB. Klett flasks were incubated at indicated temperatures (20, 30 or 37°C) in a water shaker bath set at 240 r.p.m. Anaerobic growth curves in glycerol–DMSO or glycerol–

Tagging and expression of DmsC

separated using 12.5% SDS–PAGE and then electroblotted on to nitrocellulose. DmsC:dysp was identified in Western blots using anti-dystrophin peptide antisera (α-dysp) which had been previously prepared and characterized by Milner et al. (1992). Detection of the protein bands was performed using the enhanced chemiluminescence detection system from Amersham using goat anti-rabbit IgG (H 1 L) horseradish peroxidase conjugate. Expression of DmsC:dysp in eukaryotic cells Plasmid pTW20 was transiently transfected into COS-1 cells using the procedure described by Kriegler (1990). The presence of expressed DmsC:dysp was confirmed using specific antidysp antibodies. A stable transfectant using pTW25 was also generated using McA-RH777 cells with selection using geneticin in at least three selective transfers.

Fig. 2. Growth curves of E.coli strains harboring the plasmid pTW14 which expresses DmsC:dysp. (A) Representative growth curves of wild-type strains HB101 or LE392 harboring pTW14 with no induction (s). Growth curves as a result of induction of dmsC:dysp with 1 mM IPTG at time zero (u) or after 2 h of growth (s). (B) Growth curve of F0F1 ATPase mutant LE392∆unc/pTW14 with no induction (u). Growth curves as a result of induction of dmsC:dysp with 1 mM IPTG at time zero (s) or after 2 h of growth (e).

Fig. 3. Western blot analysis of membrane fractions of HB101 expressing DmsC:dysp in LB broth media. In all cases the IPTG was added after 2 h of growth to a final concentration of 0.1 mM to induce expression. Duration time is the time after induction when the cells were harvested and lysed.

TMAO media were performed as by methods described previously (Bilous and Weiner, 1985, 1988). Expression and accumulation of DmsC in E.coli Expression in E.coli was performed using HB101(pTW14). Typically 250 ml cultures were used which had been given a 2% inoculant from an overnight culture in LB. Induction was usually at the beginning of log phase (after 2 h of growth at 37°C) with 0.1–1 mM IPTG. Incubation was continued at either 20 or 37°C for various times. Cells were harvested and lysed followed by fractionation into cytoplasmic and crude membranes by the protocol of Rothery and Weiner (1993). Protein concentrations were determined using the modified Lowry procedure (Markwell et al., 1978). Proteins were

Results Construction of DmsC:dysp fusion protein Utilizing PCR, an oligonucleotide which codes for a dystrophin specific peptide was fused to the 39 end of dmsC generating dmsC:dysp. The dmsC:dysp gene was cloned behind an inducible tac-promoter. This method also permitted the generation of a modified 59 end with an efficient ribosome binding site. This was necessary for the wild-type ribosome binding site of dmsC resides within the sequence of the dmsB gene. Unique restriction sites for cloning were added to both the 39 and 59 ends of the PCR-generated dmsC:dysp. Previous studies have shown that the vector used (pMS119EH) provides tight transcriptional control on the promoter (Turner et al., 1994b), which is important for expression of potentially lethal proteins. This vector was also chosen because it was previously shown that the expression level could be varied by titrating the IPTG levels (Taylor et al., 1994; Turner et al., 1994b,c). The dysptagged dmsC was also successfully cloned into the wild-type dmsABC operon generating dmsABC:dysp. Hydropathy analysis (not shown) suggests that the dysp tag is not expected to affect the profile of DmsC. The expected topology is shown in Figure 1B and is based on β-lactamase and alkaline phosphatase gene fusion analysis on DmsC (Weiner et al., 1993). Expression and accumulation of DmsC:dysp in E.coli The growth properties of HB101/pTW14 induced with IPTG at different times suggest that the expression of DmsC is highly lethal to E.coli (Figure 2A). The viability of the culture was found to decrease by half within 30 min after induction and completely lost after 60 min. Decreasing the IPTG level from 1 to 0.01 mM had little effect on the growth profile, suggesting that very low levels of DmsC:dysp accumulation are responsible for lethality. HB101/pTW14 cultures grown at 37°C in LB medium and induced for expression with IPTG after 2 h of growth resulted in DmsC:dysp becoming proteolysed, with the C-terminal dysp tagged end migrating at ~13 kDa as determined by Western blot analysis (results not shown). In order to alleviate this problem, we explored expression conditions of media, incubation temperature, time of induction and duration of induction. Upon examining the duration of incubation after induction, it was observed that little proteolysis occurred in the first 30 min after induction (Figure 3). However, the accumulated protein became completely proteolyzed within 2 h (Figure 3). Lowering 287

