Molecular Biology, Vol. 34, No. 6, 2000, pp. 913–920. Translated from Molekulyarnaya Biologiya, Vol. 34, No. 6, 2000, pp. 1065–1073. Original Russian Text Copyright © 2000 by Babkina, Evstafieva, Chichkova, Vartapetian, Müller, Baskunov, Petrauskene, Kochetkov, Gromova.

UDC 547.963.32.057:542.95

Recombinant Components of the EcoRII Restriction– Modification System: Restriction Endonuclease Can Interact with DNA·RNA Duplexes O. V. Babkina1,3, A. G. Evstafieva1, N. V. Chichkova1, A. B. Vartapetian1, S. Müller2, V. B. Baskunov1, O. V. Petrauskene1, S. N. Kochetkov1,3, and E. S. Gromova1 1

Belozersky Institute of Physico-Chemical Biology and Chemical Faculty, Moscow State University, Moscow, 119899 Russia; E-mail: [email protected] 2 Department of Chemistry, Humboldt University, Berlin, 10099 Germany 3 Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, Moscow, 117984 Russia Received May 29, 2000

Abstract—To obtain recombinant restriction endonuclease (R) and methylase (M) of the EcoRII restriction– modification system, bacterial strains overproducing their functional hexahistidine derivatives were constructed. Active full-length R· EcoRII was produced only in cells that also expressed M· EcoRII from a multicopy plasmid. Recombinant R· EcoRII bound with hybrid DNA·RNA duplexes. Key words: EcoRII restriction endonuclease, EcoRII methylase, histidine derivatives, DNA–RNA hybrids, nucleic acid–protein interaction

INTRODUCTION Restriction–modification systems (RMS) type II are common among bacteria, protecting cells from foreign DNA [1]. Restriction endonucleases (REs, R.) are strongly site-specific and hence are widely employed in gene engineering. One of the Escherichia coli RMS, EcoRII, includes endonuclease and methyltransferase (M.), both recognizing 5'↓CC(A/T)GG. Homodimeric R · EcoRII cleaves DNA (cleavage site indicated with arrow) and monomeric M · EcoRII methylates C5 of the internal cytosine (underlined) in the recognition site. To study their mechanism of action and functional topography, the EcoRII enzymes must be purified in appreciable amounts. One way is to obtain recombinant proteins tailed with hexahistidine. In this work we constructed several hybrid plasmids carrying the genes for R · EcoRII and M · EcoRII, obtained E. coli cells overproducing the enzymes, and preparatively isolated the individual recombinant products. It is known that R · EcoRII is a type IIe RE and, like other such endonucleases, cleaves DNA via a twosubstrate mechanism [2]. To form a catalytic complex, the dimeric enzyme must simultaneously interact with two recognition sites. This offers new possibilities of using R · EcoRII and the other IIe REs. For instance, R · EcoRII can be assumed to form the enzyme–substrate complexes with RNA · RNA and hybrid

DNA · RNA duplexes just as it does with two DNA fragments. The interaction of REs with structures containing RNA strands is poorly known. As earlier shown, type II REs do not cleave RNA, while EcoRI, HaeIII, HhaI, HindII, and MspI cleave DNA · RNA hybrids [1, 3]. More recently, BamHI and Sau3A have been reported to cleave the DNA strand in DNA · RNA hybrids [4]. There is evidence that REs also cleave DNA duplexes with point ribonucleotide inserts [5, 6]. Here we assess the interactions of R · EcoRII with short synthetic RNA · RNA and DNA · RNA duplexes. Such data are of interest for construction and analysis of hairpin ribozymes [7] and ribonucleoprotein complexes of a ribozyme and R · EcoRII and for elucidation of the cooperative function of RNA in these complexes. EXPERIMENTAL Reagents. We used T4 polynucleotide kinase, T4 DNA ligase, Taq DNA polymerase, Klenow fragment of E. coli DNA polymerase, REs (SphI, änI, êstI, LJmçI, êvuII, SmaI, and çindIII), and pUC19 from Fermentas (Lithuania). Expression vectors pQE31, pQE32, and pQE70 (Qiagen, Germany) were used to transform E. ÒÓli JM109 cells. Vectors pR224 and pT71 were kindly provided by A.S. Bhagwat (United States). We also used IPTG (Research Organics),

0026-8933/00/3406-0913$25.00 © 2000 MAIK “Nauka /Interperiodica”

914

BABKINA et al.

