THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 258, No. IS, Issue of September 25, pp. 11211-11218,1983 Printed in U. S.A.

Construction of penPAI, Bacillus licheniformis 749/C /3-Lactamase Lacking Site for Lipoprotein Modification EXPRESSION IN ESCHERICHIA COLI AND BACILLUS SUBTILIP

(Received for publication, April 4, 1983)

Peter S. F. MezesS, Wu Wang, Edward C. H. Yehg, and J. Oliver Lampen From the Waksman Institute of Microbiology, Rutgers Uniuersity, Piscataway, New Jersey 08854

Membrane-bound penicillinases in Gram-positive bacteria are glyceride-cysteine lipoproteins (Nielsen, J. B. K., and Lampen, J. 0. (1982) J. Biol. Chern. 257, 4490-4495) that can be, but are not necessarily, inter- mediates in formation of exocellular enzymes. We have now deleted from the signal region of the Bacillus licheniformis 749/C B-lactamase gene (penP) 15 base pairs that code for Ala-Leu-Ala-Gly-Cys. This se- quence includes the Cys residue that undergoes lipo- philic modification and the site of cleavage by signal peptidase and is well conserved in diverse prokaryotic lipoproteins. In the deletion gene, penPA1, the remain- ing Cys is preceded by 8 hydrophobic residues instead of 14 for the original modification site. PenPAl has been cloned in Escherichia coli and Bacillus subtilis and its expression and products compared with penP. Penicillinase synthesis by penPA1 clones was greater than or equal to amounts formed by penP clones or B. licheniformis. Lipoprotein production from penPA1 was very low in either host and appears physiologically insignificant. In E. coli carrying penPA1, 25% of the penicillinase was released into the periplasm as a proc- essed form. The remainder was a membrane-associated form of translation product size and was oriented to the periplasm. In mid-log cultures of B. subtilis carry- ing penPA1, the translation product and 2 protease- shortened species were present both in the cytoplasm and on the outer surface of the membrane. Release began at stationary phase and was dependent on the presence of enzyme processing to exo-small and a pH value >7.5. We conclude that the shortened lipophilic sequence of the penPA1 prepenicillinase is adequate for transfer of the nascent chain through the mem- brane.

There is increasing interest in utilizing Escherichia coli or various Bacillus species for the production of genetically en- gineered proteins. The DNA of the promoter-signal sequence region of secreted proteins has been fused to foreign genes to facilitate secretion of the product as a hybrid or some proc- essed form (1-3). Concomitant with these developments, the need to conveniently isolate the foreign gene products on a

* This work was supported by grants from the National Science Foundation (PCM81-16888), Miles Laboratories, Inc., and the Charles and Johanna Busch Fund. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Recipient of a Waksman-Merck Postdoctoral Fellowship.

Inc., Elkhart, IN 46515. Present address, Enzyme Products Division, Miles Laboratories,

large scale has prompted studies on the mechanism of protein translocation by these organisms (4-8). In this regard, we have focused attention on the p-lactamases (penicillin amido- P-lactamhydrolase, EC 3.5.2.6) as models of secreted proteins.

Bacillus licheniformis 749/C produces two extracellular pen- icillinases, exo-large and exo-small (9, 10) as well as a mem- brane-bound lipoprotein (11). The membrane form belongs to a diverse group of prokaryotic lipoproteins that includes the major outer membrane protein of E. coli (12). These have in common a modifiable cysteine residue surrounded by a con- served sequence in the signal region of the protein. In the processing events leading to membrane attachment, the cys- teine becomes NHp-terminal and is involved in the unique thioether linkage to a diacyl glyceride. The amino group of the cysteine becomes blocked by amide linkage to fatty acid. As the pH of growing cultures of B. licheniformis becomes slightly alkaline, proteases are activated and release most of the membrane-bound penicillinase as the hydrophilic exo- forms (9).

The penP gene has been sequenced (13) and also cloned in E. coli (14, 15) and B. subtilis (15). Expression was obtained in both hosts. In this paper we assess the role of the lipopro- tein in secretion by introducing a specific 15 bp’ deletion in the signal sequence of penP in order to remove the cysteine residue involved in the lipophilic modification. The resulting gene, penPAl, was cloned in E. coli and B. subtilis and the nature and disposition of the altered penicillinase products compared to the products of the penP gene in the same hosts.

EXPERIMENTAL PROCEDURES~

RESULTS

Preparation of penPAl in E. coli-The penPAl gene in pRW84 was prepared by deleting a 15-bp segment in the signal sequence region of penP that corresponded to five amino acids in the consensus sequence for lipophilic modifi- cation (12), including the cysteine involved in the unique thioether linkage to diacyl glyceride. Fig. 4a depicts the intact penP gene in pRW83 and the restriction sites involved in the construction and verification of the structure of pRW84 (see

The abbreviations used are: bp, base pair; SDS, sodium dodecyl sulfate; kb, kilobase pair,

* Portions of this paper (including “Experimental Procedures,” and Figs. 1-3, 6, and 7) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biolog- ical Chemistry, 9560 Rockville Pike, Bethesda, MD 20814. Request Document No. 83M-877, cite the authors, and include a check or money order for $4.80 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

11211

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11212 Secretion of Penicillinase Lacking Lipophilic Modification

b.

2 2 2 0 29

-Vol Alo Asn-

5 " G T C G C T A A C -

3 " C A G C G A T T G - C

FIG. 4. Deletion of Ala-Leu-Ala-Gly-Cys from signal se- quence of penP. a, penP on 1.4-kb HindZZZ-BarnHI fragment. Restriction sites utilized for the construction and structure verifica- tion of penPA1 in pRW84 are shown. The dashed area represents the signal sequence region of the gene. b, an expansion of the signal region; the specific deletion is bored. Numbering of the corresponding amino acids is based on fMet being 1. c, DNA and amino acid sequence resulting from the deletion.

Miniprint). In Fig. 4b, the region involving the deletion is expanded and the specific bases eliminated are boxed. Fig. 4c presents the resulting nucleotide sequence (in pRW84) for penPAl. The amino acid residues numbered 23-27, corre- sponding to Ala-Leu-Ala-Gly-Cys of the original penP gene product (13), have been deleted. The deletion does not alter the reading frame in penPAl from that in penP. Apart from the deletion of Cys 27 which is the eventual NHz-terminus of the membrane lipoprotein, the hydrophobic stretch of the signal sequence has been reduced from 16 to 11 residues, and the possibility arises that a new signal peptidase or protease site(s) has been generated in the modified translation product.

