Vol. 44, No. 2, February 1998 BIOCHEMISTRY anti MOLECULAR BIOLOGY INTERNATIONAL Pages 391-397

STUDIES ON THE HETEROLOGOUS EXPRESSION OF BstVl RESTRICTION ENDONUCLEASE IN Escherichia coli.

Claudia Saavedra ~, Enrique Gonzalez b and Claudio V~isquez ~ *

�9 aboratorio de Microbiologia Molecular, Facultad de Quimica y Biologia, Universidad de Santiago de Chile, Santiago, Chile and b Instituto de Biologia Vegetal, Universidad de Talca, Talca, Chile.

Received November 20, 1997

SUMMARY

Bacterial restriction and modification systems must be regulated to avoid self-restriction. It is generally accepted that cognate DNA methyltransferases normally protects both, the host's chromosome and extrachromosomal elements from the activity of their endonuclease counterparts. When the bstVIRM genes from Bacillus stearothermophilusV were subcloned into Escherichia coli, several clones exhibiting a r+m" phenotype were originated. The present work was undertaken to analyze the possibility that mechanisms other than DNA methylation could account for the viability of these cells. No evidence was found for an inhibitory agent or endonuclease compartmentation. In vivo experiments showed that ~, phage multiplication was poorly restricted by the heterologous enzyme. The restricting activity against the incoming phage increased however when phage adsortion was performed at higher temperatures. Analogous experiments in which a DNA-repair deficient strain was used as a host for the thermophilic R-M system suggested, to some extent, the participation of the repair machinery in the viability of r§ clones.

key words: B. stearothermophilusV/restriction-modification system/thermostable/isoschizomer

INTRODUCTION

Two highly specific enzymatic activities, an endonuclease and a DNA methyltransferase form part

of the so called type II restriction and modification (R-M) systems. While the methylase catalyzes the

transfer of methyl groups from S-adenosyl-L-methionine to defined nucleotides in the target

sequence, the endonuclease degrades any DNA lacking the modification pattern which enters the cell

(1,2).

For many years, we have been interested in studying the BstVI R-M system, specified by the

thermotolerant bacterium Bacillus stearothermophilus V, which is isoschizomeric of the XhoI (3)

and PaeR7 (4) R-M systems. Both, the endonuclease and the methylase have been characterized (5,

6), and their genes were cloned into E. coli (7) and sequenced (8). Further studies demonstrated that

*Send correspondence to Dr. Claudio V/tsquez at the address above, FAX (56) 2-6812108,

E-mail: cvasquez@lauca.usach.cl

1039-9712198/020391-07505.00/0 Copyright �9 1998 by Academic Press Australia.

391 All rights of reproduction in any form reserved.

Vol. 44, No. 2, 1998 BIOCHEMISTRYand MOLECULAR BIOLOGY INTERNATIONAL

the expression of the bstVIM gene was restricted to the vegetative life cycle of the bacterium (9).

When the bstVIRM genes were being characterized, several clones exhibiting a r+m" (restriction

proficient-modification deficient) phenotype were obtained (7). Since in restriction endonuclease-

producing strains the cognate methyltransferase represents the natural protection mechanism of the

host's DNA against autorestriction, the viability of those clones in which the MTase is absent thus

must rely in other properties of the cell and/or the system itself. In this work, we studied the

expression of the BstVI endonuclease in an attempt to explain the survivorship of these restriction

proficient-modification deficient E. coli clones.

MATERIALS AND METHODS

E. coil JMI09 {em/Al, recAl, gyrA96, thi, hsdR17 (rk , m k+), relA1, supE44, ~-, A(lac-proAB), [F', traD36, proA+B +, lacIqZAM15]}, and E. coil ES1301mutS [lacZ53, routS201 ::Tn5, thyA36, rha- 5 metB1, deoC, IN(rrND-rrNE)], were from Promega (USA). E. coil LE 392 (supE44, supF58, hsdRS14, galkz, galT22, metB1, trpR55, lacY1) was provided by Dr. Luis Meza-Basso from Universidad de Talon.

The cells were grown in Luria broth (10) at 37~ with aeration. When required, Ap was added to a final concentration of 100 ~zg/ml. Agar plates were made by adding 2% agar (DIFCO) to the above medium.

