srl-300 ::Tn 10. Mac, crpA39 fnr.Jn 10 ara9 ara9 galE503 bio203:Jn10. pepT7::Mudi(Ap"/ac cts62 X) leuBCD485 oxrA 1 zda-888:Jn10 crp773:.Jn10trpB223 bla.
Molecular Microbiology (1993) 9(1), 165-172
Regulation of the ansB gene of Salmonella enterica Michael P. Jennings, Shaun P. Scott and Ifor R. Beacham* Division of Science and Technology, Griffith University, Nathan, Queensland, Australia 4111.
Summary The expression of L-asparaginase il (encoded by ansB) in Salmonella enterica was found to be positively regulated by the cAiVIP receptor protein (CRP) and anaerobiosis. The anaerobic reguiation of the S. enterica ansB gene is not mediated by the anaerobic transcriptional activator FNR. This is uniike the situation of the ansB gene of Escherichia coli, which is dependent on both CRP and FNR. To investigate this fundamental difference in the reguiation of L-asparaginase il expression in S. enterica, the ansB gene was cioned and the nucieotide sequence of the promoter region determined. Sequence anaiysis and transcript mapping of the 5' promoter region revealed a singie transcriptionai start point (tsp) and two reguiatory sites with substantiai homoiogy with those found in E. coli. One site, centred -90.5 bp from the tsp, is homoiogous to a hybrid CRP/FNR ('CF') site which is the site of CRP reguiation in the £. coli promoter. The other site, centred 40.5 bp upstream of the tsp, is homologous to the FNR binding site of the £. coli promoter. Significantiy, however, a single basepair difference exists in this site, at a position of the reiated CRP and FNR DNA-binding site consensus sequences i(nown to be invoived in CRP versus FNR specificity. Site-directed mutagenesis indicates that this singie difference, reiative to the homoiogous £. coll site, resuits in a CRP binding site and the observed FNR-independent ansS expression in S. enterica. Thus, not oniy may CRP and FNR sites be interconverted experimentaily, with few nucieotide changes, but this has apparentiy contributed to the evoiutionary divergence of ansB reguiation in £. coli and S. enterica.
introduction Escherichia co//contains two L-asparaginase isozymes: a constitutive cytoplasmic enzyme (L-asparaginase I) of Received 15 December, 1992; revised 21 March 1993; accepted 24 March, 1993. *For correspondence. Tei. (7) 875 7445; Fax (7) 875 7656.
high K^, and a secreted low K^ enzyme (L-asparaginase il) which is regulated by oxygen and by glucose levels in the medium (Cedar and Schwartz, 1967; 1968). The latter enzyme has been intensively studied, in part because of its anti-lymphoma activity (Gallagher ef ai, 1989; Beacham and Jennings, 1990). We have recently studied the regulation of expression of the L-asparaginase IIencoding gene (ansB) from E. coli. This gene is positively regulated by the concurrent action of cAMP receptor protein (CRP) and by the product of the fnr gene (FNR; Russell and Yamazaki, 1978; Chesney, 1983; Jerlstrom etai, 1987; Jennings and Beacham, 1993 — accompanying paper). CRP and FNR are transcriptional activators that mediate responses to carbon source and anaerobiosis, respectively. They are closely related in their primary sequences and the DNA binding sites that they recognise. Consensus sequences for the DNA binding sites of CRP and FNR differ from each other at only two positions, residues 5 and 18, in the 22 bp consensus sequence (Zhang and Ebright, 1990; Bell et ai, 1989; Spiro and Guest, 1990). The two sites at which these regulatory proteins interact to transcriptionally activate ansB, in E. coli, have been identified (Jennings and Beacham, 1993). The first is an FNR site, centred 41.5 bp upstream of the major transcriptional start site. The second site, located upstream of the FNR site, is a hybrid CRP/FNR ('CF') site and is the site of CRP regulation. L-asparaginases in the closely related organism Salmonella enterica have not been investigated. It is well established, however, that by studying genes from closely related organisms, in which regulatory sequences are conserved and/or which display subtle differences in regulation, further insight into regulatory mechanisms can be gained. We report here studies on the regulation of the gene encoding L-asparaginase II, from S. enterica ('Salmonella typhimurium' LT2). We have found that the expression of this gene is regulated by glucose levels in the media, via CRP, and by anaerobiosis. However, the anaerobic regulation of this gene does not act via FNR, in contrast to the case in E. coli (see above). Nucleotide sequence analysis has revealed that the regulatory elements identified in the E. coli ansB promoter are strongly conserved in S. enterica. The observed difference in response to FNR appears to be due, at least in part, to a single base-pair difference which converts the FNR DNAbinding site found in the E. coli promoter to a CRP site in the S. enterica promoter.