Evidence for the 15-bp deletion was provided by restriction mapping of the HindIII-BamHI DNA fragment containing the penPA1 gene. In the deletion process the FnuDII site at position 334 (Fig. 4b) was destroyed, as well as the HhaI sites at positions 336 and 349. Thus, a FnuDII digestion gave only 2 bands of 477 and 905 bp, while H h I treatment left the DNA uncut. Direct evidence for the deletion was provided by DNA sequencing. A 971-bp PstI-BamHI DNA fragment from pRW84 was cloned into the corresponding sites of M13mp8 and colorless plaques selected from infected cells of E. coli JM103. DNA sequencing from the PstI site agreed with the published sequence of the B. licheniformis 749/C gene (13) and verified the deletion of the 15 bp boxed in Fig. 4b.

Cloning of penPA1 in B. subtilis-The hybrid plasmid pRWlOl contains a 1.15-kb DNA fragment from pBR322 that includes the origin of replication, as well as pUBllO linearized at the PuuII site, and the penP gene, carried on the 1.397-kb HindIII-BanHI DNA fragment of pRW83 (Fig. 3, Miniprint). The BarnHI site was recovered when the filled-in BarnHI terminus of the 1.397-kb DNA fragment was ligated next to a G nucleotide from the blunt-ended PuuII cut of pUB110. pRW185, which containspenPA1, was constructed by remov- ing a 623-bp PstI-ClaI DNA fragment from pRWlOl and introducing the comparable 608-bp DNA fragment from pRW84. Except for the 15-bp deletion in the penPAl gene, pRW185 is identical to pRW101.

Growth and Production of Penicillinase in E. coli-E. coli RR1 cultures transformed with pRW83 or pRW84 were grown simultaneously, and aliquots were taken at regular intervals, sonicated, and assayed for total penicillinase activity. The growth of each clone was normal with no cell lysis; however,

total penicillinase production by the penPAl clone was 3-5 times greater than that produced by its penP counterpart. The amount of penicillinase produced in E. coli with pRW83 is comparable to that produced in B. licheniformis 749/C itself. The reason for the increased production of p-lactamase with pRW84 has not been established, but clearly the plasmid is not toxic to the host organism.

Distribution of Penicillinase in E. coli-The distribution of penicillinase activity in late log cultures of E. coli carrying penP or penPA1 is presented in Table I. Neither culture had any activity in the medium. Practically all of the penicillinase produced by the penP clone was in the cell envelope fraction and required detergent (Triton X-100) to be solubilized the small amount of activity (4%) in the periplasmic fraction could be there as a result of sloughing from the membrane or contamination with membrane fragments. After the shock procedure, the damaged bacteria were pelleted and suspended in 50 mM Tris-HC1, pH 7.5. Direct assay of this suspension for penicillinase gave the same activity as that of sonicated intact cells. Furthermore, trypsin treatment of such cells released all the activity as a solubie penicillinase the same size as that produced from the lipoprotein membrane penicil- linase of B. licheniformis 749/C protoplasts (11).

The distribution of p-lactamase in the penPAl clone differs from its penP counterpart in that there is a sizeable peri- plasmic fraction (24%) and Triton X-100 is not required to solubilize the membrane enzyme (Table I). After sonication of a culture sample, 83% of the total activity was recovered in the soluble fraction. Most of the remaining activity was recovered and solubilized when the sonicated envelope frac- tion was treated with buffer containing 1% Triton X-100 or simply suspended in 50 mM Tris-HC1, pH 7.5, without deter- gent and resonicated. When cells carrying penPAl were shocked to remove their periplasmic penicillinase, the bulk of the bound penicillinase (80-85%) was available to substrate and was released as a smaller form by trypsin (results not shown). Alkaline phosphatase was used as a marker enzyme for periplasmic proteins. It was present in the shockates from cells grown in low phosphate medium (40) but not in cells cultured in L-broth. Glucose-6-phosphate dehydrogenase was found only in the cytoplasmic fractions; none could be de- tected in the shockates. These results indicate that the peri- plasmic fraction containing the processed penPAl penicilli- nase was not contaminated by cytoplasmic proteins. We have

TABLE I Distribution of penicillinase produced in E. coli RRI by penP and

penPA1 Total penicillinase activity was determined on culture samples by

sonication and direct assay of the suspension. The periplasmic frac- tion was prepared according to the shock method described. The shocked cells were resuspended and sonically disrupted, and the envelope was pelleted at 100,000 X g. The supernatant represented the soluble fraction. The pelleted envelopes were suspended in buffer (50 mM Tris-HC1, pH 7.5) containing 1% Triton X-100, mixed, and centrifuged again at 100,000 X g for 40 min. The extract contained the solubilized membrane-bound penicillinase in the case of penP, while the enzyme recovered in this fraction for the penPAI clone represented penicillinase loosely bound to membrane. The latter activity could also be extracted by repeated sonication of the envelope fraction in buffer without detergent.

( ~ R w 8 3 ) % (pRW84) penP wnPA1 % total

units rnl" units rnl" Total 7,600 100 22,000 100 Periplasm 310 4.1 5,200 24 Soluble after 0 0 13,000 59

Membrane 6.400 84 2,900 13 sonication

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Secretion of Penicillinase Lacking Lipophilic Modification 11213

no evidence for any penicillinase in the cytoplasm itself for either the penP or penPA1 clones of E. coli.

Nature of the Penicillinase Products-When cultures of E. coli carrying pRW83 (penP) were incubated with [3H]palmi- tate, a labeled band having the mobility of B. licheniformis membrane penicillinase was obtained as the sole immunopre- cipitated product (Fig. 5, lunes 1 and 2) . This is in agreement with earlier observations (11, 14) that the penicillinase pro- duced in E. coli with the penP gene on a X lysogen (Xpen) is a lipoprotein that contains the glyceride-cysteine thioether modification and resembles the structure of the major outer membrane lipoprotein of E. coli (41).