Oligonucleotide duplexes used and their thermodynamic stability Duplex I II III IV

Structure 5'-d(AGAGCCAGGTTGGC)-3' 3'-(TCTCGGTCCAACCG)d-5' 5'-r(AGAGCCAGGUUGGC)-3' 3'-(TCTCGGTCCAACCG)d-5' 5'-d(AGAGCCAGGTTGGC)-3' 3'-(UCUCGGUCCAACCG)r-5' 5'-r(AGAGCCAGGUUGGC)-3' 3'-(UCUCGGUCCAACCG)r-5'

Tm , °C* 58 57 52 65

* The accuracy was ±1°C.

15 Ci/mol S-adenosyl-L-[3H]methionine (Amersham, United States), and 1000 Ci/mol [γ-32P]Äíê (Izotop, Russia). Oligonucleotide synthesis. Oligoribonucleotides [7] and oligodeoxyribonucleotides were synthesized on a Pharmacia synthesizer by the standard amidophosphite method, using commercial reagents and solvents. Plasmids. The ecoRIIR gene was obtained from pR224, a derivative of pR209 containing the E. coli RMS genes [8]. To construct pQER2, the ShI site was generated upstream of the ecoRIIR start codon with the use of PCR, and two gene fragments were ligating into the ShI and änI sites of pQE32. To obtain the 5' gene fragment, the êstI-LJmçI fragment (about 700 nt) of pR224 was ligated into the corresponding sites of pUC19, the resulting plasmid was used as a template for PCR with direct primer M13 and primer R 5'-ÄÄëÄGGGëÄíGëííÄíG-3' (the ShI site is underlined), and the product (about 350 nt) was digested with ShI and BamHI. The 3' fragment of eÒÓRIIR was excised with BamHI and KpnI from the plasmid obtained by ligating the BamHI-êvuII fragment (about 1500 nt) of pR224 into the BamHI-SmaI sites of pUC19. The PCR-amplified DNA fragment was sequenced by Sanger’s method [9]. The enzyme encoded by pQER2 contained additional MetArgGlySerçis6GlyIleArg at the N end. To construct pQER10, DNA fragments coding for the N and C regions of R · EcoRII were consecutively inserted in pQE70, with the termination codon removed via PCR. First, the 324-nt ShI–BamHI fragment of pQER2 was ligated into the corresponding sites of pQE70. The resulting plasmid was cleaved with BamHI and ligated with the 882-nt fragment which was obtained by digesting the product amplified from pR224 with primers R and R3 (5'dëGGGATëCíGGÄÄíÄíëíGëGíÄÄÄG-3', the BamHI site is underlined) with BamHI. Orientation of the two fragments was checked with ShI. The