Preparation of cell-free extracts and endonuclease or methylase assays have been described previously (5-7). Protein concentrations were determined by the method of Lowry et al. (11). ci857 bacteriophage was a generous girl of Dr. Xavier Jordana, Universidad Catolica de Chile. All chemicals were molecular biology grade reagents.

RESULTS AND DISCUSSION

Characterization of recombinant plasmids from E. coil r*m clones. The molecular analysis of

one of the bstVIR gene-containing pBluescript derivatives, which was purified from a recombinant

r+m E. coli clone is shown in Fig. 1. Hereafter we will refer to this molecule as the PPt plasmid

Following digestion with the isoschizomeric endonucleases X'hoI and/or BstVI, the DNA molecule

was linearized in both cases (Fig. 1 c and d, respectively), thus indicating that the corresponding target

sequence in the vector's multiple cloning site had not been methylated in vivo. Since the cognate DNA

methyltransferase normally protects the host's chror~osome and extrachromosomal DNAs from

autorestriction (1, 12), the possibility that alternative mechanisms to DNA methylation could operate

to explain the viability of these clones was investigated.

Studies on the in vivo inhibition of BstVI endonoclease. Previous observations made during the

purification ofBstVI endonuclease from its natural host, suggested the presence of an inhibitor of

the endonuclease in vivo. To test this possibility, crude or fractionated cellular extracts prepared from

392

Vol. 44, No. 2, 1998 BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL

Figure 1.

Restriction analysis of the recombinant plasmid (PP0 carrying the bstVIR gene from B. stearothermophilus V. Agarose (1%) gel electrophoresis of the digestion products of pP~ with c) XhoI, d) BstVI, e) EcoRI + PstI, f) e + HindIII. Wells a) and b) contained molecular weight standards (NEB's 1 kb ladder) and undigested pP1, respectively. Gels were run horizontally at room temperature until the dye (bromophenol blue) reached the bottom, stained in a 1 #g/ml ethidium bromide solution and photographed under short-wave UV light using Polaroid 667 films.

pPl-containing E. coli clones were assayed for an inhibitory effect of the activity of BstVI

endonuclease in vitro. Aliquots of these extracts were preincubated with the BstVI endonuclease at

37 or 60~ prior to the standard restriction assay. No inhibitory effect was detected (not shown). In

this context, it is interesting to note that the presence of specific restriction endonuclease inhibitors,

has been documented mainly for phage-specified proteins inhibiting type I EcoB and EcoK R-M

systems (13, 14).

Endonuelease eompartmentation, The localization ofBstVI endonuclease and genornic DNA in

different compartments, could also explain the viability of r+m" clones. Since some BstVI activity has

been consistently detected in the supernatants of culture media after cell sedimentation, the sorting

of the enzyme into the periptasmic space and/or its secretion to the extracellular medium, as

previously described for TthI (15) and PstI (16), was investigated. Spheroplasts derived from pP,-

containing E. coil cells were prepared following three different standard experimental approaches:

lysozyme, CaCI~ and sucrose treatments. Analogous results were obtained in all cases. Nearly all the

endonucleolytic activity was found in the particulate fraction (containing the spheroplasts) and only

4% of it could be detected in the supernatant fluids. Cell lysis was monitored by assaying the cytosolic

enzymes LDH and PEPCK (not shown). The data obtained indicated that the endonucleolytic activity

in the superuatant fractions was not a consequence of cell rupture, so its origin is being further

investigated.

393

Vol. 44, No. 2, 1998 BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL

In spite that BstVI endonudease has been recovered sistematically as a soluble protein, we decided

to explore the possibility that a fraction of the enzyme could be present as inclusion bodies-like

aggregates in E. coll. This assumption was based in the fact that many foreign proteins are

insolubilized upon expression in heterologous hosts (17-19). Electron micrographs of pBluescript-

and pP,-carrying cells were indistinguishable, and inclusion bodies were not evident. Similar results

were obtained when various standard protocols designed to purify inclusion bodies were employed

(not shown). Although such structures could not be visualized, the formation of microaggregates of

BslVI endonuc!ease cannot be ruled out.