166 /W. P. Jennings, S. P. Scoff and I. R. Beacham Results Aerobic
Regulation of L-asparaginase II expression in S. enterica To determine whether L-asparaginase II synthesis in S. enterica is environmentally controlled by glucose and oxygen levels, L-asparaginase assays were performed on cultures of S. enterica grown aerobically or anaerobically, with or without 0.4% glucose. The specific activity of Lasparaginase II in aerobically grown cultures was reduced by approximately 10-fold relative to anaerobic cultures. The level of L-asparaginase activity in anaerobic cultures containing glucose were also reduced by approximately 10-fold (see Fig. 1). To determine whether the observed regulation by glucose and anaerobiosis was mediated by CRP and FNR, respectively, L-asparaginase assays were performed on crp and oxrA mutant strains of S. enterica (oxrA is the S. enterica homologue of the £ coli fnr gene which encodes the FNR protein; Jamieson and Higgins, 1984; Strauch ef ai, 1985). Consistent with the effect of glucose, Lasparaginase II levels in the crp mutant strain were very low, corresponding to about 10% of that in a wild-type strain (Fig. 1). Surprisingly, in view of the effect of aerobic versus anaerobic growth conditions, the level of Lasparaginase II is reduced by less that 50% in the oxrA mutant of S. enterica (Fig. 1). Hence L-asparaginase II synthesis in S. enterica is only moderately influenced by FNR.
Cloning, nucleotide sequence analysis, and transcript mapping of the 5'promoter region of the S. enterica ansB gene Since the mechanism of anaerobic regulation of ansB in S. enterica seems to be a variation of that in E. coli, the promoter region was isolated, and its nucleotide sequence determined (see the Experimental procedures), in order to see if regulatory sequences are conserved and for use in gene-fusion studies. A comparison of the nucleotide sequences of the 5'promoter regions of ansB from S. enterica and E. coli (Fig. 2) reveals 58% identity over a 147 bp region. Most of this homology is due to the conservation in S. enterica of two regulatory elements previously identified as being the sites of CRP and FNR interaction in the £. coli promoter (the CF and FNR sites; see Introduction and Jennings and Beacham, 1993). Homology is also evident between the ribosomal binding sites and one of the -10 regions present in the E. co//promoter proximal to the major transcriptional start point (tsp; Fig. 2). There are a number of notable differences within the regions which are conserved. In the S. enterica homologue of the E. coli FNR binding site (S. enterica nucleotide positions (ntp) -30 to - 5 1 , Fig. 2), there are two changes present in the site
Anaerobic Anaerobic.»^iu
i80galE
TN2062 (oxrA)
PP1037(crp)
Fig. 1. L-asparaginase expression in wild-type (180ga/£) and mutant strains of S. enterica, grown aerobically and anaerobically. 180pa/£was also grown both in the presence and absence of 0.4% glucose. The values are expressed in ixmoles ammonia released minute"' mg protein"' and are the result of at least three independent experiments.
