Cloning and Characterization of the Bacillus licheniformis Gene Coding for Alkaline Phosphatase

JOURNAL OF BACTERIOLOGY, June 1984, p. 978-982 Vol. 158, No. 3 0021-9193/84/060978-05$02.00/0 Copyright C) 1984, American Society for Microbiology Cloning and Characterization of the Bacillus licheniformis Gene Coding for Alkaline Phosphatase F. MARION HULETT Department of Biological Sciences, University of Illinois at Chicago, Chicago, Illinois 60680 Received 19 December 1983/Accepted 8 March 1984 The structural gene for alkaline phosphatase (orthophosphoric monoester phosphohydrolase; EC 3.1.3.1) of Bacillus licheniformis MC14 was cloned into the PstI site of pmk2004 from chromosomal DNA. The gene was cloned on an 8.5-kilobase DNA fragment. A restriction map was developed, and the gene was d on a 4.2-kilobase DNA fragment. The minimum coding region of the gene was localized to a 1.3- kilobase region. Western blot analysis was used to show that the gene coded for a 60,000-molecular-weight protein which cross-reacts with anti-alkaline phosphatase prepared against the salt-extractable membrane alkaline phosphatase of B. licheniformis MC14. A number of biochemical, physiological, and immunological studies concerning the synthesis and localization of alkaline phosphatase (APase; orthophosphoric mnonoester phosphohydrolase; EC 3.1.3.1) in Bacillus subtilis and Bacillus licheniformis have been reported. APase of B. licheniformis has been identified in a secreted form (F. M. Hulett, K. Stuckmann, D. Spencer, T. Sanopoulou, and F. Abedinpour, submitted for publication), a cell-bound form released by lysozyme (8, 9), a salt-extractable membrane form (10, 17), a detergent-extractable membrane form (19), and an enzymatically inactive form in the cytosol (unpublished data). These various APase species are immunologically related. It has been shown that culturing conditions significantly affect both the distribution and the amount of synthesis of APase (8, 15, 19). Although there are less extensive localization studies of APase in B. subtilis, seemingly contradictory reports of cell-bound (22, 24) or membrane-associated (inner leaflet) (7) and secreted APase (4, 27) in the organism have also been made. It is difficult to interpret the localization data (active dimer associated with inner leaflet of cytoplasmic membrane [6, 21] versus active dimer secreted [Hulett et al., submitted for publication]) based on any of the current models for protein secretion (3, 7, 23-25) or insertion of proteins into membranes (13, 18). There are at least two possibilities which would explain these data. Either there is a single structural gene which is under uniquely complex transcriptional and translational regulatory control or there are mnultiple structural genes for APase which account for the synthesis of different species directed to different locatioris depending on the growth conditions. (Thd lattor possibility is supported by the fact that there have been no APase-negative mutants isolated in Bacillus strains which are due to mutation in the structural gene [phoa].) We have cloned the phoa gene of B. licheniformis (MC14) on multicopy plasmid pmk2004 to facilitate regulatory and structural studies on APase. This paper reports the steps involved in cloning and subcloning the phoa gene, a restriction map defining a minimum coding region, and evidence of relatedness of the cloned gene product to the APase isolated and characterized previously. MATERIALS AND METHODS Bacterial strains and plasmids. The bacterial strains and plasmids used are given in Table 1. Escherichia coli Xph9Oa, 978 which contains E15 within phoa (phoa8), was obtained from J. Beckwith via P. Berg (1, 11). DNA isolation. Small-scale plasmid DNA isolation was done by a modification of the method of Birnboim and Doly (2). Cesium