R.J.Turner et al.

the temperature to 20°C after induction did not rescue the culture or stop the proteolysis (Figure 3). The proteolysis could be inhibited if the culture was grown in the nutrient-rich medium Terrific broth and incubated at 20°C overnight (20 h), although both the yield of cells and accumulation of protein were fairly low (Figure 4). A comparison of the expression methods is presented in Table I. Expression and accumulation of DmsC:dysp in an E.coli unc mutant Since the anchor subunit DmsC is involved in the membrane distribution of protons between the periplasm and cytoplasm (Figure 1A), the lethality of DmsC accumulation may be the result of the protein acting as a proton channel. This would result in ATP depletion as the cell attempted to maintain the proton motive force. To alleviate this, an F0F1 ATPase mutant (unc) was employed. Expression of DmsC:dysp was found to be less harmful to cell proliferation when harbored in an E.coli unc– strain (Figure 2B). Cultures of LE392∆unc/pTW14 grew at the same rate as the control after an extended lag phase of between 300 and 400 min when IPTG was added at time zero along with the inoculant. Induction after 2 h gave rise to the same ‘instantaneous’ plateau of growth but after ~300 min the culture recovered. The greatest level of DmsC:dysp accumulation was observed using LE392∆unc/pTW14 grown in LB overnight at 20°C (Figure 5, Table I). These cultures were still viable whereas the wild-type LE392/pTW14 grown in either LB or TB had lost all viability. Expression of DmsC:dysp within the dms operon To provide a method of identification of the DmsC protein, which previously could not be readily monitored, DmsC:dysp

was expressed within the native operon as dmsABC:dysp. No lethality was observed during anaerobic growth with glycerol– DMSO or glycerol–TMAO media. Further, this DMSO reductase was functional and would support anaerobic respiratory growth. Examination of crude membranes by Western blot analysis, from DSS301/pTW34 grown anaerobically on glycerol–fumarate medium, showed the DmsC:dysp band migrating at ~28 kDa with no proteolysis observed (not shown). The activity of DmsABC:dysp was the same as the wild-type DmsABC as assessed by standard assay protocols (Cammack and Weiner, 1990; Sambasivarao and Weiner, 1991) using either benzyl viologen or dimethylnaphthoquinone as the electron donor and DMSO or TMAO as the substrate. The DmsABC:dysp also showed the same activity as wild type upon Triton X-100 solubilization. DmsC:dysp expression and localization in eukaryotic cell lines In order to alleviate the problem arising from the toxicity of DmsC accumulation in wild-type E.coli, expression in eukaryotic cells was explored. COS-1 cells were transiently transfected. Transfection efficiencies ranged between 1/25 and 1/50 cells. In contrast to wild-type E.coli, expression of DmsC:dysp did not appear to be toxic to the COS-1 cells. Confocal immunofluorescence microscopy of COS-1 cells expressing DmsC:dysp showed a distribution pattern consistent with the protein being associated with the endoplasmic reticulum (Figure 6). DmsC:dysp was identified within the membrane fraction from the transfected COS-1 cells using SDS–PAGE and anti-dysp. However, owing to the low transfection level, the yield of DmsC:dysp was low (Table I). A stable transfectant in McA-RH777 cells was used to see

Fig. 4. Western blot analysis of membrane fractions of HB101 expressing DmsC:dysp in the nutrient rich medium Terrific broth. Induction was after 2 h of growth with 0.1 mM IPTG. Duration time is the time after induction when the cells were harvested and lysed.