Studies of in vivo restriction. It had been sistematically observed that recombinant r~m - clones grew

slower than those containing the vector alone. As this observation was indicative of the expression

&the BslVI R-M system to some extent, experiments were undertaken to evaluate quantitatively the

behaviour of the BstVI endonuclease in vivo, as judged by J~ bacteriophage infection. The bstVIR

gene, alone or as part of the complete BstVI R-M system, was mobilized into a host sensitive to ~,

phage. Competent E. coli LE392 cells were transformed individually with control (pBluescript) and

recombinant [pE 3 (r+m +) and PP1, pC5 and pS~ (r+m')] plasrnids (7). Transformants were selected for

ampicillin resistance. Cell extracts were prepared in each case and all of them but those from

pBluescript-containing clones, exhibited endonuclease activity in vitro. The cells were then infected

with ~. ci857 bacteriophage and the resulting plaques counted the day after. The results are

summarized in Table 1. The plating efficiency of all the r+m clones was indistiguishable from that

of the r'm" control. This result suggested that the enzyme, besides of being unable to degrade the

host's chromosomal or plasmid DNAs, was also not playing the defence physiological role that this

kind of enzymes supposedly perform (1,20).

A possible explanation to the almost negligible restriction observed, could be attributed to the non-

optimal temperature (37~ at which the thermophilic enzyme was functioning in vivo. To assess this,

the temperature of phage adsorption was increased to give the enzyme the chance of a better

restriction. Again, the pPt-containing clone showed a plating efficiency at 37~ which was not

different from that of control bacteria. However, restricting activity raised to 10% and 20% when the

adsorption of phage was carried out at 42 and 47~ respectively. In the case of the clone that

harbored the complete BstVI system, plating eff• diminished to 70, 58 and 30% of the control

when phage was adsorbed at 37, 42 and 47~ respectively (Table 1). Although E. coli LE392/pE 3

clone produces M.BstVI, these results are not surprising since it produces as much as ten times more

BstVI endonuelease than the r+m clone (Vasquez, unpublished observations). These results suggest

394

Vol. 44, No. 2, 1998 BIOCHEMISTRYend MOLECULAR BIOLOGY INTERNATIONAL

Table 1

In vivo restriction of L ci857 bacteriophage by E. coli LE392 carrying the bstV1RM genes from B. stearothermophilus V

Plating efficiency E. coli/plasrnid Phenotype (% of control)

37~ 42~ 47oc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

LE 392/none (rm-) 100 100 100

LE 392/pBluescript (r-m-) 105 100 102

LE 392/pE3 (r+m § 70 58 30

LE 392/pP~ (tim) 101 90 81

LE 392/pC5 (r+m ) 95 ND ND

LE 392/pS4 (r+m -) 97 ND ND

The hosts containing the indicated plasmids were infected with ~ ci857 phage as follows: 100 #1 of cells were mixed with 100/ul of an apropriate phage dilution (to give an average multiplicity of int~tion of 0.0002) and incubated at 37~ for 15 rain. The mixtures were then further incubated for 15 rain at the indicated temperatures. Soft agar (0,7%) was then added and the mixtures were poured over Luria-agar (2%) plates which contained Ap (100/~g/ml). Once solidified, plates were incubated overnight at 37~ The numbers shown represent the average of four independent determinations. Plasmids pC5 and pS4 were subcloned from the complete BstVI system as described previously (7). ND, not determined.

that regulatory elements, probably controlling the expression of the bstVlR gene in the homologous

host, had been removed during subcloning. Regulatory elements of this kind have been communicated

for other R-M systems (21, 22). A little ORF coding for a 19-aminoacid peptide was found preceding

the bstVIR gene (8). This sequence shares no homology to the family of regulatory clemens described

by Tao et al. (21) and it is not known if is expressed in B. stearothermophilus V and/or in E. coli.

Since the host's DNA-repair system could contribute significantly to the viability ofr+m clones,

another experimental strategy was devised to evaluate the restriction in vivo by the BstVI R-M

system. To accomplish this, pP~ and pE 3 plasmids were transferred to E. coli ES 130lmutS. After

bacteriophage infection, a strong increase in restricting activity was observed in both cases, and

plating efficiency diminished to less than 4% from that of control (Table 2). These results were

intriguing, since the endonuclease should produce staggered cuts instead of mismatchs. Attempts to

395

Vol. 44, No. 2, 1998 BIOCHEMISTRY and MOLECULAR BIOLOGY INTERNATIONAL

Table 2

In vivo restriction of ~ ci857 phage by E. coli 1301mutS carrying the bstVIRM genes from B. stearothermophilus V