chloride equilibrium centrifugation DNA isolation was used for large-scale purification of plasmid DNA. Media. Antibiotic screening was performed on Luria agar plates, using penicillin G at 150 jig/ml, kanamycin (Kan) at 50 jig/ml, or tetracycline at 30,ug/ml. Screening and selection were carried out on 1% Neopeptone plates (Difco Laboratories) containing 1.5% Noble agar, 0.1 M Trisma Base (ph 7.5), 1% NaCl, 50 pg of kanamycin per ml, and 50,ug of 5-bromo-4-chloro-3-indolylphosphate-p-toluidine (XP) (Neo-XP-Kan plates). The indicator dye (XP) was obtained from Sigma Chemical Co. When selection on penicillin G was required, screening was carried out as a second step on the above Neopeptone plates with an antibiotic. It has been reported (1) and we confirmed that XP significantly decreases the effect of penicillin concentration. Enzymes. Restriction enzymes were obtained from Bethesda Research Laboratories or Amersham Corp. Digestions were carried out in high, medium, or low buffer as described previously (14). Digestions were at 37 C for 1 h. Ligation was performed with T4 ligase at 40C overnight in 66 mm Tris-hydrochloride (ph 7.9)-33 mm NaCl-10 mm MgC12-1 mm,13-mercaptoethanol-25 mm ATP. Transformation. Transformation was carried out by the method of Cohen et al. (5). Gel electrophoresis. DNA fragments generated by restriction enzyme digests were analyzed in 0.8% agarose gels in 40 mm Tris-20 mm sodium acetate. Smaller fragments were analyzed on acrylamide gels in 0.089 M Tris-borate buffer containing 2 mm EDTA. Insertional Tn5 mutagenesis. E. coli Xph9Oa carrying plasmid pmh81 was infected with bacteriophage lambda carrying transposon Tn5. the multiplicity of infection was one. Adsorption was carried out on ice for 30 min before the cells were grown under plasmid selection (penicillin) at 30 C for 3 h. Cells were then plated on Luria plates containing 400 jig of kanarmycin per ml. (TnS carries a kanamycin resistance gene.) We had previously determined that E. coli with TnS inserted into the chromosome were killed by concentrations of kanamycin of 200,ug/ml but that E. coli cells carrying TnS on a multicopy plasmid could grow in the presence of 600 to 800 jig of kanamycin per ml. Therefore, the Luria plates

VOL. 158, 1984 B. LICHENIFORMIS GENE CODING FOR ALKALINE PHOSPHATASE 979 TABLE 1. Bacterial strains and plasmids Bacterial strain Genotype or Origin or plasmid phenotype Strain E. coli Xph9Oa F- lacz624 phoae15 J. Beckwith (11) proc+ phor+ trp rpsl B. licheniformis phoa+ F. M. Hulett (10) MC14 Plasmid pmk2004 Ampr Tetr Kanr Whitea M. Khan (12) pmh8 Amps Tetr Kanr Bluea This study pmh8a1 Amps Tetr Kanr Bluea This study (EcoRI2-EcoRII) pmh8a2 Amps Tetr Kanr Whitea This study (PvuII2-PvuII3) pmh81 Ampr Tet' Kan' Bluea This study (Xho2-EcoRI2) pmh82 Ampr Tetr Kan' Whitea This study (PvuII2-PvuII3) pmh81a1 Ampr Tet' Kan' Whitea This study (HindIII3-HindIII4) a Indicates the ability to cleave XP on indicator plates. containing 400,ug of kanamycin select for cells containing TnS inserts on pmh81. The colonies on Luria plates containing 400 jig of kanamycin were washed off, and the plasmids were isolated from the combined colonies. The resulting plasmid collection was used to transform strain Xph9Oa. Screening and selection were carried out on Neo-XP-Kan plates as described above. White colonies represent colonies carrying plasmid pmh81 with Tn5 inserted into the APase gene, and blue colonies represent colonies carrying plasmid pmh81 with inserts external to the gene. ExolI-Sl mapping. Deletion mapping was carried out by using a slight modification of the procedure described by Roberts and Lauer (16). pmh81 (10 jig) was digested with XhoI, ethanol precipitated, and suspended in ExoIll buffer (16) containing 50 mm NaCl. Thirty units of Exolll nuclease were added and incubated