such that the left half-site in S. enterica is not significantly homologous to either an FNR or a CRP site, having only three or one match(es), respectively. These two nucleotide changes also alter a -10 promoter sequence associated with a minor tsp found in the E. co//promoter. A further, potentially significant difference is at position 18 of the same site (ntp -34; see Fig. 2). This is one of the specificity determining bases which differentiate CRP from FNR binding sites: in E. coli this base is an A residue, as in the FNR consensus sequence, whereas in S. enterica it is a C residue as in the CRP consensus binding site. These two conserved sites in the S. enterica promoter are therefore referred to as the CF site (S. enterica ntp -80 to - 1 0 1 , Fig. 2) and the CRP site (S. enterica ntp -30 to - 5 1 , Fig. 2), by analogy to the CF and FNR sites, respectively, in E. coli (see Introduction, Fig. 2). The transcriptional start point (tsp) was determined using both primer extension and SI nuclease analyses (Fig. 3). The analysis revealed a single transcript initiating with an A residue at ntp +1 (Fig. 2). Located 6bp upstream of the tsp is a sequence homologous to the -10 promoter sequence. There is no corresponding -35 sequence present, a feature common in positively regulated promoters (Raibaud and Schwartz, 1984). Analysis of RNA isolated from a culture containing glucose revealed no detectable transcript (Fig. 3A, lane 3), consistent with the expression studies (Fig. 2). However, transcription was observed in the aerobic culture (Fig. 3A, lane 2), despite a strong reduction in ansB expression (Fig. 4). Although the measurement of the transcript by S1 nuclease protection analysis may be regarded as only
Salmonella enterica ansB gene regulation -no
^^-^"gA
TCaAA-tt (FNR)
-70
-60
EC ATGCCGTTTAATTCTTCGTTTTGTTACCTGCCTCTAACT^^^^ aa -TGTgA
Se EC
_2o +1 aa-tgaGa TCACA-Tt; (CRP) • -10 * TCGCTGGTTAATCGTATGGCGTCACATTATTCGTCTGCAATATAGAGATAATGCGACCAGT
::::::::: :::::::: : :
;
Fig, 2, Comparison Of the 5'promoter regions Of
S
TCacA-TT (CRP)
:: TTTTTT:::
::
CACGTTGTAAATTGTTTAACGTCAAATTTCCCATACAGAGCTAAGGGATAATGCGTAGCGT aA-tTTGa TCAAA-TT (FNR) ~^-To *
SD Met Se TGACATAACIGGAGATATAACATGG : :: ::::::::: :: : :::: EC TCACGTAACTGGAGGAATGAAATGG SD Met
semi-quantitative, this discrepancy raises the possibility that the observed regulation of ansB expression by anaerobiosis may be post-transcriptional. Analysis o^ansB expression using an ansB'-'lacZ gene fusion In order to facilitate further investigations of the regulation of the ansB gene, an ansS'-'/acZ fusion, pSSF16, was constructed (see the Experimental procedures and Fig. 2), The fusion contains 143bp of the 5'-untranslated region including the first codon of the ansB signal peptide, followed by /acZ sequence. The effects of glucose and of oxygen levels on p-galactosidase activity in strains containing this fusion are consistent with those observed for L-asparaginase II: a strong reduction in aerobic cultures as compared to anaerobic cultures,-and strong catabolite repression by glucose (Fig. 4), The expression of the fusion was examined in the E. CO//strain M^82recA and was found to be almost identical to the S, enterica strain 180gfa/E (Fig, 4); this permitted the use of a set of isogenic fnr and crp mutants, derived from M182, in further experiments (see the Experimental procedures). Mutational analysis of ttie ansB promoter region In order to determine whether the CF sitd is involved in
167
(-10,
''"'^ '^® Shine-Dalgarno (SD) sequences. The initiating methionine codon is indicated (Met). The underiined regions of the S. enterica sequence, at the beginning and end of the sequence shown, indicate the sequence used to synthesise oiigonucieotides for ampiification of the DNA used to construct the ansS'-'/acZfusion (see the Experimen/a/pracedures).Transcriptionaistart sites (tsp) are indicated by asterisks, and the S. enfer/ca sequence is numbered with respect to the tsp (+1). The CRP and FNR consensus site sequences are shown above and beiow their putative binding sites. Upper case indicates identity with the respective consensus sequence, whereas iower case represents iack of identity; this homoiogy is shown above the sequence in the case of S. enterica, and beiow the sequence for E. coli. The upstream site (nucieotide positions -101 to-80) is referred to as the CF site, and the tsp proximai site as the CRP site (nucieotide positions -51 to-30), in S. enferica, orthe FNR site in £ CO//(see text; Jennings and Beacham, 1993). The S. enter/ca ansSpromoter region sequence will appear in the EMBUGenBank/DDBJ Nucieotide Sequence Data Libraries under the accession number X69868.