enzyme encoded had additional GlyêheÄrgSerHis6 at the C end. To construct pQER15, two DNA fragments were inserted into the ShI and êstI sites of pQE32. One was the 363-nt ShI-çindIII fragment of pQER2 which coded for the N-terminal region of R · EcoRII. The other fragment of about 3900 nt contained the rest of ecoRIIR, the intercistronic region, and ecoRIIM with its own promoter. To obtain this fragment, pR224 was digested with êstI and EcoRI and the 5' ends were filled with Klenow fragment. The resulting DNA fragment was ligated with êstI linkers and digested with PstI and çindIII. The 3.9-kb fragment was electrophoretically purified in 0.8% agarose gel. The resulting plasmid coded for the same N-terminal sequence of R · EcoRII as pQER2. To construct pQEMet2, the 1.7-kb êstI–BamHI fragment of pT71 [10] was inserted into the corresponding sites of pUC19. The fragment was then excised with ShI and änI and cloned into the corresponding sites of pQE31. The resulting plasmid codes for M · EcoRII with N-terminal MetArgGlySerçis6ThrAspProHisAlaCysSerGlyGlyAsn. Isolation of R · EcoRII. Overnight cultures of E. coli JM109 cells transformed with pQER15, pQER2, or pQER10 and grown in L broth (LB) [11] supplemented with 0.15 mg/ml ampicillin were diluted 1:20 in the same medium and grown at 37°ë with intense aeration for 3 h, to éD550 = 0.6. Transcription of ecoRIIR was induced with 1 mM IPTG. Expression was carried out at 37°ë for 2 h. Cells were collected by centrifugation at 4000 rpm for 10 min and disrupted by sonication at 22 kHz. In express isolation, cells were sonicated in buffer A (20 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5% (v/v) glycerol, 0.01% Triton X-100, 7 mM mercaptoethanol (ME), 17 µg/ml PMSF). Cell fragments were sedimented by centrifugation at 18,000 rpm in a Beckman ultracentrifuge (United States). The pellet (inclusion bodies) was dissolved in buffer B (buffer A containing 6 M guanidine hydrochloride (Gua-HCl) in place of PMSF). The supernatant was combined with Ni-NTA agarose and incubated for 1 h with continuous stirring. The affinity sorbent with immobilized protein was pelleted by centrifugation at 12,000 rpm and washed with buffer C (buffer A containing 25 mM imidazole). The protein was eluted with buffer D (20 mM Tris-HCl, pH 6.3, 500 mM imidazole) and analyzed by SDS-PAGE. The activity of R · EcoRII was tested with pBR322 (37°C, 30 min). To isolate R · EcoRII from inclusion bodies, buffers C and D were supplemented with 6 M Gua-HCl. The enzyme was renatured via dialysis against 2 × storage buffer (20 mM Tris-HCl, pH 7.6, 100 mM NaCl, 2 mM DTT, 5% glycerol) overnight. MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

R · EcoRII INTERACTS WITH DNA · RNA DUPLEXES SphI BamHI

pQER2 PT5

RBS

His6

SphI

pQER10 PT5

RBS

KpnI

ecoRIIR

BamHI

to

BamHI

ecoRIIR

His6

to

SphI HindIII

pQER15 PT5

RBS

His6

PT5

RBS

PstI

ecoRIIR

SphI PstI

pQEMet2 His6

915

ecoRIIM

to

BamHI KpnI ecoRIIM

to

Fig. 1. Plasmids constructed on the basis of pQE: PT5, T5 promoter containing two regions of the lac operator; RBS, ribosomebinding site; His6, DNA sequence coding for six histidine residues; to, transcription terminator of phage λ.

Preparative isolation of R · EcoRII on a column with Ni-NTA agarose followed a published protocol [12]. The protein was eluted with 100–500 mM imidazole in buffer E (50 mM NaH2PO4, pH 6.0, 1 M NaCl, 10% glycerol, 0.1% Triton X-100, 10 mM ME). The R · EcoRII activity was detected in fractions containing 200–400 mM imidazole. With 0.5 l of cell culture, the procedure yielded a preparation containing 12 mg protein (purity more than 90%, protein concentration 1 mg/ml, activity 100 units/µl). Isolation of M · EcoRII. E. coli JM109 cells transformed with pQEMet2 were grown as above and disrupted in buffer F (10 mM Tris-HCl, pH 8.0, 500 mM NaCl, 20% glycerol, 8 M urea). After centrifugation, the supernatant was combined with Ni-NTA agarose and incubated for 3.5 h at room temperature. The sorbent was washed with buffer F containing 25 mM imidazole, and the protein was eluted with 500 mM imidazole at pH 6.3. The enzyme was renatured via dialysis against 2 × storage buffer (40 mM Tris-HCl, pH 7.5, 400 mM NaCl, 2 mM EDTA, 14 mM ME, 5% glycerol). The activity of M·EcoRII was inferred from incorporation of 3ç in DNA duplex I (Cd 0.3 mM) in 30 min at 20°ë in 10 µl of 10 mM Tris-HCl (pH 7.5) containing 50 mM NaCl, 15 mM MgCl2, 0.1 mM DTT, and tritiated S-adenosyl-L-methionine. MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