at 22 C for 90 min. Samples were taken between 30 and 90 min, and the reaction was stopped by adding an equal volume of 2X SI buffer (16). Five units of SI was added and incubated for 30 min at 20 C. The DNA was then phenol-chloroform extracted, ethanol precipitated, ligated, and used to transform strain Xph9Oa. Screening for APase activity was carried out on Neo-XP plates. Preparation of DTP paper and Western blotting of protein from gel to DTP paper. The procedures have been outlined previously (26). The only modification was that iodinated goat anti-rabbit immunoglobulin G was substituted for iodinated protein A for labeling the primary antibody. The primary antibody was made to the salt-extractable, membrane-associated APase (10, 17). RESULTS Cloning and restriction mapping of the phoa gene. To clone the APase gene of B. licheniformis MC14, we constructed a PstI chromosomal DNA fragment library containing 6,000 independently isolated clones. B. licheniformis MC14 chromosomal DNA was cut with PstI and mixed with the vector pmk2004 (kindly provided by P. Matsumura) which had been cut with PstI and pretreated with calf intestinal APase. These DNA fragments were ligated and used to transform E. coli Xph9Oa (Table 1). Selection and screening were carried EcoRl, Sal 12 Hind 1113 Xho 12 FIG. 1. Restriction map of pmh8 (13.65 kb). Vector DNA is represented by a heavy solid line. Insert DNA is represented by double lines. Fragments are referred to by the flanking restriction site numbers, e.g., XhoI2-PvuII4 identifies the region which includes the phoa gene. out by using 1% Neopeptone plates containing kanamycin and XP. In addition, colonies were screened for ampicillin and tetracycline resistance. Since only one PstI recognition site exists in pmk2004 and this site is in the gene coding for ampicillin resistance, loss of ampicillin resistance was an indication of insertional inactivation. In these experiments it was found that between 80 and 85% of colonies which were kanamycin and tetracycline resistant were ampicillin sensitive. Putative APase colonies turned blue. Approximately 1 in 3,000 colonies was blue. Plasmid DNA from the blue colonies was isolated by a modified method of Birnboim and Doly (2) and cut with PstI. The sizes of the two DNA fragments observed corresponded to linear pmk2004 (5.2 kilobases [kb]) and to an insertion fragment of 8.45 kb. When the plasmid DNA was used to transform E. coli Xph9Oa, all transformed cells (kanamycin and tetracycline resistant, ampicillin sensitive) were blue on XP indicator plates. This plasmid, pmh8, is shown in Fig. 1 with restriction sites indicated. The following plasmids and s were generated to determine the locus of the phosphatase gene (Table 2). An EcoRIj-EcoRI2 (pmh8a1) of pmh8 resulted in a plasmid 2.75 kb smaller TABLE 2. Deletion and s of pmh8 and pmh81 Transformed Plasmid Construction Fragment Cloned into: Xph9oa phenotype (colony) pmh8 EcoRI 10.95-0 Blue pmh8 PvuII 5.9-8.3 White pmh8 PvuII 5.9-8.3 SmaI site of White pmk2004 pmh8 XhoI-RI 6.4-10.95 RI-XhoI site of Blue pmk2004 pmh81 HindlIl 4.1-7.2 HindIll site of White pbr322 pmh81 BglII 6.0-7.0 Blue 1112

980 HULETT Pvu2I.tr m 7.25 ( ) 2g e Sal12 4 Sma 12 = Hind 1113 n 3 mal 0 12 3.64 FIG. 2. Restriction map of pmh81 (7.25 kb). Fragments are referred to by the flanking restriction site numbers as in Fig. 1. which retained the blue phenotype on XP indicator plates. A PvuII2-PvuII3 (pmh8a2) of pmh8 resulted in a plasmid 2.5 kb smaller which did not exhibit the phosphatase phenotype (i.e., colonies are white on XP indicator plates). This indicated that the phosphatase gene was either wholly in the PvuII DNA or partially deleted. A of the PvuII2-PvuII3 fragment into the SmaI site on pmk2004 (pmh82) resulted in a plasmid 2.5 kb larger than pmk2004 which did not confer