Fig. 5. Western blot analysis of the membrane fractions from E.coli wildtype strain LE392 and a mutant strain deficient in the F0F1ATPase (LE392∆unc) harbouring pTW14. Cultures were grown in LB medium. Cultures were induced after 2 h of growth with 0.1 mM IPTG. Duration time is the time after induction when the cells were harvested and lysed.

Table I. Comparison of DmsC:dysp expression with different induction conditions and expression hosts Hosta strain

Induction temperature (°C)

Duration of inductionb

Mediumc

DmsC:dysp observedd

Relative yielde

HB101 HB101 HB101 HB101 HB101 LE392∆unc LE392∆unc COS-1 McA-RH777

37 37 37 20 20 37 20 – –

2 2 30 20 20 2 20 – –

LB TB LB TB LB LB LB – –

2 deg 2 111 deg 1 2 2 111 1 1

0 0 25 20 0 0 60 1 1–3

h h min h h h h

aExpression plasmid used in bacterial hosts was pTW14. Eukaryotic tissue culture COS-1 bIn all cases cultures were incubated at 37°C for 2 h, at which time IPTG was added and

cells harbored pTW20 and McA contained pTW25. allowed to incubate at the designated temperature and time before

harvesting. Luria-Bertani broth; TB, Terrific broth. dObservation of DmsC:dysp on a Western blot originating from loading 25 µg of membrane protein. The presence of degradation products is designated with deg. eRelative yield was calculated by taking into consideration the intensity of the band on the Western blot and the yield of cells (mg of membrane protein). cLB,

288

Tagging and expression of DmsC

if the protein yield was strictly due to transfection levels. Immunofluorescence microscopy of the transfected McARH777/pTW25 cells showed the same pattern as with the COS-1 cells but with the immunofluorescence being less pronounced (not shown). Examination of cell-free extracts of McA-RH777/pTW25 by Western blot showed only trace amounts of DmsC:dysp. Discussion In order to obtain amplified levels of membrane protein expression, a number of variables must be considered: transcription, translation, protein folding, protein targeting, protein processing and assembly. Transcription is important to consider with regard to the promoter and promoter regulation on the expression plasmid being used. Not all strong promoters work equally well and empirical assessment of a number of different ones is needed in order to optimize expression levels. This was dramatically demonstrated with FucP where the λPL promoter worked very well and the T7 promoter fairly poorly (Gunn et al., 1994). Translation is a factor which includes codon usage and RNA secondary structural elements with regard to ribosome interaction. This was found to be the case in expressing the liver mitochondrial phosphate transporter in E.coli (Ferreira and Pedersen, 1992). Efficient membrane protein folding, targeting and assembly are dependent on many environmental factors as well as physiological functions. In our study we explored: growth temperature after induction, duration of growth between induction and harvesting, growth phase prior to induction and media conditions. We observed all these variables played a role in the amount of DmsC:dysp that was accumulated with the correct molecular weight. Many of the variables discussed above can be controlled or varied by the choice of host harboring the expression plasmid. The metabolism of the host can play a role in post-translational modifications, cellular targeting and processing (proteolysis). In our study on DmsC expression, the most important consideration was the host. Investigating the growth properties of