Circular dichroism (CD) spectra were obtained in a Jouan-III unit (France), with oligonucleotide duplexes used at ëd 3 µM, in buffer G (40 mM TrisHCl, pH 7.6, 50 mM NaCl). Binding and hydrolysis of oligonucleotide duplexes with R·EcoRII. Oligonucleotides were 5'endlabeled with T4 polynucleotide kinase (10 units/µl) and [γ-32ê]-Äíê. Complexation of R · EcoRII with duplexes I–IV (table) was analyzed in 7–15% gradient PAG, in nondenaturing conditions. DNA, RNA, and DNA · RNA duplexes (ëd 0.6 µM) were incubated with R · EcoRII (6 µM/monomer) in 10 µl of buffer G containing 7 mM DTT and 8% glycerol (buffer H) with or without 10 mM CaCl2. Samples were incubated 10 min at 37°ë and then 20 min at 4°ë. Hydrolysis of oligonucleotide duplexes (ëd 0.3 µM) with R · EcoRII (2 units) was carried out at 37°ë in 10 µl of buffer H containing 5 mM MgCl2 for 30 min. The reaction mixtures were electrophoretically separated in 20% PAG in the presence of 8 M urea, gels were autoradiographed. To activate hydrolysis of the noncleavable analog of R · EcoRII substrate III, the reaction was carried out in the presence of 0.3 µM substrate I.

916

BABKINA et al. (a) M, kDa

1

2

(b) 3

M, kDa

1

2

(c) 3

M, kDa

45

45

45

25

25

25

1

2

Fig. 2. Production of R·EcoRII and its inactive derivatives in E. coli cells transformed with (a) pQER15, (b) pQER2, and (c) pQER10. Total protein from cells grown (a, 2; b, 1) without and (a, 3; b, 2; c, 2) with IPTG and (a, 1; b, 3; c, 1) protein eluted from Ni-NTA agarose were separated by SDS-PAGE in 15% gel. M, markers.

RESULTS AND DISCUSSION Recombinant Enzymes of the EcoRII Restriction–Modification System The EcoRII RMS genes have been cloned and the corresponding proteins purified to homogeneity via chromatography on several sorbents [13–17]. We tried to obtain His6 derivatives of the EcoRII enzymes which can be isolated in one step, via affinity chromatography of Ni-NTA agarose. Reuter et al. [12] have 1

2

3

4

5

briefly described such a construct for R · EcoRII, and have demonstrated that the enzyme containing His6 remains active and specific. To obtain a strain overproducing R · EcoRII, three plasmids were constructed (Fig. 1). Plasmids pQER2 and pQER10 code for R · EcoRII containing His6 at the N and C ends, respectively. Plasmid pQER15 codes for R · EcoRII with N-terminal His6 and M · EcoRII. In all plasmids, ecoRIIR is controlled by the bacteriophage T5 promoter containing the operator regions of the lac operon. The M · EcoRII gene is controlled by its own constitutive promoter and is transcribed in the opposite direction (Fig. 1). Nucleotide sequence analysis of the constructs showed that the ecoRIIR sequence differs from one published sequence [8] by insertion of G between í675 and G676 and deletion of ë710, and is identical to another [14].

Fig. 3. Digestion of pBR322 with R·EcoRII isolated from E. coli cells transformed with (2) pQER2, (3) pQER10, and (4) pQER15. Lane 1, intact plasmid; 5, pBR322 hydrolyzed with commercial R·EcoRII (Fermentas, Lithuania). Samples were resolved in 0.8% agarose gel.

As expected, cells transformed with pQER15 produced a 46 kDa protein (Fig. 2a). A protein expressed from pQER2 (about 30K) was shorter than R · EcoRII (Fig. 2b). Cells transformed with pQER10 produced a 46K protein which did not differ in electrophoretic mobility from R · EcoRII, but its production was markedly lower than with pQER2 and pQER10 (Fig. 2c). With all plasmids, synthesis of the recombinant proteins was significantly higher in the presence of IPTG, while only trace amounts of the proteins were detected in its absence. Specific activity of R ·EcoRII was detected in preparations obtained with pQER15, but not with pQER2 and pQER10 (Fig. 3). Therefore, transformants carrying pQER15 were used in further experiments. MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