Plating efficiency

E. coli/plasmid Phenotype ( % of control) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ES 130 lmutS/pBluescript (rm) 100,0 ES 130 lmutS/pP1 (r+m ) 3,5 ES1301mutS/pE 3 (r*m ~) 1,4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E. coli ES 1301mutS cells which carried the indicated plasmids were infected with ~, ci857 phage essentially as indicated in the legend to Table 1. The numbers are the mean of four independent determinations.

transfer the BstVI R-M system to other DNA-repair defective E. coil strains have been unsuccesful

so far. This probably reflects the inherent difficulty in the establishment of R-M systems into naive

hosts. We have also failed in transferring the BstVI R-M system into B. sutilis. In this case, it is

possible that the endonuclease be produced at higher levels in this bacterium than in E. coli, as occurs

in the case of the BamHI R-M system (23).

Taken together, these results suggest that BstVI endonuclease, although poorly, is actually active

in vivo, as opposed to the isoschizomeric PaeR7 r+m clones (24). Thus, the viability of BstVI

restriction-proficient modification-deficient E. coil cells could be explained, at least in part, invoking

the participation of the host's own DNA repair system, and a poor restricting activity of the

heterologous endonuclease in vivo.

ACKNOWLEDGEMENTS

This work was supported in part by Grants from Direcci6n de Investigaci6n, Universidad de Santiago de Chile and Grant 1950440 from Fondecyt.

REFERENCES

1. Bickle, T. and Krtiger, D. (1993) Microbiol. Rev. 57, 434-450. 2. Roberts, R. and Macelis, D. (1997) Nucleic Acids Res. 25, 248-262. 3. Gingeras, T., Myers, P., Olson, J., Hanberg, E and Roberts, R. (1978) J. Mol. Biol. 118, 113-122. 4. Jacoby, G. and Sutton, L. (1977) Plasmid 1, 115-116. 5. Vasquez, C. (1985) Biochem. Intern. 10, 655-662.

396

Vol. 44, No. 2, 1998 BIOCHEMISTRYGnd MOLECULAR BIOLOGY INTERNATIONAL

6. Barra, R., Chiong, M., Gonzalez, E. and V~isquez, C. (1988) Biochem. J. 255, 699-703. 7. Vhsquez, C., Saavedra, C. and Gonzhlez, E. (1991) Gene 102, 83-85. 8. Gonz~ilez, E. and Viisquez, C. (1993) Gene 131,103-106. 9. Gonz~lez, E., PadiUa, C., Saavedra, C. and Vfisquez, C. (1994) Microbiology 140, 1337-1340. 10. Sambrook, J., Fritsck, E.F. and Maniatis, T. (1989) Molecular Cloning: a laboratory manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 11. Lowry, O., Rosebrough, N., Farr, A. and Randall, R. (1951) J. Biol. Chem. 193, 265-275. 12. Wilson, G. (1991) Nucleic Acids Res. 19, 2539-2566. 13. Bandyopadhyay, P., Studier, W., Hamilton, D. and Yuan, R. (1985) J. Mol. Biol. 182, 567-578. 14. Krtiger, D and Bickle, T. (1983) Microbiol. Rev. 47, 345-360. 15. Venegas, A., Vicufia, R., Alonso, A., Vatdes, F. and Yudelevich, A. (1980) FEBS Lett. 109, 156-158. 16. Walder, R., Hartley, J., Donelson, J. and Walder, J. (1981) Proc. Natl. Acad. Sci. USA 78, 1503- 1507. 17. Mitraki, A. and K~ng, J. (1989) Biotechnology 7, 690-697. 18. Schein, H. (1989) Biotechnology 7, 1141-1149. 19. Moore, J., Uppal, A., Maley, F. and Maley, G. (1993) Prot. Exp. Pur. 4, 160-163. 20. Matic, I., Taddei, F. and Radman, M. (1996) Trends Microbiol. 4, 69-72. 21. Tao, T., Bourne, J. and Blumenthal, R. (1991) J.Bacteriol. 173, 1367-1375. 22. Ires, C., Sohail, A. and Brooks, J. (1995) J. Baeteriol. 177, 6313-6315. 23. Brooks, J., Landry, D., Sznyter, L., Waite-Rees, P., Ives, C., Moran, L., Slatko, B. and Benner, J. (1991) Nucleic Acids Res. 19, 841-850. 24. Gingeras, T. and Brooks, J. (1983) Proc. Natl. Acad. Sci. USA 80, 402-406.

397