Two forms of penicillinase are present in E. coli trans- formed with pRW84 (penPA1) (Fig. 5, lune 4) . The larger is associated with both the envelope and soluble fractions (after sonication of either intact or osmotically shocked cells) and has the same mobility as the translation product (lune 3) produced from pRW84 in the E. coli in uitro transcription- translation system of Zubay (42). The periplasmic fraction contained only the smaller form (lune 8). A culture labeled with [3H]palmitate gave no 3H-labeled band on the autoradi- ogram of an immunoprecipitate of total penicillinase after 2 days exposure (lune 5 ) . Exposure of the same sample for 2 weeks gave 2 faint bands, one of translation product size and one slightly smaller (lune 6 ) . These bands may be the result of metabolic recycling of 3H or of a physiologically insignifi- cant amount of modification of Cys 21, or Ser 20 (Fig. ll), in the translation product which has been partially processed (see “Discussion”). Labeling with [35S]cysteine indicated that Cys 21, the only such residue of the protein, is present in the full translation product (lune 7), but not in the smaller periplasmic enzyme (lanes 8 and 9). Consequently, the latter form cannot extend beyond Val 22 at its NH, terminus. When 10 pg of the envelope-associated product was treated with 100 pg of the periplasmic protein fraction (representing 1.0 ml of

31 K-

1 2 3 4 5 6 7 8 9 FIG. 5. Nature of penicillinase products in E. coli. Samples

were electrophoresed in 10% polyacrylamide gels containing 0.1% SDS. All lanes except lane 3 represent immunoprecipitates. Lane I, membrane penicillinase from B. licheniforrnis 749/C stained with Coomassie blue. The lipoprotein has an approximate molecular weight of 32,000 but has the same, or slightly faster, mobility that the hydrophilic exo-large form does on 10% polyacrylamide gels containing SDS (11). Lane 2, autoradiogram of [3H]palmitate-labeled total penicillinase obtained after sonication and Triton X-100 solu- bilization of the envelope fraction from E. coli carrying the penP gene. Lane 3, [3SS]methionine-labeled products of an in vitro tran- scription-translation system with pRW84 as template. The penPAI translation product a t 34 kDa is the most prominent band. Lane 4, [35S]methionine-labeled totalpenPAI penicillinase. All the remaining lanes also represent products from the penPAI clone. Lanes 5 and 6, autoradiograms of [3H]palmitate labeling of total penicillinase. In general, gels were dried, enhanced if [3H]palmitate was used, and radioautographed for 2 days, except for lane 6, which was a 2-week exposure. Lane 7, autoradiograms of [?3]cysteine-labeling of total penicillinase. Lane 8, Coomassie blue staining of periplasmic fraction obtained from an [%3]cysteine-labeled culture. Lane 9, autoradiogram of lane 8.

original late log culture) for 60 min at 37 “C, no processing to the periplasmic form was observed.

Characterization of the Periplasmic Penicillinase-The pu- rification of the periplasmic penicillinase produced bypenPA1 in E. coli RR1 was achieved in two steps as described under “Experimental Procedures” (see also Figs. 6 and 7, Miniprint). The enzyme was >95% pure, giving a single band of approx- imately 31 kDa upon SDS-polyacrylamide gel electrophoresis (Fig. 7, lune 4 ) . We compared its mobility with the three known B. licheniformis penicillinases on a 15-22% gradient polyacrylamide gel containing 8 M urea and 0.1% SDS. This gel system resolves the membrane lipoprotein penicillinase from exo-large, each form having a mobility consistent with its molecular weight. The periplasmic form from penPA1 ran just slightly faster than exo-large did (Fig. 8, lunes 2 and 5 ) . The membrane-associated penicillinase of the penPA1 clone is included for comparison (Fig. 8, lune 1 ).

A sample of the purified periplasmic penicillinase (300 pg) was sequenced by the automated Edman procedure. The NHz terminus was found to be blocked, and conventional deblock- ing methods for acyl groups proved to be futile. The apparent size of the protein by gel electrophoresis suggested that its NHz terminus was near Gln 36, and this residue could have cyclyzed in the isolation and purification steps to pyrogluta- mate (43) which would block sequencing. When the immobi- lized penicillinase was treated with pyrrolidone carboxylyl peptidase, an enzyme acting on pyroglutamate to yield glu- tamic acid (44), still no sequence data could be obtained. However, whenever a pyroglutamate residue is followed by proline, as in this section of the leader sequence, the peptidase does not catalyze the reaction to give glutamic acid (43). Finally, after mild base treatment (2 N KOH, 22 “C, 12 h), which hydrolyzes the amide bonds of pyroglutamate at a relatively faster rate than the amide bonds of the polypeptide backbone (43), a glutamic acid residue was uncovered at the NH2 terminus. It was followed by proline. These data, along with the relative size in the gradient gel (Fig. 8) , indicate that the periplasmic form ofpenPA1 penicillinase begins with Gln 36 (numbering according to the penP gene product) and is one residue shorter than the exo-large form found in B. licheniformis cultures (Fig. 11).

Growth and Penicillinase Production in B. subtilis withpenp and penPA1-B. subtilis carrying pRWlOl or pRW185 was grown in 2% CH/S medium for direct comparison of its growth and total penicillinase production with that of B.

M assoc - peri -

E. coli B. licheniformis

- MLP exo-large - exo-small

-

1 2 3 4 5

FIG. 8. Comparison of periplasmic and membrane-associ- ated penicillinase from E. coli containing penPAZ with B. lichenitorrnis penicillinases. The immunoprecipitated samples were electrophoresed on a 15-22% gradient SDS-polyacrylamide gel containing 8 M urea and stained with silver. Lane I, membrane- associated (M assoc) penicillinase obtained from E. coli. Lanes 3, 4, and 5, E. licheniforrnis penicillinases: exo-large, membrane lipopro- tein (MLP) , and exo-small, respectively.