R · EcoRII INTERACTS WITH DNA · RNA DUPLEXES

In the above experiments, the proteins were isolated from the soluble fraction according to the express method (see Experimental). However, we found that most RMS EcoRII proteins occur in inclusion bodies and developed methods for isolating virtually homogeneous R · EcoRII (pQER15) and M · EcoRII (pQEMet2) from the soluble fraction (Fig. 4) and from inclusion bodies. The express method yielded preparations with a low protein concentration (0.3–0.4 mg/ml), whereas chromatography on Ni-NTA agarose proved suitable for large-scale isolation of R · EcoRII and yielded more concentrated preparations (1 mg/ml). As shown above, active R · EcoRII was obtained only with pQER15 which also encodes M· EcoRII. The plasmids coding for only R · EcoRII directed production of inactive proteins. Possibly, R · EcoRII has a toxic effect on E. coli and even on dcm+ cells, which results in selection of cells producing mutant proteins devoid of enzymic activity. Indeed, pQER2 determined production of inactive shortened R · EcoRII. This probably resulted from a nonsense mutation occurring about 800 nt away from the start codon. Although similar in electrophoretic mobility to the full-length R · EcoRII, its mutant form synthesized from pQER10 acted as a weak nuclease but lacked the specific R · EcoRII activity (Fig. 3). The lethal effect of endonuclease gene expression has been earlier observed when the methylase gene was inactivated by Tn5 insertion [15]. Thus, like with most type II endonucleases, expression of R · EcoRII in the absence of M · EcoRII is lethal. Plasmids used to produce R · EcoRII commonly contain two genes, ecoRIIR and ecoRIIM [8, 16–18]. In the only exception [12], cells which carry a multicopy plasmid containing only the endonuclease gene have been maintained viable. There is evidence that M · SsoII inhibits expression of its own gene and enhances expression of the R · SsoII gene [19]. Matvienko et al. [18] have suggested a translational (posttrancriptional) control of endonuclease gene expression in the EcoRII RMS. In our plasmids, ecoRIIR was under the control of T5 promoter containing the operator regions of the lac operon. When transcription was induced with IPTG, synthesis of R · EcoRII increased to a similar extent in cells transformed with pQER2 and pQER15, regardless of whether cells produced M · EcoRII or not (Fig. 2). These data suggest that expression of ecoRIIR in our system is not regulated by M · EcoRII. Interaction of R·EcoRII with RNA·RNA and Hybrid DNA·RNA Duplexes To study the interaction of R·EcoRII with substrates of unusual conformation, we constructed 14-mer DNA (I), RNA (IV), and hybrid DNA · RNA (II, III) MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

M, kDa

1

917 2

66 45

Fig. 4. SDS-PAGE in 15% gel of (1) R·EcoRII and (2) M·EcoRII purified on Ni-NTA agarose. M, markers.

duplexes which contained the EcoRII site (I) (table). The duplexes are analogous in sequence, except that U was used in place of T in RNA strands. To characterize the thermodynamic stability and conformation of the duplexes, we estimated their melting temperature (Tm) and obtained their CD spectra. As expected, the Tm of RNA duplex IV was higher than that of DNA duplex I (table). The CD spectra of duplexes I and IV corresponded to the B and A conformations, respectively [20] (Fig. 5). Hybrid duplexes II and III were less stable than RNA duplex IV, duplex III displaying considerably lower Tm than duplex II (table). In CD spectra, duplexes II and III markedly differed from the homogeneous duplexes, being more similar to the A form (Fig. 5). These duplexes only slightly differed from each other, the CD spectrum of duplex II being closer to that of ADNA. This is consistent with the data that DNA · RNA oligonucleotide hybrids differ in conformation from both A- and B-DNA, being more similar to the former [21, 22]. DNA · RNA duplexes with the DNA strand mostly consisting of purines (dR·rY) are less thermodynamically stable and differ in conformation from analogous duplexes with a purine-rich RNA strand (rR·dY) [21, 22]. It has been assumed that rR·dY duplexes are more similar to RNA, whereas the structure of dR·rY duplexes is intermediate between DNA and RNA [21, 22]. The DNA strand was purine-rich in duplex III and pyrimidine-rich in duplex II, and the former melt at a

918

BABKINA et al. ∆ε 6 5 4 3

4

2

2

3

1

1

0 –1 –2 220

240

260

280

300

320 λ, nm

Fig. 5. CD spectra of duplexes (1) I, (2) II, (3) III, and (4) IV in 40 mM Tris-HCl (pH 7.6) containing 50 mM NaCl; C d 3 µM.