the blue colony phenotype on strain Xph9Oa upon transformation. Thus, pmh8a2 (PvuII2-PvuII3) represents a partial of the phosphatase gene. The EcoRI2-XhoI2 fragment of pmh8 was d into the EcoRI-XhoI region of pmk2004 (resultant plasmid was pmh81, see Table 1). Xph9Oa transformed with pmh81 exhibited the blue-colored colony phenotype on indicator plates. Double digests (XhoI and EcoRI) of this plasmid showed a 4.2-kb insert and a 3.05-kb vector. This indicates that the APase gene is located between the XhoI2 and EcoRI2 sites of pmh8 and contains the PvuII3 site. Figure 2 shows the restriction map of pmh81. The PvuII site (PvuII3) implicated in the coding region maps at 5.0 kb. Deletions and s used to map the coding region are given in Table. 2. The HindIll (HindIII3-HindIII4) did not contain the gene; the BglII (BglII2-BgIII3) did not interrupt the gene. Therefore, the HindIll site at 4.1 kb must be in the coding region. TnS insertional mutagenesis of pmh81 was used to further map the gene. The procedures used are described above. pmh81 plasmids carrying TnS were used to transform strain Xph9Oa. The resulting transformed cells were selected and screened on Neo-XP-Kan plates. The location of the TnS insertion was mapped in all of the plasmids which did not cause APase production in transformed Xph9Oa, indicating that the insertion was in the coding region of the gene. Of the 15 separate clones, all mapped between XhoI2 and PvuII3. The one which mapped furthest from the PvuII3 site was at 3.64 kb, as indicated on the map of pmh81 (Fig. 2). Deletions were constructed by cutting pmh81 with XhoI, followed by nuclease(s) ExollI-SI digestion and ligation (Fig. 3). ExoIII-SI nuclease mapping was performed as outlined above. Deletion plasmids were used to transform strain Xph9Oa. Plasmids from blue and white colonies were isolated, and the extent of the in each was mapped. Colonies containing plasmids with s from XhoI2 to 3.6 (on pmh81) remained blue. Plasmids with larger s (from XhoI2), past the point (3.64 kb) at which TnS insertion caused inactivation of the APase gene on pmh81, showed no APase production when used to transform Xph9Oa. This locates one terminus of the gene between 3.60 and 3.64 kb. The minimum size of the coding region is calculated to be 1.3 kb with the right terminus of the gene in pmh81 close to 3.64 kb and the left terminus of the gene containing the PvuII3 site. Preliminary transcription mapping studies indicate that transcription starts at least 150 bases before the PvuII3 at 5 kb. Expression of B. licheniformis gene in E. coli. Although the blue colony color of cells (Xph9Oa) containing pmh8 or pmh81 on XP plates is easily detected after 2 days, no APase production in growing cultures could be measured. Western blot analysis was used to determine (i) the relatedness to the APase species previously studied and (ii) the size of the cloned gene product. Figure 4 shows an autoradiogram of a Western blot which had been treated with rabbit anti-apase, followed by 125I-labeled goat anti-rabbit immunoglobulin G. Lanes containing purified APase (lane 1) or cell lysates of Xph9Oa carrying pmh81 (lane 3) show a 60,000-molecular-weight band (Fig. 4). (The subunit size of B. licheniformis MC14 APase is 60,000 [10, 17].) Lane 2 (Fig. 4) which contains a lysate of Xph9Oa carrying pmk2004 does not show this band. DISCUSSION We have cloned the structural gene for APase into an E. coli plasmid, pmk2004. The original DNA fragment (8.45 Pvu Hind 1114 Sma I Bgl 113 Eco RI2.3 Tn5 3." J. BACTERIOL. FIG. 3. ExoIII-SI of pmh81. Deletion plasmids were digested with Hindlll, and the length of the HindIII2-HindIII3 fragment was determined on 1% agarose gels. (Mapping was based on the assumption that the rate of digestion by ExoIl from XhoI2 was equal in both directions.) Solid lines indicate the distance from XhoI, in pmh81, that retained the APase phenotype when the constructed plasmid was transformed into strain Xph9Oa. Dashed lines indicate the length of DNA deleted which resulted in an APase-negative phenotype. 1112