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11214 Secretion of Penicillinase Lacking Lipophilic Modification

FIG. 9. Growth, pH, and penicil- pH linase production by B. lichenifor- mis 749/C and by B. subtilis BD170 transformed with pRWlOl or pRW185. of 2% CHIS Cultures medium were starting grown at in pH 15 6.5. ml 1 .(:IF -

Growth (0) and pH (A) were measured at the indicated times. Aliquots (100 pl) f 400 were removed and assayed for total (0) e and secreted (0) penicillinase. a, B. lich- eniformis 749/C; b, B. subtilis trans- formed with pRWlOl (penP); c, B. sub- tilis transformed with pRW185

Y

0

0 5 15 25

(penPAI).

licheniformis 749/C. The total amount of penicillinase pro- duced in B. subtilis with either penP or penPA1 was greater (by "3000 units ml") than that produced in B. licheniformis 749/C. In addition, the B. subtilis clones grew somewhat more slowly than B. licheniformis (Fig. 9) and underwent some lysis in late log phase. B. licheniformis secreted penicillinase more quickly than either of the B. subtilis clones. The products found in the medium of 12-h cultures (-10% of the total penicillinase activity) of B. subtilis carrying either penPA1 or penP indicate that sloughing from the membrane or lysis is occurring. By 24 h the penP and penPA1 clones had released 85-90% of their total penicillinase, but only 70% had been released by B. licheniformis (Fig. 9). In all cases, cell-bound enzymes were more rapidly released if 1% CH/S medium was used, or if the initial pH was >7.0. For example, when the starting pH of the 2% CH/S medium was 7.2, all three cultures had released 60-70% of their total penicillinase by 12 h, at which point the pH of the cultures was 7.7. Lysis could be prevented by growing the B. subtilis clones in 2% CH/S medium supplemented with sucrose (15% w/v) as osmotic support. Secretion is clearly dependent upon proteases with an alkaline pH optimum, as previously observed for B. lich- eniformis (9).

Characterization of Penicillinase Products in B. subtilis- The penicillinase in the supernatants of 24-h cultures of the penP and penPA1 clones or of B. licheniformis was entirely in the exo-small form as determined by immunoprecipitation and SDS-polyacrylamide gel electrophoresis. The initial se- creted product from the penP clone was exo-large at substan- tially decreased levels when compared to B. licheniformis; exo- small was first observed at the beginning of stationary phase. Treatment of 10-h cells of B. licheniformis or of the penP clone with lysozyme to promote gentle lysis released a mixture of exo-large and membrane lipoprotein penicillinases, the former predominating with B. licheniformis, the latter with the penP clone (Fig. 10, lanes 1 and 2). The penPA1 clone, similarly treated, released three distinct penicillinase products which were immunoprecipitable from the lysis supernatant. The protein corresponding to the slowest band was translation product size (cf. with the membrane-associated product from E. coli penPA1 (Fig. 10, lanes 5 and 6 ) . The other 2 bands represented processed products which were still larger than lipoprotein membrane penicillinase as determined by electro- phoresis on the 15-22% polyacrylamide gradient gel (Fig. 10, lanes 4 and 5).

Protoplasts prepared from 10-h cultures of B. subtilis con- taining either penP or penPA1 had 55-60% of the total penicillinase activity available on the outer membrane sur- face, when assayed in osmotically supported media. This activity could be removed in both cases as a soluble, hydro- philic product the size of exo-small, by trypsin treatment. In the case of the penP clone the remainder of the total activity had been secreted as exo-large, while the remaining 45% of the total activity of the penPA1 clone was released upon

0 5 15 25

Time, h

exo-lorge MLP =

t

1 2 3 4 5 6 FIG. 10. Penicillinase products released by B. subtilis con-

taining penP or penPAI and by B. licheniformis 749/C after treatment with lysozyme. Cells from 8-h cultures were lysed by treatment with lysozyme in osmotically unsupported media. After centrifugation, penicillinase was immunoprecipitated from the super- natants, electrophoresed in a 15-22% gradient polyacrylamide gel containing 8 M urea and stained with silver. Lane I, B. licheniformis. Lane 2, B. subtilis penP clone. Lane 3, exo-large standard. Lane 4, membrane lipoprotein (MLP) standard. Lane 5, B. subtilis penPAI clone. Lane 6, membrane-associated (M assoc) penicillinase from E. coli penPAI clone.

sonication of the trypsin-treated protoplasts. This penicilli- nase was translation product size and must have been present in the cytoplasm or on the inner face of the plasma membrane of B. subtilis.

Labeling of the B. subtilis clones with [3H]palmitate gave results similar to those obtained with the E. coli clones. Palmitate was incorporated into the membrane penicillinase of the penP clone, while the penPA1 clone had a negligible incorporation (<2%) in its penicillinase products.

DISCUSSION

B. licheniformis 749/C penicillinase belongs to a class of Gram-positive proteins that have a membrane-bound form which is a glyceride-cysteine lipoprotein. This form can be a precursor of the exopenicillinase, but does not appear to be an essential intermediate (45). From the available evidence, however, it was still possible that the lipoprotein is a transient intermediate a t a localized site of synthesis where it is not in equilibrium with the bulk of the membrane-bound material. As a means of determining the role of the lipoprotein in secretion we have constructed a penicillinase gene that lacks 15 bp coding for residues 23-27 of prepenicillinase (Fig. 11) which include the Cys residue at which the modification occurs and the site of cleavage by signal peptidase (46). Penicillinase production and distribution were examined in clones of E. coli and B. subtilis carrying the shortened penP gene (penPA1) on the appropriate plasmid vectors. These were compared with the results obtained for penP cloned in the same hosts and with penicillinase production in the par- ent, B. licheniformis.

The amount of penicillinase produced in cultures carrying

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Secretion of Penicillinase Lacking Lipophilic Modification 11215 5 I

Met-Lys-Leu-Trp-Phe-Ser-Thr-Leu-Lys-Leu-Lys

b e x o - l a r g e -

I deletion

+ex0 - large rn

+penPAI periplasm - oxo - small I--

FIG. 11. Signal regions and processed forms of penicillinase from peep and penPA1. The asterisk denotes the cysteine residue that undergoes lipophilic modification and becomes the NH, terminus of membrane penicillinase when cleaved by signal peptidase (S.P.). Leu 24, Cys 27, and Asn 30 are underlined to indicate the position and identity of the most highly conserved residues in the consensus sequence for lipoprotein formation. The corresponding positions are underlined for the penPA1 penicillinase indicating Cys 21 as a possible site for 1iDoDhilic modification. Numbering is from M e t = 1. The various processed forms are indicated as well as the site " Y

of cleavage by trypsin.

the penPAl gene was greater than or equal to that produced with the penP gene or in the parent itself. The transcriptional promoters of the B. licheniformis 749/C penP gene have been determined (47) and are apparently operational in the plas- mids used in this study. The amount of penicillinase formed in the host systems is not proportional to the gene copy number, which is high, since nearly the same amount of enzyme is produced from the single chromosomal copy in B. licheniformis. The increased production of penicillinase in the penPAl clone of E. coli RR1 over that of the penP clone indicates that processing to the membrane lipoprotein in the latter may be a limiting factor on expression.