a/is6 expression in S, enterica, as it is in E. coii, a deletion was created within a/isS'-7acZfusion which removes this site (to give pSSF16Acf: see Fig. 5B) whilst the downstream CRP site is retained. Expression from this construct is reduced by about fivefold in a wild-type strain, consistent with this conserved site being required for optimal expression (Fig. 5A), The residual activity is still regulated by oxygen and expression in mutant E. coii host strains also demonstrates that ansB expression is FNRindependent and CRP-dependent when the CF site is deleted (Fig, 5A). Thus the CF site is required for optimal expression and is not essential for oxygen- or CRP-mediated regulation. The sequence of the CRP site predicts that the C residue at position 18 of the consensus sequence (nucieotide position -34, Fig. 2) may be required for CRP binding, since only the right-hand half-site has homology to the CRP binding-site consensus sequence and this residue determines FNR or CRP specificity (Bell et ai., 1989). This base was therefore changed to a T residue which is neither an FNR nor a CRP determinant. This mutation was introduced into pSSF16Acf, resulting in pSSF16Acf/crp18 (see Fig. 5B), Assays of p-galactosidase activity expressed from this construct in all strains revealed very low levels of unregulated activity (Fig, 5A). This result is consistent with the identification of the CRP site as a site necessary for CRP interaction. The lack of expression as a result of the C ^ T mutation
168
M. P. Jennings, S. P. Scott and I. R. Beacham
A GATC
B 1 2 3
12
4
G A T C
PROBE
• TSP
Fig, 3, Transcript mapping using SI nuclease protection (A) and primer extension anaiysis (B). in both cases, lanes G, A, T and C represent nudeotide sequencing reactions. The oiigonucleotide corresponding to the transcriptional start point (tsp) is indicated by an arrow. The ^^P-iabelled probe is aiso Indicated by an arrow. A. SI nuciease protection analysis. Lanes 1 to 4 contain 30 ^g of total RNA from cultures of 180ga/£containing pSSF16 grown under different cuiture conditions. The RNA was anneaied with a ^^P-labelled probe, and treated with 100 units of SI nuciease (see the Experimental procedures). Lanes: (1) anaerobic culture; (2) aerobic cuiture; (3) anaerobic culture containing 0.4% glucose; (4) probe only, no SI nuclease. B. Primer extension analysis. Lanes 1 and 2 contain 5 ng and 10 ng of total RNA respectively, annealed with a ^^P-labelled probe, and then extended with reverse transcriptase. Total RNA was isoiated from anaerobic cuitures of 180gaiE containing pSSF16.