wild-type E.coli strain HB101 overexpressing DmsC:dysp demonstrated that the expression of DmsC is lethal to E.coli. Lethality of overexpression of membrane proteins is not new and is a common problem (Grisshammer and Tate, 1995). The lethality from DmsC expression could be overcome by utilizing an E.coli F0F1 ATPase (unc) mutant as the host. This suggests that the reason for the lethality observed in wild-type hosts is a result of DmsC acting as a proton motive force uncoupler. In the case of a protein leaking protons to the periplasm, the F0F1 ATPase would deplete the cells’ ATP levels trying to maintain the proton motive force. In order for DmsC to act as an uncoupler the protein would contain a form of a proton well or channel. This is consistent with the present model of the enzyme (Weiner et al., 1992; Rothery and Weiner, 1996). We have also employed the use of an unc mutant as the host for expression of the proton–drug efflux antiporter, EmrE as well as a number of its homologues (Paulsen et al., 1996). The results demonstrated that the use of an unc mutant gives 3–5-fold more protein accumulation over wild-type hosts (unpublished results). Poor Coomassie Blue staining of some integral membrane proteins on SDS gels has hindered their observation. This is especially true of DmsC, which is additionally difficult to monitor by immunodetection owing to its poor antigenicity. Identification of integral membrane proteins is compounded by the poor transfer efficiency to nitrocellulose or other blotting materials. DmsC is no exception and does not transfer well to nitrocellulose membranes (Sambasivarao et al., 1990). The present study on the expression and accumulation of DmsC was possible only because of the use of a specific epitope tag. This tag allowed for easy identification even with poor transfer to blotting membranes. Considerations in choosing an epitope tag include availability of the antibody, size of the epitope (number of units of the epitope required for maximal antibody interaction), specificity of the available antibodies (with regard to host background reactions) and antigenicity of epitope (required titer of antibody for detection). With these considerations, antibody affinity columns that bind to the epitope tags

Fig. 6. Confocal immunofluorescence microscopy of COS-1/pTW20 cell expressing DmsC:dysp. The fluorescence distribution pattern suggests that DmsC:dysp is associated with the endoplasmic reticulum.

289

R.J.Turner et al.

can be used to purify the membrane protein. Alternatives to epitope tagging include the hexahistidine and biotinylation tags, which are useful primarily for purification purposes (Grisshammer and Tate, 1995; Phizicky and Fields, 1995). Generally the concept of heterologous protein expression systems find E.coli to be the default choice for many membrane proteins, although a number of alternatives are arising which include yeast (Sekler et al., 1995) and insect baculovirus systems (Domingo and Trowbridge, 1988; Grisshammer and Tate, 1995). In our study we examined the possibility of expressing our lethal prokaryotic protein in tissue cultures. We found that DmsC:dysp could be expressed and localized to the endoplasmic reticulum. Additionally, DmsC:dysp was not lethal in the cell lines used. Although it was successful, the yield of protein was not at levels which would be amenable for further study. We have expressed the integral membrane anchor subunit, DmsC, of DMSO reductase in both prokaryotic and eukaryotic systems. Identification of the DmsC was by the utilization of a unique tagging of this protein with a dystrophin-specific sequence. The successful expression of DmsC represented here will allow, for the first time, the ability to carry out structure and function studies on this subunit independent of DmsA and DmsB. It is likely that in addition to DmsC, other integral membrane anchor subunits of respiratory enzymes would also be difficult to express on their own. The data obtained in this study provide methodologies to overcome lethality problems associated with the overexpression of integral membrane proteins such as DmsC. The work performed here may have general usefulness in the study of membrane proteins for identification, targeting, overexpression and in purification. Acknowledgments This work was funded by the Medical Research Council of Canada and the Muscular Dystrophy Association of Canada. J.H.L. was a recipient of an Alberta Heritage Foundation for Medical Research summer studentship. We thank Dr Kimberly Burns for dysp antisera and useful suggestions on its use. We also thank Dr Michal Opas, University of Toronto, for taking the confocal microscopy images.