E. coli carrying penP contained only the glyceride-cysteine lipoprotein which was oriented into the periplasm and was not cleaved by protease. The product frompenPAl was trans- ferred across the membrane, remaining loosely attached, and was removable by sonicaton. I t was cleaved by an unidentified protease releasing 20-25% of the total penicillinase to the periplasmic space as a form one residue shorter than the exo- large initially released in B. licheniformis (Fig. 11). The en- zyme that remained membrane-associated had the same mo- bility on SDS-polyacrylamide gels as the in uitro translation product of penPA1.

When penP was cloned in B. subtilis, the major products were the membrane-bound lipoprotein and the exo-large form (as in B. licheniformis); late in growth exo-small appeared. If penPAl was present, three forms including the full length translation product and material shortened at the NH2 ter- minus were initially detected. They appeared to be located both in the cytoplasm and on the outer aspect of the plasma membrane. Secretion of penicillinase into the medium was delayed, but by early stationary phase essentially all the penicillinase was in the medium as exo-small. Secretion in this system occurred only when the processing enzyme yield- ing exo-small was present in the culture supernatant (late logarithmic phase) and the pH was >7.5. When cultures of B. subtilis containingpenPA1 (orpenP) were grown at pH values starting above neutrality, release of exo-small was enhanced

indicating that this protease is activated at higher pH values. Exo-large was not detected as a product of intact cells, but was observed as an intermediate form when lysis of some cells occurred. Presumably, the processing enzyme yielding exo- large acts on the membrane-associated products of cells con- taining penPAl only after lysis releases them from the mem- brane.

In the penPA1 clone of B. subtilis only about 55% of the cell-bound material appeared to be on the outer surface of the membrane. The remainder was not retained by the membrane during gentle lysis with lysozyme in hypotonic medium. It was either free in the cytoplasm or very loosely attached to the inner surface of the plasma membrane. We conclude that the shortened lipophilic sequence is adequate in E. coli for transfer of the nascent chain through the membrane without intervening formation of the lipoprotein; however, in B. sub- tilis entry into the membrane may be delayed or abortive. We cannot say whether or not the internal enzyme was eventually secreted. It may have been degraded while synthesis and secretion continued.

The remaining Cys residue in the penPAl signal could conceivably be an acceptable site for the thioether type of modification. Nevertheless, only traces of palmitate labeling were found in cells containingpenPA1; the degree of lipophilic modification at the remaining Cys was physiologically insig- nificant. For known bacterial thioether lipoproteins the con- sensus sequence is flanked by Leu and Asn, and these are in the correct positions in penPAl (Fig. 11). However the large nonpolar residues Phe and Val may be unacceptable substi- tutes for the usual small intervening residues. In addition, Cys 21 may be buried in the membrane bilayer and inacces- sible to the processing enzymes.

It is likely that with penPA1 cloned in B. subtilis part of the site for cleavage to exo-large is buried in the plasma membrane. This is indicated by the large amounts of exo- large obtained from these cells after lysis by lysozyme (Fig. 10, lunes land 2 ) . The processing enzyme may also be less active in B. subtilis than in B. licheniformis. We conclude that

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11216 Secretion of Penicillinase Lacking Lipophilic Modification

the lipoprotein structure enhances secretion initially by virtue of its orientation at the membrane surface and its resultant accessibility to processing enzymes. For the convenient iso- lation of genetically engineered products, B. subtilis may prove to be useful for immobilizing foreign gene products at the outer surface of the membrane when the signal sequence region of penPAl would be used to direct such localization and a delay in secretion. Optimum conditions for release could then be determined in each case.

Acknowledgments-We wish to thank Dr. Brammar for a generous gift and Drs. Nielsen and Hussain for helpful discussions. The tech- nical assistance of Julia Sohm is gratefully appreciated. We also thank Eli Lilly and Co. for a gift of benzyl penicillin and The Upjohn Co. for spectinomycin.

REFERENCES 1. Palva, I., Sarvas, M., Lehtovaara, P., Sibakov, M., and Kaiiri-

ainen, L. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,5582-5586 2. Talmadge, K., Stahl, S., and Gilbert, W. (1980) Proc. Natl. Acad.

Sci. U. S. A. 77,3369-3373 3. Talmadge, K., Kaufman, J., and Gilbert, W. (1980) Proc. Natl.

Acad. Sci. U. S. A. 77,3988-3992 4. Inukai, M., and Inouye, M. (1983) Eur. J . Biochem. 130 , 27-32 5. Emr, S. D., Hedgpeth, J., Clement, J.-M., Silhavy, T. J., and

6. Ito, K., Bassford, P. J., and Beckwith, J. (1981) Cell 2 4 , 707-717 7. Chang, C. N., Blobel, G., and Model, P. (1978) Proc. Natl. Acad.

8. Zwizinski, C., Date, T., and Wickner, W. (1981) J. Biol. Chem.

9. Izui, K., Nielsen, J. B. K., Caulfield, M. P., and Lampen, J. 0.

10. Ambler, R. P., and Meadway, R. J. (1969) Nature (hd.) 222 ,

11. Nielsen, J . B. K., Caulfield, M. P., and Lampen, J. 0. (1981)

12. Nielsen, J. B. K., and Lampen, J. 0. (1982) J. Biol. Chem. 2 5 7 ,

13. Neugebauer, K., Sprengel, K. R., and Schaller, H. (1981) Nucleic

14. Brammar, W. J., Muir, S., and McMorris, A. (1980) Mol. Gen.

15. Gray, O., and Chang, S. (1981) J. Bacteriol. 145,422-428 16. Pollock, M. R. (1965) Biochem. J. 94,666-675 17. Boyer, H. W., and Roulland-Dussoix, D. (1969) J. Mol. Biol. 41 ,

18. Dubnau, D., Goldthwaite, C., Smith, I., and Marmur, J. (1967) J.

19. Messing, J., Crea, R., and Seeburg, P. H. (1981) Nucleic Acids

Hofnung, M. (1980) Nature ( h n d . ) 85,82-85

Sci. U. S. A. 75, 361-365

256,3593-3597

(1980) Biochemistry 19,1882-1886

24-26

Proc. Natl. Acad. Sci. U. S. A. 78, 3511-3515

4490-4495

Acids Res. 9,2577-2588

Genet. 178,217-224

459-472

Mol. Biol. 27, 163-185

Res. 9.309-321

20. Miller, J. H. (1972) in Experiments in Molecuhr Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

21. Mandell, M., and Higa, A. (1970) J. Mol. Biol. 5 3 , 159-162 22. Clewell, D., and Helinski, D. R. (1972) J. Bacteriol. 110, 1135-

23. Dubnau, D., and Davidoff-Abelson, R. (1971) J. Mol. Bwl. 5 6 ,

24. Gryczan, T. J., Contente, S., and Dubnau, D. (1978) J. Bacteriol.

25. Neumark, R., and Citri N. (1962) Biochim. Biophys. Acta 59,

26. Sherratt, D. J., and Collins, J. F. (1973) J. Gen. Microbwl. 76,

27. Sargent, M. G. (1968) J. Bacteriol. 95, 1493-1494 28. Inouye, H., Pratt, C., Beckwith, J., and Torriani, A. (1977) J.