VOL. 158, 1984 B. LICHENIFORMIS GENE CODING FOR ALKALINE PHOSPHATASE 981 I 2 I...7.'l: '::.:; '.. F.:r: r. N z FIG. 4. Cross-reactivity of the APase of B. licheniformis and of pmh81-encoded protein with anti-apase. Purified APase and cell lysates of strain Xph9Oa carrying pmk2004 or pmh81 were resolved on sodium dodecyl sulfate-polyacrylamide gel, electrophoretically transferred to DPT paper, and reacted with anti-apase antiserm and 1251Ilabeled goat anti-rabbit antibody. The blot was subjected to autoradiography for 48 h. Lane 1, purified APase; lane 2, pmk2004 lane 3 pmh81. kb) was cloned from a Pstl digestion of B. licheniformis DNA into a kanamycin- and tetracycline-resistant plasmid, pmh8. A 4.2-kb fragment, Xho12-EcoR12, of pmh8 containing the gene was d onto a penicillin-resistant plasmid pmh81. Restriction mapping of the cloned region indicates that the PVU113 site at 5.0 kb on pmh81 is in the coding region of the gene and that the BglII12 site at 6.0 kb is not. The other end of the coding region was mapped at ca. 3.64 kb by Tn5 insertional mutagenesis and ExoIll-SI mapping. Therefore, the minimum coding region of the gene is 1.3 kb, and the maximum coding region is 2.4 kb. The subunit molecular weight and amino acid analysis (ca. 550 amino acids per subunit) of APase implicate a 1.65-kb coding region. Thus, the APase structure gene fits well within the 2.4-kb fragment. Expression of the cloned gene in E. coli was quite low. Possible reasons for low expression could be: (i) the absence of specific positive regulatory proteins in the heterologous system necessary for full expression of the Bacillus gene or (ii) low E. ccli RNA polymerase efficiency in transcribing the Bacillus gene. Western blot analysis of whole cell lysates of E. coli Xph90a carrying the, pmh81, showed that the cloned DNA coded for a protein of 60,000 daltons (subunit size of B. licheniformis MC14 APase) that cross-reacted with anti-apase. Thus we have shown by three criteria that the cloned B. licheniformis DNA codes for APase, i.e., molecular weight, antigenicity, and enzymatic hydrolysis of XP. Further characterization of this gene should reveal the reasons for APase compartmentalization in the cell (9, 20, 21), its secretion as an extracellular enzyme (Hulett et al, submitted for publication), and the mechanisms involved in the regulation of expression of the APase gene (19). For 3 these and other purposes, in vitro transcription, Si nuclease mapping, and sequencing studies are being initiated. ACKNOWLEDGMENTS I thank Philip Matsumura for his generous contributions of bacterial stains, plasmids, enzymes, and equipment, as well as technical guidance and innumerable helpful discussions. I am grateful to Jung-Wan K. Lee for restriction mapping of pmh81. I thank Diana Rivera for her excellent technical assistance. LITERATURE CITED 1. Berg, P. E. 1981. Cloning and characterization of the Escherichia coli gene coding for alkaline phosphatase. J. Bacteriol. 146:660-667. 2. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. 3. Blobel, G. 1980. Intracellular protein topogenesis. Proc. Natl. Acad. Sci. U.S.A. 77:1496-1500. 4. Cashel, M., and E. Freese. 1964. Excretion of alkaline phosphatase by Bacillus subtilis. Biochem. Biophys. Res. Commun. 16:541-544. 5. Cohen, S. N., A. C. Y. Chang, and L. Hsu. 1972. Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R factor DNA. Proc. Natl. Acad. Sci. U.S.A. 69:2110-2114. 6. Emr, S. D., M. N. Hall, and T. J. Silhavy. 1980. A mechanism of protein localization: the signal hypothesis and bacteria. J. Cell Biol. 86:701-711. 