References Bilous,P.T. and Weiner,J.H. (1985) J. Bacteriol., 162, 1151–1155. Bilous,P.T. and Weiner,J.H. (1988) J. Bacteriol., 170, 1511–1518. Bishop,R.A. and Weiner,J.H. (1993) Eur. J. Biochem., 213, 405–412. Boyer,H.W. and Roulland-Dussoix,D. (1969) J. Mol. Biol., 41, 459–472. Cammack,R. and Weiner,J.H. (1990) Biochemistry, 29, 8410–8416. Domingo,D.L. and Trowbridge,I.S. (1988) J. Biol. Chem., 263, 13386–13392. Ferreira,G.C. and Pedersen,P.L. (1992) J. Biol. Chem., 267, 5460–5466. Grisshammer,R. and Tate,C.G. (1995) Q. Rev. Biophys., 28, 315–422, and references cited therein. Gunn,F.J., Tate,C.G. and Henderson,P.J.F. (1994) Mol. Microbiol., 12, 799–809. Kriegler,M. (1990) Gene Transfer and Expression: A Laboratory Manual. Stocton Press, New York. Markwell,M.A., Haas,S.M., Bieber,L.L. and Tolbert,N.E. (1978) Anal. Biochem., 87, 206–210. Mead,P.A., Szczesna-Skorupa,E. and Kemper,B. (1986) Protein Engng, 1, 67–74. Milner,R.E., Busaan,J.L. and Michalak,M. (1992) Biochem. J., 288, 1037– 1044. Paulsen,I.T., Skurray,R.A., Tam,R., Saier,M.H.,Jr, Turner,R.J., Weiner,J.H., Goldberg,E.B. and Grinius,L.L. (1996) Mol. Microbiol., (19), 1167–1175. Phizicky,E.M. and Fields,S. (1995) Microbiol. Rev., 59, 94–123. Rothery,R.A. and Weiner,J.H. (1993) Biochemistry, 32, 5855–5861. Rothery,R.A. and Weiner,J.H. (1996) Biochemistry, 35, 3247–3257. Rothery,R.A., Simala Grant,J.L., Johnson,J.L., Rajagopalan,K.V. and Weiner,J.H. (1995) J. Bacteriol., 177, 2057–2063.

290

Sambasivarao,D. and Weiner,J.H. (1991) J. Bacteriol., 173, 5935–5943. Sambasivarao,D., Scraba,D.G., Trieber,C. and Weiner,J.H. (1990) J. Bacteriol., 172, 5938–5948. Sambrook,J., Fritsch,E.R. and Maniatis,T. (1989) Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sekler,I., Kopito,R. and Casey,J.R. (1995) J. Biol. Chem., 270, 21028–21034. Strack,B., Lessl,M., Calendar,R. and Lanka,E. (1992) J. Biol. Chem., 267, 13062–13072. Taylor,D.E., Hou,Y., Turner,R.J. and Weiner,J.H. (1994) J. Bacteriol., 176, 2740–2742. Traxler,B., Boyd,D. and Beckwith,J. (1993) J. Membr. Biol., 132, 1–11. Turner,R.J., Burns,K., Busaan,J., Michalak,M. and Weiner,J.H. (1994a) Biochem. Cell Biol., 71, Axii. Turner,R.J., Weiner,J.H. and Taylor,D.E. (1994b) Biochem. Cell Biol., 72, 333–342. Turner,R.J., Weiner,J. H. and Taylor,D.E. (1994c) Microbiology, 140, 1319– 1326. Weiner, J.H., Rothery,R.A., Sambasivarao,D. and Trieber,C.A. (1992) Biochim. Biophys. Acta, 1102, 1–18. Weiner,J.H., Shaw,G., Turner,R.J. and Trieber,C.A. (1993) J. Biol. Chem., 268, 3238–3244. Yanisch-Perron,C., Vieira,J. and Messing,J. (1985) Gene, 33, 103–119. Received July 25, 1996; revised October 4, 1996; accepted October 15, 1996