Mol. Biol. 110 , 75-87 29. Banejee, S., and Fraenkel, D. G. (1972) J. Bacteriol. 110, 155-

160 30. Merril, C. R., Goldman, D., Sedman, S. A., and Ebert, M. H.

(1981) Science (Wash. D. C.) 211 , 1437-1438 31. Heppel, L. A. (1971) in Structure and Function of Biological

Membranes (Rothfield, L. I., ed) pp. 223-247, Academic Press, New York

32. Josefsson, L.-G., and Randall, L. L. (1981) J. Biol. Chem. 256,

33. Crane, L. J., and Lampen, J. 0. (1974) Arch. Biochern. Biophys.

34. Clarke, A. J., Mbzes, P. S. F., and Viswanatha, T. (1980) J. Appl.

35. Machleidt, W., Wachter, E., Scheulen, M., and Otto, J. (1973)

36. Smith, H. 0. (1980) Methods Enzymol. 65, Part 1,371-380 37. Prentki, P., Karch, F., Iida, S., and Meyer, J. (1981) Gene 14,

38. Holmes, D. S., and Quigley, M. (1981) Anal. Biochem. 114,193-

39. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl.

40. Schlesinger, M. J . (1967) J. Biol. Chem. 242 , 1604-1611 41. Hantke, K., and Braun, V. (1973) Eur. J . Biochem. 34,284-296 42. Zubay, G. (1973) Annu. Rev. Genet. 7,267-287 43. Doolittle, R. F. (1977) in Advanced Methods in Protein Sequence

Determination (Needleman, S. B., ed) pp. 28-54, Springer- Verlag, New York

44. Podell, D. N., and Abraham, G. N. (1978) Biochem. Biophys. Res. Commun. 8 1 , 176-185

45. Crane, L. J., Bettinger, G. E., and Lampen, J . 0. (1973) Biochem. Biophys. Res. Commun. 50,220-227

46. Chang, C. N., Nielsen, J . B. K., Izui, K., Blobel, G., and Lampen, J. 0. (1982) J. Biol. Chem. 257,4340-4344

47. McLaughlin, J. R., Chang, S.-Y., and Chang, S. (1982) Nucleic Acids Res. 10,3905-3919

1146

209-221

134,318-329

749-751

217-230

2504-2507

160,655-666

Biochem. 2, 183-189

FEBS Lett. 37,217-220

289-299

197

Acad. Sci. U. S. A. 74,5463-5467

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Secretion of Penicillinase Lacking Lipophilic Modification 11217 PurifIcatmn of Periplasmic Peniclllmase from E. c a t ' penPOl clone Supplemental Material to:

lun6lcillin. kansmycii, t;trs;ycline, shlor&phenicol. diisopropG1 fluorophos- phate, dithiothreitol, p-nltrophenyl phosphate, a-NmP, gluc0~e-6-phosphate. and trypsin were purchased from slgna Chemicals 1st. Louie. NO). Benzyl wnicillin was kindly Drovlded bv Ell Lrllv and Co. flndianamlls. IN1 whlle

polyacrylamide gels were obtalned from Blo-Rad Laboratories. Polybuffer ex- ipectinomycin WaB th; kift Of Th; Upiohn Co. fX~lm1100, MI): !&gents for

changer PEE94, Polybuffer 74 Sepharose 4 0 and Sephadex G-75 were purchased from Phemacia. [9,10-3H1Paimrtic acid 118 Cilmonol~ and 135Slcyateine 11108.5 Crlmmol) were product* of New England Nuclear (Boston, MA). I35SlMeth- lonlne 11350 Cilmoll and lo-32PldATP I l l 0 C i l m l l were obtained from mer-

25972) was Obtained from M.R. Pollock and gram on 25 casein hydrolysatelslte

HBlOl 1171, and 8. 6 u b f d C d ED170 (18) were used as recipients for transforma- (CHIS) medium 116). E. eo& K-12 straln RRl, a reCAI derivative Of straln

tion. E. c o l i JH103 was the reclprent strain for transfection by N13 (191.

a suitable antibiotic for selection. 1t wae t r a n s f o m d according to publieh- E . e o l i RR1 was g m y n at 37'C in L broth I 2 0 1 supplemented with 20 vglml Of

method of Clevell and Hellnski 122). after amplifrcation for 10 h by either ed procedure 121) and plasmids were isolated by the CaCl density gradient

Chloramphenlcd or spectinomyein (200 uglml). 8. d ~ b f i l ~ d transformation was p e r f o m d 08 described by Dubnau (231 and plasmids purified as reported earl- ier (241. Antiblotlc medlum 3 IDzfcOl was supplemented with 20 vglml kana- mycin and 30 vglml amplclllln when eelecting for recanbinant clones. Cultures Of 8 . 6ubttfCd contalnlng genes penP or enPAl were gram in 2% CHIS medium atartmg at pH 6.5 or pH 7.2 for c m p a r i k the penicillinaae prodwte with those m 8. Luhen*dana*a 749lc. It wan found that protease formatux was

8 . L~chrni6onn~d 749Ic. Constitutive for paniclllinase synthesis IATCC

repressed in thls richer medium, as prev~auely observed With 8 . centu6 cu1-

penicillinase mtlvity could be qualitatively assigned for reconbinant clones turea (25). All Bacil lud cultures were grown at 34-C. Relative levels Of

on L agar plates containing polyvinyl alcohol and the necesllazy antibiotics needed for Selection according to the method of Sherratt and Collins 126).