7. Ghosh, B. K., J. T. M. Wouters, and J. 0. Lampen. 1971. Distribution of the sites of alkaline phosphatase(s) activity in vegetative cells of Bacillus subtilis. J. Bacteriol. 108:928-937. 8. Glynn, J. A., S. D. Schaffel, J. M. McNicholas, and F. M. Hulett. 1977. Biochemical localization of the alkaline phosphatase of Bacillus licheniformis as a function of culture age. J. Bacteriol. 129:1010-1019. 9. Hansa, J. G., M. Laporta, M. A. Kuna, R. Reimschuessel, and F. M. Hulett. 1981. A soluble alkaline phosphatase from Bacillus licheniformis MC14: histochemical localization, purification and characterization and comparison with the membrane-associated alkaline phosphatase. Biochim. Biophys. Acta 675:340-401. 10. Hulett, F. M., S. D. Schaffel, and L. L. Campbell. 1976. Subunits of the alkaline phosphatase of Bacillus licheniformis: chemical, physiochemical, and dissociation studies. J. Bacteriol. 128:651-657. 11. Inouye, H., S. Michaelis, A. Wright, and J. Beckwith. 1981. Cloning and distribution mapping of the alkaline phosphatase structural gene (phoa) of Escherichia coli and generation of mutants in vitro. J. Bacteriol. 146:668-675. 12. Kahn, M., R. Kolter, C. Thomas, D. Figurski, R. Meyer, E. Remaut, and D. Helinski. 1979. Plasmid cloning vehicles derived from plasmids CalEl, F, R6K, and RK2. Methods Enzymol. 68:268-280. 13. Lodish, H. F., and J. E. Rothman. 1979. The assembly of cell membranes: the two sides of a biological membrane differ in structure and function; studies of animal viruses and bacteria have helped to reveal how this asymmetry is preserved as the membrane grows. Sci. Am. 240:48-63. 14. Maniatis, T., E. F. Fritsch, and S. Sambrook. 1982. Molecular cloning. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 15. McNicholas, J. M., and F. M. Hulett. 1977. Electron microscope histochemical localization of alkaline phosphatase(s) in Bacillus licheniformis. J. Bacteriol. 129:501-515. 16. Roberts, T. M., and G. D. Lauer. 1979. Maximizing gene expression on a plasmid using recombination in vitro. Methods Enzymol. 68:473-482. 17. Schaffel, S., and F. M. Hulett. 1978. Alkaline phosphatase from Bacillus licheniformis solubility dependent on magnesium, purification and characterization. Biochim. Biophys. Acta 526:457-467.

982 HULETT 18. Silhavy, J. T., S. A. Benson, and S. D. Erm. 1983. Mechanisms of protein localization. Microbiol. Rev. 47:313-344. 19. Spencer, D. B., C.-P. Chen, and F. M. Hulett. 1981. Effect of cobalt on synthesis and activation of Bacillus licheniformis alkaline phosphatase. J. Bacteriol. 145:926-933. 20. Spencer, D. B., J. G. Hansa, K. V. Stuckmann, and F. M. Hulett. 1981. Membrane-associated alkaline phosphatase from Bacillus licheniformis that requires detergent for solubilization: lactoperoxidase 1251 localization and molecular weight determination. J. Bacteriol. 150:826-834. 21. Spencer, D. B., and F. M. Hulett. 1981. Lactoperoxides-'251 localization of salt-extractable alkaline phosphatase on the cytoplasmic membrane of Bacillus licheniformis. J. Bacteriol. 145:934-945. 22. Takeda, K., and A. Tsugita. 1967. Phosphoesterases of B. J. BACTERIOL. subtilis. J. Biochem. 61:231-241. 23. von Heijne, G. 1980. Trans-membrane translocation of proteins: a detailed physics-chemical analysis. Eur. J. Biochem. 103:431-438. 24. Wickner, W. 1979. Assembly of proteins into biological membranes: membrane trigger hypothesis. Annu. Rev. Biochem. 48:23-45. 25. Wickner, W. 1980. Assembly of proteins into membranes. Science 210:861-868. 26. Wong, S. L., and R. H. Doi. 1982. Peptide mapping of Bacillus subtilis RNA polymerase factors and core-associated polypeptides. J. Biol. Chem. 257:11932-11936. 27. Wood, D., and H. Tristam. 1970. Localization in the cell and extraction of alkaline phosphatase from Bacillus subtilis. J. Bacteriol. 104:1045-1051.