• TSP
PROBE
at position 18 (with respect to the consensus sequence) in the right half-site CRP site (see above) also strongly indicates that the base at this single position may be responsible for the lack of FNR-dependent regulation. A C ^ A change was therefore introduced at this position, In the ansS'-7acZ fusion, converting this site to a potential FNR site. Expression of p-galactosidase from this construct (pSSF16ds18; Fig, 5B) showed a fivefold reduction in activity in the wild-type strain compared to pSSF16 (Fig, 5A). This was further reduced in a crp mutant strain, presumably as a result of the lack of cAMP-CRP interaction at the upstream CF site. Significantly, however, the remaining activity was virtually abolished In the crp fnr mutant strain (Fig. 5A), and the expression from the construct was reduced In an fnr strain to a greater degree than pSSF16, These results are consistent with the
change in pSSF16ds18 creating a promoter that is partially responsive to FNR, Discussion The expression of the ansB gene from E, coli and S. enterica is fundamentally different with respect to the mechanism of oxygen regulation. The E, coli gene is dependent on the transcriptional activator FNR whilst anaerobic regulation of the S. enterica gene is essentially FNR-independent. Comparison of the promoter regions of these genes reveals many similarities. Both promoters possess two apparently functional regulatory protein binding sites upstream of a single tsp (major tsp in the case of E, coli). In E, coli it is thought that FNR binds to the downstream
Salmonella enterica ansB gene regulation FNR site to initiate a low level of transcription. This low \evel of transcription is increased when CRP binds at the upstream CF site, resulting in co-dependent regulation by CRP and FNR (Jennings and Beacham, 1993), In the S. enterica promoter, the downstream site has a left-hand half-site homologous to neither CRP nor FNR and a right-hand half-site homologous to a CRP site. The results of site-directed mutagenesis of the specificity determining residue of this putative CRP site, which either inactivate the site or change the specificity of the site to an FNR site, are consistent with the identification of this site as a functional CRP-binding site. Deletion of the upstream CF site results in a reduction in activity whilst the residual activity remains responsive to both CRP and anaerobiosis. This suggests that the S. enterica promoter requires CRP bound at both the CRP and CF sites for optimal expression. The S, enterica promoter has only one tsp. The absence of the minor tsp, present in the E, coli promoter, may be the result of expression below detectable levels. Another possibility is that differences in the S, enterica upstream -10 region, relative to the E. coli homologue, may have resulted in the silencing of the upstream minor tsp in S, enterica.
•
Ml B2recA. aerobic
n
MlSSrecA
0
fnr
0
crp
mil
crpfnr
6000-
PSSF16
B
pSSF16/Sct
pSSF16Acf/crp18
CONSENSUS SEQUENCES TT-TCTgA TCaAA-aa (FNR) T-TGTQA
TCacA-aa
(CRP)
•
Aerobic
D
Anaerobic
169
EB Anaerobic+glu
e
pSSF16/180galE
pSSF16/M182recA
Fig, 4, Beta-galactosidase expression in wild-type S. enterica (180ga/£) and E. co//(M182), containing pSSF16. This plasmid contains the S. enterica ansB 5' region fused with iacZ (see text and Fig. 2). The strains were grown aerobically in the absence of glucose and anaerobically, both in the presence and absence of 0.4% glucose. The values are expressed in units mg protein"' (see Miller, 1972), and are the result of at least four independent experiments.
Fig, 5 A. Beta-galactosidase expression in wild-type and mutant strains of E. co//containing a site-directed mutant version of pSSF16 (text and Fig. 5B). M^e2recA was grown anaerobically and aerobically. All other strains were grown anaerobioally. pSSFI 6/M1 e2recA crp fnr and pSSF16Acf/crp18/M182rec/^ (aerobic) were not done. The values are expressed in units mg protein"' (see Miller, 1972), and are the result of at least four independent experiments. B. The ansB'-'/acZfusion plasmid, pSSF16, and deletion and sitedirected mutants of the ansS promoter region. The transcriptional start point is indicated by an asterisk. The -10 consensus sequence is underlined. Regions homologous to FNR and/or CRP DNA binding sites are indicated by boxes above the sequence; matches with the ansB sequence are in upper case and mismatches in lower case. These sites are the CF site (upstream) and the CRP site (downstream). The nucieotide sequence of deletions and site-directed mutants are shown under the sequence of pSSF16. Site-directed changes are indicated by underlined characters.
