A Coxiella burnetii gene encodes a sensor-like protein

Gene, 151(1994)185-190 0 1994 Elsevier Science B.V. All rights reserved.

185

0378-l 119/94/$07.00

GENE 08404

A Coxiella burnetii gene encodes a sensor-like protein (Two-component

regulatory system; signal transduction; PCR; obligate intracellular bacterium; Rickettsia)

Yin-yuan MO and Louis P. Mallavia Department of Microbiology, Washington State University, Pullman, Received by R.E. Yasbin: 1 July 1994; Revised/Accepted:

WA 99164-4233, USA

10 August/l

1 August

1994; Received at publishers:

5 September

1994

SUMMARY

Two-component regulatory systems play important roles in the adaptive responses of many bacteria to environmental changes. The sensor proteins of these systems are highly conserved near their C-termini. We exploited this feature to isolate a gene encoding a putative sensor component from the obligate intracellular rickettsial parasite Coxiella burnetii (Cb). Using degenerate primers and the polymerase chain reaction (PCR), we isolated a DNA fragment from a genomic library of Cb containing an open reading frame (ORF), sufficient to encode a 48-kDa protein. Sequence comparison revealed that the deduced protein shared high homology to members of the bacterial sensor protein family, particularly at three conserved regions of the C terminus. When the Cb sensor-like gene was cloned into a high-copy-number vector and introduced into an E. coli strain (phoM, phoR), the mutant expressed low levels of alkaline phosphatase activity, suggesting that the gene functioned as a sensor protein in E. coli. Consequently, we designated this gene qrsA (for Q fever agent regulatory sensor-like gene). Because two-component regulatory systems have been implicated in a variety of cellular processes, including virulence determinants in some pathogenic bacteria, the identification of qrsA in Cb may shed light on how the pathogen adapts to extracellular changes during infection, as it proliferates in the phagolysosome.

INTRODUCTION

Coxiella burnetii (Cb) is an obligate intracellular bacterium which causes Q fever in humans (Baca and Paretsky, 1983). A unique feature of this organism is its ability to proliferate in the phagolysosome, an environCorrespondence Washington

to: Dr. L.P. Mallavia,

State

University,

Department

Pullman,

of Microbiology,

WA 99164-4233,

USA.

Tel.

(l-509) 335-3323; Fax (l-509) 335-3517; e-mail: [email protected] Abbreviations: aa, amino acid(s); ACP, acid phosphatase(s); acp, gene(s) encoding ACP; bp, base pair(s); C., Coxiella; Cb, C. burn&i; E., Escherichia; GCG, Genetics Computer Group (Madison, WI, USA); IVTT, in vitro transcription/translation; kb, kilobase or 1000 bp; nt, nucleotide(s); ORF, open reading frame; PAGE, polyacrylamide-gel electrophoresis; PCR, polymerase chain reaction(s); P, promoter; QrsA, sensor-like protein; qrsA, gene encoding QrsA; RBS, ribosome-binding site(s); SDS, sodium dodecyl sulfate; chloro-3-indolyl-B-D-galactopyranoside; lyl phosphate. SSDI 0378-1119(94)00615-6

Xaa, any aa; XGal, 5-bromo-4XP, 5-bromo-4-chloro-3-indo-

ment that is usually lethal to other intracellular bacteria. Although several chromosomal and plasmid-carried genes have been characterized for Cb (Mallavia, 1991), little is known concerning virulence factors, requirements for intracellular growth or gene regulation. How Cb enters host cells and establishes intracellular infection is of particular interest to our laboratory. Previous studies have shown that during infection Cb produces an acid phosphatase (ACP) activity, which may act as a virulence factor (Baca et al., 1993). In Salmonella, expression of ACP-encoding genes is regulated by a pair of genes, phoQ and phoP (Groisman et al., 1989; Miller et al., 1989) that are homologous to a family of two component regulatory systems (Stock et al., 1989). Moreover, in S. typhimurium, phoQ and phoP have been implicated as virulence determinants enhancing survival in macrophages (Fields et al., 1989; Miller et al., 1989). Two-component regulatory systems have been identified in a large number of bacterial species and shown

186 to regulate a variety of cellular processes (Stock et al., 1989). The sensor protein family is defined by a section of conserved sequences consisting of three distinct regions of conservation, generally located near the C terminus. Within this section seven highly conserved aa have been identified: Asn in region II and Asp-Xaa-Gly-Xaa-GlyXaa,,_,,-Gly-Xaa-Gly in region III. In addition, all sensor proteins have a conserved His in region I, presumably the site of autophosphorylation. As the first step towards identifying genes controlling the expression of the acp genes in Cb, an attempt was made to isolate a phoQ analog. The conserved residues in region III of the sensor proteins made it possible to amplify that segment using degenerate primers corresponding to the highly conserved aa residues. The aim of this study was to clone, sequence and express a Ch gene coding for a sensor-like protein in E. coli.