lower temperature. Possibly, duplexes II and III differ in conformation, the latter being closer to B-DNA. We studied whether R · EcoRII hydrolyzes duplexes II–IV analogous to its substrate. To detect cleavage of individual strands, the 32P label was introduced in either strand of the duplex. R · EcoRII did not cleaved any strand of duplexes II–IV (data not shown). Complexation of R · EcoRII with duplexes I–IV was studied using the gel retardation assay (Fig. 6). The enzyme was used in a tenfold excess with respect to a duplex, and binding was carried out in the absence of Mg2+ or in the presence of Ca2+. It is known that Ca2+ does not act as a cofactor of REs but is able to replace Mg2+ in the enzyme catalytic center and thereby to inhibit hydrolysis [23]. In addition, we verified that R · EcoRII did not cleave the canonical substrate I in the presence of Ca2+. Either without Mg2+ or with Ca2+, R · EcoRII bound only with hybrid duplex III which had the upper DNA strand and the lower RNA strand (Fig. 6). Complexation was not observed with RNA duplex IV and hybrid duplex II. The canonical substrate I formed two complexes, “upper” and “lower,” with R · EcoRII (Fig. 6). We have previously shown that the less mobile complex consists of a dimer of the protein bound with two oligonucleotide substrates, while the

lower complex is formed by one protein and one substrate [24]. With duplex III, only the upper, dimeric complex was formed. As earlier shown, 14-mer substrate analogs, which contain modified heterocyclic bases and are resistant to hydrolysis with R · EcoRII, can be cleaved upon enzyme activation with the canonical substrate of the same size [25]. The critical condition is that the enzyme–substrate complex contain both modified and canonical substrates (the two-substrate mechanism of hydrolysis). We tried to activate hydrolysis of duplex III by adding duplex I in the reaction mixture. However, R · EcoRII did not cleave the hybrid duplex even in the presence of the canonical substrate (data not shown). Our data demonstrate that the A or similar conformation of the recognition site strongly prevents its cleavage with R · EcoRII in DNA·RNA duplexes. In addition to different dimensions of the major and minor grooves in the A conformation, 2'-OH groups may sterically hinder formation of the catalytic complex. However, R · EcoRII binds DNA·RNA duplexes which have a purine-rich DNA strand and hence are similar in conformation to B-DNA. Thus, we have elaborated efficient one-step methods for isolating R · EcoRII and M · EcoRII. The data on the interaction of R · EcoRII with RNA and RNA·DNA duplexes suggest that the enzyme can be MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

R · EcoRII INTERACTS WITH DNA · RNA DUPLEXES I

(a) 1

II 2

3

919

III 4

5

IV 6

7

8

Upper complex Lower complex

Oligonucleotide duplexes

I

(b) 1

II 2

3

III 4

5

IV 6

7

8

Upper complex Lower complex

Oligonucleotide duplexes

Fig. 6. Binding of R·EcoRII with duplexes I–IV in buffer H in the presence (a) or absence (b) of Ca2+ (10 mM). The enzyme (6 µM/monomer) was incubated with duplexes (2) I, (3, 4) II, (5, 6) III, and (7, 8) IV (Cd 0.6 µM) 32P-labeled at the upper (1–3, 5, 7) or lower (4, 6, 8) strands. Lane 1, duplex I incubated without the enzyme. Samples were resolved in nondenaturing 7–15% gradient gel; the gel was autoradiographed.

used to construct model ribonucleoprotein complexes in studies on ribozymes.

technology (Enzyme Engineering), and a grant from INTAS (IR-97-1635).

ACKNOWLEDGMENTS

REFERENCES

We are grateful to A.S. Bhagwat (United States) for plasmids pR224 and pT71 and to O.I. Andreeva for recommendations on isolation of R · EcoRII. This work was supported by a grant from Howard Hughes Medical Institute (HHMI 75 195-54 5501), the Russian Foundation for Basic Research (project no. 9804-49 108), the State program Novel Methods in Bio-

1. Modrich, P. and Roberts, R.J., Nucleases, Linn, S.M. and Roberts, R.J., Eds., New York: Cold Spring Harbor Lab. Press, 1982, pp. 109–154. 2. Krüger, D.H., Kupper, D., Meisel, A., et al., FEMS Microbiol. Rev., 1995, vol. 17, pp. 177–184. 3. Molloy, P.L. and Symons, R.H., Nucleic Acids Res., 1980, vol. 8, pp. 2939–2946.