pSSF16cls18
CRP CONSENSUS TT-tat.Ga TCACA-Tt GAGATAATGCGACCA -10
pSSF16
GGTTAATCGTATGGCGTCACATTATTCGTCTGCAATATAGAGATAATGCGACCA
pSSFl 6Ac f
GGTTAATCGTATGGCGTCAIATTATTCGTCTGCAATATAGAGATAATGCGACCA
pSSF16Ac£/crpl8
TCTGTTTTTTCCTGCAnTTGTTATCCATCTCTAAAAAATACTCTCTGTCGGTTATATATCGCTGGTTAATCGTATGGCGTCAMTTATTCGTCTGCAATATAGAGATAATGCGACCA
pSSF16dsl8
170
M. P. Jennings, S. P. Scott and I. R. Beacham
The mechanism of oxygen regulation of ansB in S. enterica remains to be determined; one possibility is that it may be due to changes in DNA supercoiling (Higgins ef al., 1990). Such changes may affect the structure of the cAMP-CRP-induced nucleoprotein complex that is required for transcription. Another consideration may be variations in cAMP concentrations during anaerobic growth (Unden and Duchene, 1987). These changes, however, are quite small in magnitude and could not fully account for the observed regulation. The precise role of the L-asparaginase II in S. enterica and E. co//is not known. It has been suggested, however, that this enzyme may be involved in the utilization of Lasparagine to provide L-fumarate as a terminal electron acceptor during anaerobic respiration. In E. co/; all of the enzymes involved in this pathway are regulated by FNR (Jennings and Beacham, 1993). The anaerobic regulation of ansB expression in S. enterica is also consistent with the proposed role for L-asparaginase II as a source of an alternative electron acceptor. The strong similarity between known CRP and FNR binding sequences suggests that one, or a few, base-pair changes, during evolution, could, in principle, switch the regulation of a gene from FNR to CRP dependence and vice versa. The data presented here suggest that this has actually occurred in the case of the ansB promoter, some time after the divergence of E. coli and S. enterica.
Experimental procedures Bacteriai strains and piasmids The bacterial strains and plasmids are listed in Table 1.
Cuiture conditions Cells were grown in Luria-Bertani (LB) medium (Sambrook et al., 1989); ampicillin (lOOngmr') was included for plasmidcontaining strains.
Table 1. Strains and plasmids.
Strain/ Piasmid E CO//strain DH5a M182 Ml 82 rec/4 M^82fnr W82crp recA M^82crpfnr
Relevant genotype
Source/ Reference
&lacU169 ($80 /acZAM15), Sambrook eta/. (1989) recA1 AUaclPOZY) X74, galK, gaiU,Casadaban and Cohen strA (1980) Mac, recA56, sri-300:Jn10 From Ml82 by transduction Alac, fnrwTniO Belief a/. (1989) Mac, crpA39, recA56, Jennings and Beacham srl-300 ::Tn 10 (1993) Mac, crpA39 fnr.Jn 10 Beil era/. (1989)
Salmonelia strain SGSC180 ara9
Whitefield and Levine (1973) From 180 by transduction Strauchefa/. (1985)
SGSC180ga/£
ara9 galE503 bio203:Jn10
TN2062
PP1037
pepT7::Mudi(Ap"/ac cts62 X) leuBCD485 oxrA 1 zda-888:Jn10 crp773:.Jn10trpB223 D. McPhee
Piasmid pUC118
bla
Vieira and Messing (1987) pSK Biuescript bla Stratagene pNM482 bla Minton(1984) pMJS6 bla, ansB This work pSSIO bla, AansB This work pSSF16 bla, ansB'-'lacZ This work pSSF16Acf pSSF16 containing a deietion This work of sequence upstream of nucieotide position -53, resuiting in the deletion of the CF site. pSSF16Acf/crp18 pSSF16Acf containing a This work C—>T mutation at nucieotide position -34 pSSF16ds18 pSSFI 6 containing a C->A This work mutation at nucieotide position -34 pMJII bla, AansB Jennings and Beacham (1990)
Enzyme assays Beta-galactosidase was assayed in sonic extracts (Miller, 1972). The same extracts were also assayed for p-lactamase (Jones et al., 1982) in order to correct for variation in plasmid copy number (Jennings and Beacham, 1993). Protein concentration was measured using the BCA protein assay reagent kit (Pierce).