Labelling of cell components

Assaya. antrbody PreParltlOn and Polyacrylamide qel electrophcresie

Penicillinase lctivlty was detemined as deacrlbed by Sargent (27). One unit is the amount Of enzyme hydrolyzing 1 vmol benzyl penLCillin in 1 h at 30.C. Alkaline phosphatase and glucose-6-phosphate dehydrogenaee activities in E. C O C A RR1 were determined using published procedures 128.29). Antibdy preparatron, immunoprecipitation, SDS polyacrylamide gel electrophoreau and autoradiography were performed as previoullly described 19) . A 15-225 gradent polyacrylamide gel containing 8 H urea and 0.1% SDS 171 was vtllxred to re- solve bands representing membrane and exo-large penicillinaee. Silver stain- ing was used for Some gels 130).

Lmalization Of Peniclllmase Activity in E. e o l i

shock procedure Of Heppel 131) or a s modlfled by Randall for use with smll quantitlee Of cells 132). Soluble proteins that are cytoplasmic or loosely bound to membranes were released from the shocked cells fdevoid Of their perr- plasmic contents) by sonication. In a typical experiment involving 2.5 ml cultures, cells were resuspended in 1.0 ml 50 mM TelS-HCl buffer pH 7.5 and sonically disrupted, IHSE 49nicatOr) for 15 h periods (12 microns peak to peak) with coollng for 1 mi" intervals for a total Of 3 tunes. This suspen- SlOn was centrxfuged at 100,000 x g for 40 m m : the supernatant contazned the

pH 7.5 contalnlng 150 mM NaCl and 1% TrItOn X-100. incubated at amblent soluble proteins. The pellet W ~ B suspended xn 300 u l Of 10 mM Tris buffer

temperature for 10 m m vlth occasional mrxing and centrifuged at 100 000 x g for 40 mi"; this supernatant contained the solubilized membrane p r d m s . Total peniclllinase from E. c o i ~ Could be Obtaxned by perfomlng the sonica- t m n and ultracentrlfuge steps on a culture pellet resuspended dlrectly m the above buffer contaming detergent. Trypsin cleavage of expoSed membrane- bound or membrane-associated proteins Of €. c a t * was performed on shocked cells. In a typical experiment, after separating the periplasmnlc contents from cells representing 2.5 ml Of orlgmal Culture, the damaged ce l l s were

added and the mixture was incubated at 3 7 k for 30 mi". Samples of the SYS- suspended in 0.5 ml 50 mM Trls-HC1 buffer pH 7.5. TrypBln 120 vglml) was

penslon were assayed for pen~c~llrnese drrectly. The remamder was centrz- fuged for 10 mln (12,000 X 91 and the Supernatant and pellet (resuspended xn

an identical mannez except for the orn~5slon Of trypsln. In all fractlonatlon 0.5 ml Of the same buffer) Were assayed separately. A Control was treated I"

Steps, dlisopropylfluoro phosphate (1 mMl was present to lnhlbit proteases.

The soluble periplaamic pmtelns of E. e o t c were isolated by the Osmotic

performed ebsentldlly as described earller for 8 . Z-rehen~6anaca ( 9 11 12). The harvestrng of secreted and membrane protelns from E. d u b f d * d was

Total penlclllinase activity was determined by arnsay~ng aliquot. 0; a'sonl-

were isolated by mild lysozyme treatment Vlthout an osmotic support. 1n a cated culture. cytoplasmic protelns. end those loosely bound to the membrane.

typlcal experiment, cells were concentrated 5 fold from an 8 h Culture and suspended In 10 mM Trls-HC1 buffer pH 7.5 contalnlng MgC12 I 1 0 m M 1 EDTA 12 mM) and lysozyme 1200 pg/mll. The mixture was incubated at rook tempera- ture for 30 mrn or untll lysis occurred. then redlmented at 100,000 X g for

plasm or loosely attached to the membrane as well as proceSsed forms Of these. 40 m m . The supernatant contained any penlcrlllnaee released from the cyto-

The remaining penicllllnase could be released from the membrane pellet by treatment Wlth detergent 01 ~n the case of the clone, a160 by sonlca-

plasma membrane. protoplasts were prepared (11) and assayed on rnedla osmotl- tlon. To determine the amount Of penlcllllnase on t e outer Surface Of the

cally supported by sucrose (40% w i v l . Trypsin treatment of protoplasts was carried out a5 described for E . Z4chen~60nm4~ 111) .

The soluble perzplasmrc proteins Of a 4 L late log Culture Of E. C o t & RR1 were isolated by the osmotic shock procedure of Heppel I291 and freere- drled after dialyals. The lypaphzllred materlal (1.3 4, Contllnlng CB. 20 mg peniclllinase) was dissolved in 5.0 ml of start buffer 125 mM plperazlne-HC1 pH 5.70 containing 15% v l v glycerol1 and chromatofocused on 1 column 10.3 X 28 cm) Of polybvffer exchanger PBE94 according to the manufacturer's dlrec- tlons. Proteins with pK1,5.4 were eluted as a large slngle peak at the beginning of the gradlent and did not Contain penlcltllnase. Penlcllllnase activity was associated with 2 peaks eluting from the chramatafacunmg column (Fig. 6) at pH 5.04, 1 Value Consistent wlth the 1IOeleCtrlC POlnt Of 8. Lehencdonm~b secreted penicillinase 1331, and at pH 4.82-4.76. Thls hetero- geneity can probably be explained by a p ~ e v ~ o u s ObserVat10n that some amlde groups of Gln and A m resxdues in exapenic~ll~naae can be hydrolyzed to the corresponding acids and thus grve specre5 wlth lowered lsoelectrlc p01nt5 (91. Fractions Wlth the highest penlolllinase actlvzty from the pH 5.04 peak were pooled. concentrated and applled to a sephadex G-75 column Y S m g NLCl 10.9% w l v ) as eluent to remve plybuffer and glycerol. The total yleld from the peni=illin=.se-cDnt=i"i"g fractmns was 7.4 mg 1 3 5 % ) . €mail purification was achieved using a 5.0 ml column Of N-acetyl-penlclllBmlne SepharOSe 4 B 1341. A sample from the chromtofocusmg step contalnlng 1 mg of penlcllllnase was applied in 0.5 m1 buffer 125 Iw cltrate pH 4.51 to the afflnlty adsorbent equilibrated I" the same buffer. A11 the penlclllinase was bound to the column and was eluted Wlth 25 Iw cltrate pH 4.3 buffer. The Y I ~ ~ O Y S Stages of purification are presented in Fig. 7.