MOLECULAR BIOLOGY

Vol. 34

No. 6

2000

920

BABKINA et al.

4. Nazarenko, I.A., Gorbunov, Yu.A., and Malygin, E.G., Bioorg. Khim., 1987, vol. 13, pp. 928–933. 5. Ohtsuka, E., Ishino, J., Ibaraki, K., and Ikehara, M., Eur. J. Biochem., 1984, vol. 139, pp. 447–450. 6. Gromova, E.S., Kubareva, E.A., Vinogradova, M.N., et al., J. Mol. Recogn., 1991, vol. 4, pp. 133–141. 7. Schmidt, C., Welz, R., and Müller, S., Nucleic Acids Res., 2000, vol. 28, pp. 886–894. 8. Bhagwat, A.S., Johnson, B., Weule, K., and Roberts, R.J., J. Biol. Chem., 1990, vol. 265, pp. 767–773. 9. Sanger, F., Nicklen, S., and Coulson, A.R., Proc. Natl. Acad. Sci. USA, 1977, vol. 74, pp. 5463–5467. 10. Wyszynski, M.W., Gabbara, S., and Bhagwat, A.S., Nucleic Acids Res., 1992, vol. 20, pp. 319–326. 11. Sambrook, J., Fritch, E.F., and Maniatis, T., Molecular Cloning: A Laboratory Manual, New York: Cold Spring Harbor Lab. Press, 1989, 2nd ed. 12. Reuter, M., Kupper, D., Meisel, A., et al., J. Biol. Chem., 1998, vol. 273, pp. 8294–8300. 13. Kosykh, V.G., Buryanov, Y.I., and Bayev, A.A., Mol. Gen. Genet., 1980, vol. 178, pp. 717–718. 14. Kosykh, V.G., Repik, A.V., Kaliman, A.V., et al., Dokl. Akad. Nauk SSSR, 1989, vol. 308, pp. 1497–1499.

15. Som, S., Bhagwat, A.S., and Friedman, S., Nucleic Acids Res., 1987, vol. 15, pp. 313–332. 16. Kosykh, V.G., Puntezhis, S.A., Bur’yanov, Ya.I., and Baev, A.A., Biokhimiya, 1982, vol. 47, pp. 619–625. 17. Kupper, D., Reuter, M., Mackeldanz, P., et al., Protein Express. Purif., 1995, vol. 6, pp. 1–9. 18. Matvienko, N.N., Zheleznaya, L.A., Chernysheva, E.E., et al., Biokhimiya, 1997, vol. 62, pp. 1314–1318. 19. Karyagina, A.S., Shilov, I.A., Tashlitsky, V.N., et al., Nucleic Acids Res., 1997, vol. 25, pp. 2114–2120. 20. Ivanov, V.I., Minchenkova, I.E., Minyat, E.E., and Frank-Kamenetskii, M.D., J. Mol. Biol., 1974, vol. 87, pp. 817–833. 21. Gyi, J.I., Conn, G.L., Lane, A.N., and Brown, T., Biochemistry, 1996, vol. 35, pp. 12538–12548. 22. Gyi, J.I., Lane, A.N., Conn, G.L., and Brown, T., Biochemistry, 1998, vol. 37, pp. 73–80. 23. Vipond, I.B., Baldwin, G.S., and Halford, S.E., Biochemistry, 1995, vol. 34, pp. 697–704. 24. Karpova, E.A., Kubareva, E.A., Gromova, E.S., and Buryanov, Ya.I., Biochem. Mol. Biol. Int., 1993, vol. 29, pp. 113–121. 25. Petrauskene, O.V., Gromova, E.S., Romanova, E.A., et al., Gene, 1995, vol. 157, pp. 173–176.

MOLECULAR BIOLOGY

Vol. 34

No. 6

2000