Assay of high affinity L-asparaginase activity by direct Nessierisation L-asparaginase II activity was assayed using sonic extracts by measuring the production of ammonia with Nessler's reagent (Wriston and Yellin, 1973). The substrate, L-asparagine, was used at a concentration of 0.1 mM at which L-asparaginase I
activity is not significant (Cedar and Schwartz, 1967). Cells were grown anaerobically in sealed 10 ml centrifuge tubes or aerobically with vigorous shaking to mid-log phase in LB medium. The amount of ammonia released was estimated from a standard curve of ammonia concentration. The protein concentration of extracts was estimated using the BCA protein assay reagent kit (Pierce).
Transcript mapping Transcript mapping was carried out using primer extension analysis (Jennings and Beacham, 1990) and S1 nuclease protection (Sambrook et al., 1989). Total RNA was isolated from cells containing pSSF16 according to Aiba et al. (1981). Probes were generated using the Ml 3 -40 primer, which
Salmonella enterica ansB gene regulation anneals to /acZ sequence in pSSF16. The extended primer was digested with BamHl for primer extension experiments and EcoRI for SI nuclease experiments.
DNA amplification and nucieotide sequence analysis Enzymatic amplification of DNA was carried out essentially as described by Saiki et al. (1988), using an annealing temperature of 55°C and a total of 35 cycles. Nucieotide sequencing was by the dideoxy termination method, using supercoiled plasmid templates and T7 polymerase (Sambrook etal., 1989; Limand Pene, 1988).
Molecular cloning of the 5' region of the ansB gene of S. enterica, and construction of an ansB'-'lacZ fusion Southern hybridization of S. enterica DNA, using a 2.3 kb Pst I fragment of the E. coli ansB gene from pMJ11 (Jennings and Beacham, 1990) as a probe revealed that the S. enterica gene was contained within a 3.0 kb EcoRI-/V/ndlll fragment. DNA was digested with these enzymes, electrophoresed on an agarose gel, and DNA in the size range of approximately 2.5 kb to 3.5 kb excised, purified, and cloned into pUC118 digested with EcoRI and HindW. From the resulting recombinant plasmid, pMJS6, identified by colony screening using the same probe, a Pvu\-Hind\\\ fragment was subcloned into Smal- and H/ndlll-cleaved pSK Biuescript (Stratagene) to give pSS10. The nucieotide sequence was determined on one strand using the SK and T7 primers. Using this sequence, two oiigonucieotides, containing EcoRI and BamH\ restriction sites (underlined), were synthesized: 'ssmet', 5'-CGGATCCATGTTATATCTCCAG-3'; and 'ssup', 5'-GCGAATTCTG I I I I I ICCTGCA-3'. They encompass the ansB 5' region, including the first (methionine) codon of the signal peptide (see Fig. 3), and were used to amplify a DNA fragment which was cloned into the EcoRI and BamHl sites of the /acZ fusion vector pNM482, to give pSSF16 (see Fig. 3). The sequence of the inserted DNA was determined to ensure the absence of errors introduced during DNA amplification.
Deletion and site-directed mutagenesis of the ansB promoter Deletions and site-directed mutations, within pSSF16, were made using the polymerase chain reaction (PCR) and pSSF16 as template. The degenerate oligonucleotide 'ssdncrp' (5'GCGAATTCGGTTAATCGTATGGCGTCAC/TATTATTCGT-3': EcoRI site underlined) and the M l 3 - 4 0 primer were used to generate a PCR product which incorporated a deletion in the sequence and/or a mutagenic change in the original ansB'-'lacZ construct (pSSF16). The amplified DNA was cloned into the EcoRI and BamH\ sites of the /acZfusion vector pNM482, to give pSSF16Acf and pSSF16Acf/crp18. Both constructs have the sequence upstream of nucieotide position -53 deleted, but pSSF16Acf/crp18 also contains a site-directed change (C->T) at nucieotide position -34. pSSF16ds18 was derived using two oiigonucieotides, 'ssup' (see above in this section) and 'St^Ec': 5'-CGGATCCCATGTTATATCTCCAGTTATGTCAACTGGTCGCATTATCTCTATATTGCAGCGAATAATTTGAC-3' (SamHI site underlined). A PCR product was
171
generated which incorporated a mutagenic change at nucieotide position -34 (C—»A) in the promoter region. The amplified DNA was cloned into the EcoRI and SamHI sites of the /acZfusion vector pNM482. For each cloned PCR product the nucieotide sequence was determined in order to confirm the desired change and to ensure the absence of errors introduced during DNA amplification.