Protein Sequence Determlnatlon

Amino acid sequence analysis was carried out m collaboratlon Wlth M. Horn of Sequemat, Inc. fwatertown, ml. Pzotezn 1300 ug of purlfled p e n -

derivative Of 3-aminopropyl glass and svblected to Edmdln degradation 1 3 5 ) . cilllnase) was immobilized on the aolld-phase p-phenylene dllsothlocyanate

Phenylthiohydantoin derivatives were d e t e m m e d by hlgh pressure l q u d chromatography with monitoring at 269 and 313 m, and compared to a Chromato- gram Of a Standard mixture.

Construction Of plasmid vectors contalnlng penP and penP01

All DNA fragments were recovered from 5% polyacrylamide gels etamed wlth ethldlum bromide according to the gel elutlon method of Smlth (361. The 8-lilctmse gene in pER325 (37) was inactivated by Bal 31 digestion of Pst I treated Vector and the DNA was subsequently recncularined by blunt-end liga-

mphenlcol and tetracycline resistant clone With approximately 1.4 kbl removed tion and "Bed to transform E. cati RR1. An ampxlllin-sensitive and chlor-

from the plasmld, was the 80urce Of pRW33 (€19. 1). The 8 . l ~ c h e n ~ 6 o n m ~ a 7491'2 penicllllnase gene originating from i en 114) was recovered as a 1.4 kb ECOR I* DNA fragment. The sticky ends were h e d m, Hind Ill and BamH I phosphorylated linkers added, and the resulting penP fragment was cloned between the Hind 111 and Ban3 I sites of pRw33. to glve the 6.0 kb pRw83

RR1. (Fig. 1). Thls vector containing the penP gene was Used to transform E. coic

Figure 1:

treated =multaneouSly wlth Hlnd I11 120 Units) and BamH I (20 Unltsl for

MgC17 17 Iwl and bovine serum slbmln 1100 V 4 l m l l . The 1.4 kb Hlnd 1 1 1 - B a m H 1 11 h at 37'C I" 7 mM Trls-HC1 buffer, pH 7.4, contalnmg NaCl 160 mM),

The construction of pRW84 1s outlined I" Elg. 2. pRW83 ( 5 0 ug) was

fragment containing the enP gene was lsolatid, and 5.0 ug was digested fur- ther Wlth FDYD I1 (10 u n h for 2 h. Of the 3 fragments obtalned (€19. 21. the 334 bp Hlnd 111-FnuD I1 and 586 bp FnuDII-EnuD I1 DNA fragments were isolated. The 586 bp €DUD 11-FnuD I1 fragment 12.0 ~ 4 1 was then treated wlth Hha I (10 units) for-1 h. The nolvtlon wis heated to-65°C for 10 m m and the

was blunt-ended by digestLon vlth 51 nuclease I 5 0 unltsl at 4'c overnight. DNA was ethanol precipitated. The re9ultlng 5 1 1 bp Hha I-FnUD I1 DNA 11.9 vg1

The DNA was precipitated wlth ethanol and flnally restricted with B g l I1 110 units) for 1 h. The DNA fragment used In the constructlon Of pRW84 had been

ended Bgl I1 erte ae the other end. Thls plece IC), plus the 334 bp Hlnd 111- therefore blunt-ended at the Hha I site and was 227 bp in length to the stlcky-

FnuD I1 fragment Isolated earlier (8) and the large fragment 1-5.4 kb) of a separate Hind 1 1 1 - B 9 1 I1 digest Of pRw83 ( A I were mlxed I= equ~molar amounts ln the presence of TI DNA ligase (200 units1 (Fig. 21. After lncubatlon at 10'C for 14 h, the SOlutlOn was used dlrectly to transform E. C O P < RR1.

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

11218 Secretion of Penicillinase Lacking Lipophilic Modification

Figurn

Figure 3: -striction m p of pRWlO1, a bifYnetiona1 plasmid. The P.nP gens can b. expr*ssed in both E. e o t i and (1. b u b t i t i a .

culture1 by the rapid boiling method of Holmes and Ouigley 1381. Rapid E. e o t i pla-ids w r a prepared for reatsiction e n r p analysis 15.0 ml

screening of 8 . A u b t i t i a plasmids was carried out for 1.5 nl culture. g r m overnight. The cell. were pallsted in the microfuge and sumpended In 100 yl of 10 rAl Trig-HCl, pH 7.5 buffer containing EDTA I1 ml and sucrose I251 wlv). Pour microlitars of lysozyna solution I10 nglml) were added and the mixture

EDTA I10 ull pH 7.5 and 2( SDS in 0.7 H NaCl I 100 111 vera added and'nixed incubated at 37.C with gentle agitation. After 1 h 5 H NaC1 120 ul) 0.2 W

gently. The suspension wan *tored on ice for 30 min. and then microfuasd for

Solution Stored at -20.C for 20 min. The pracipitated DNA vas recovered by 10 mi.. To the iupamatant, an equal volume of is6piopsnoi vas iddid-iid thi centrifugation and vashed with cold 70( ethanol. The pellet van air-dried and dissolved in water. The yield of plasmid DNA vas from 1-2 yg.

Figure 7:

- = ... .c ""

1 2 3 4

mrification of panicillinam f r m pariplam of E. c o t i RR1 con- t i i n i n g PRW84.

Saaples 13-50 ug protein1 at various st.-. of purification yere .ubjsct.d to SDS polyacrylamide gel electrophoresis. Lane 1: II1Unoprecipitate of total pariplasmic proteins shown in lane 2 . Lane 3 : After chrmtofocusing of proteins s h m in lane 2 and ChromatOgraphy on Ssphadex C-75 to remove Polybuffer and glycerol. l a n e 4: Purified wnicillinase after adsomtion and elution from N-acetyl-penicillbninc-sepksro.. 18 sol-: The gel was-stained with Cwmassie blue.

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from

P S Mézes, W Wang, E C Yeh and J O Lampensubtilis.

site for lipoprotein modification. Expression in Escherichia coli and Bacillus Construction of penP delta 1, Bacillus licheniformis 749/C beta-lactamase lacking

1983, 258:11211-11218.J. Biol. Chem. 

  http://www.jbc.org/content/258/18/11211Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/18/11211.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 13, 2018

http://ww

w.jbc.org/

Dow

nloaded from