Acknowledgements These studies were supported by the Australian Research Council. S.P.S. is the recipient of an Australian Postgraduate Research Award. We thank Dr S. Busby for generously supplying strains.
References Aiba, H., Adhya, S., and de Crombrugghe, B. (1981) Evidence for two functional gal promotors in intact Escherichia coli cells. J Biol Chem 256:11 905-11 910. Beacham, I.R., and Jennings, M.P. (1990) L-Asparaginase. Today's Life Science 2: 40-43. Bell, A.I., Gaston, K.L, Cole, J.A., and Busby, S.J.W. (1989) Cloning of binding sequences for the Escherichia coli transcription activators, FNR and CRP: location of bases involved in discrimination between FNR and CRP. NucI Acids ResM: 3865-3874. Casadaban, M.J., and Cohen, S.N. (1980) Analysis of gene signals by fusion and cloning in Escherichia coli. J Mol Biol 138:179-207. Cedar, H., and Schwartz, J.H. (1967) Localisation of two Lasparaginases in anaerobically grown Escherichia coli. J Biol Chem 242: 3753-3755. Cedar, H., and Schwartz, J.H. (1968) Production of L-asparaginase II by Escherichia coli. J Bacteriol 96: 2043-2048. Chesney, R.H. (1983) E. co//L-asparaginase II production in the presence and absence of catabolite activating protein. FEMS Microbiol Lett A7: 161-162. Gallagher, M.P., Marshall, R.D., and Wilston, R. (1989) Asparaginase as a drug for treatment of acute lyrfiphoblastic leukaemia. Essays Biochem 24:1-40. Higgins, C.F., Dorman, C.J., and N.I. Bhriain, N. (1990) Environmental influences on DNA supercoiling: a novel mechanism for the regulation of gene expression. In The Bacterial Chromosome. Driica, K., and Riley, M. (eds). Washington: American Society for Microbiology, pp. 421-431. Jamieson, D.J., and Higgins, C F . (1984) Anaerobic and leucine-dependent expression of a peptide transport gene in Salmonella typhimurium. J Bacteriol AGO: 131-136. Jennings, M.P., and Beacham, I.R. (1990) Analysis of the Escherichia coli gene encoding L-asparaginase II, ansB, and its regulation by cyclic AMP receptor and FNR proteins. J eacter/o/172:1491-1498. Jennings, M.P., and Beacham, I.R. (1993) Co-dependent positive regulation of the ansS promoter of Escherichia coli by CRP and the FNR protein: a molecular analysis. Mol Microbiol 9:155-^
64.
Jerlstrom, P.G., Lui, J., and Beacham, I.R. (1987) Regulation of Escherichia co//L-asparaginase II and L-aspartase by the ^nrgene product. FEMS Microbiol Lett 4A: 127-130. Jones, R.N., Novick, W.W., Barry, A.I., and Thornsberry, C. (1982) In vitro evaluation of CENTA, a new p-lactamasesusceptible chromogenic cephalosporin reagent. J Clin /W/crob/o/15: 954-958.
172
M. P. Jennings, S. P. Scott and I. R. Beacham
Lim, H.M., and Pene, J.J. (1988) Optimal conditions for supercoiled DNA sequencing with the Escherichia coli DNA poiymerase iarge fragment. Gene Anal Tech 5: 32-39. IVIiiler, J.H. (ed.) (1972) Experiments in Molecular Genetics. Coid Spring Harbor, New York: Coid Spring Harbor Laboratory Press, pp. 403-404. Minton, N.P. (1984) Improved plasmid vectors for the isoiation of transiational lac gene fusions. Gene 31: 269-273. Raibaud, O., and Schwartz, M. (1984) Positive controi of transcriptional initiation in bacteria. Annu Rev Genet 18: 173-206. Russeil, L., and Yamazai