EXPERIMENTAL AND DISCUSSION

(a) Cloning the PCR products PCR amplification using degenerate primers has been successfully used to clone a gene coding for an EnvZ (Comeau et al., 1985) analog from Staphylococcus aureus (Bayles, 1993). We employed the same approach to isolate a Cb sensor-like gene. A DNA segment to be amplified in this study was anticipated to be within region III of the family of sensor proteins that ranged from 33 to 61 aa. Within this region, 3 aa (Asp-Xaa-Gly-Xaa-Gly) at the N terminus and 2 aa (Gly-Xaa-Gly) at the C terminus are extremely conserved among different bacterial species (Stock et al., 1989). To reduce non-specific PCR products, various concentrations of components of the PCR were tested. These variations included different concentrations of primers (0.2, 1, 2 and 4 PM), Mg2+ (1, 1.5, 2, 2.5, 3, 3.5, 4 and 5 mM) and DNA template ( 10, 50, 100 and 200 ng/ml). We found that a concentration of 2 uM primer, 2 mM Mg2+ and 100 ng/ml of DNA template at an annealing temperature of 48°C gave the best result, i.e., under these conditions major PCR products were in the range of 100 to 300 bp in size. PCR products (100-300 bp) were isolated and cloned into the vector pCRI1. Restriction analysis of approx. 200 recombinant plasmids showed a random distribution of size from 50 to 300 bp. Among 45 clones that were sequenced, two clones carried the same DNA fragment ( 115 bp). These differed by two nt at a primer region due to the use of degenerate primers, suggesting that they were not merely a duplicated clone. One of the clones was named pRS33 (Fig. 1 and Table I). A search of the protein database revealed that this portion of the deduced aa sequence was highly homologous to a variety of pro-

karyotic sensor proteins. For example, it shared 71% aa similarity with CpxA (Weber and Silverman, 1988) and PhoR (Makino et al., 1989), and 61% with PhoQ (Miller et al., 1989). To confirm that the cloned DNA fragment was originally amplified from Cb, the DNA insert present in pRS33 was labeled with [a-32P]dCTP and used to probe Cb DNA. This fragment hybridized to Cb but not to E. coli DNA, suggesting its Cb origin (data not shown).

(b) Isolation of a clone carrying a sensor-like gene from the C. burnetii library To verify that the DNA insert of pRS33 encoded a portion of the sensor-like protein, the next step was to isolate the entire gene and flanking regions. Therefore, a Cb genomic library was constructed in hZAPI1 and plaque hybridization screening conducted using the [a-32P]dCTP-labeled 115-bp insert of pRS33 as a probe. Among approx. 2000 plaques, two showed a strong signal to pRS33. Both clones were in vivo excised and were indistinguishable by physical mapping. This clone, pRS33.C carried a 5.4-kb EcoRI fragment of Cb DNA. To localize the cloned DNA fragment of pRS33 on pRS33.C a restriction map was generated and Southern hybridization localized it within the XhoI-NheI region (Fig. 1). Subsequently, a subclone, pRS33.1 containing this region was used for DNA sequencing. To examine whether other isolates contain this gene, Southern hybridization was conducted with Cb genomic DNAs using the cloned fragment as a probe. We chose isolates Nine Mile, ‘S’, Priscilla and ‘K’ to represent various disease-causing agents and genetic backgrounds (Samuel et al., 1983; 1985). The Nine Mile isolate has only been associated with acute Q fever in humans and carries the QpHl plasmid; ‘S’ is a plasmidless isolate associated with chronic disease in humans; Priscilla was isolated from aborted goat tissues and ‘K’ has only been associated with chronic endocarditis in humans. Both ‘K’ and Priscilla appear to carry the same QpRS plasmid. When total DNA from other isolates of Cb were tested with this probe, all gave strong signals with the same restriction pattern for Clal and Hind111 (Fig. 2). In addition, when a fragment corresponding to the DNA insert of pRS33 was cloned from the isolate Priscilla, its nt sequence was identical to that of pRS33. These data suggest that the gene is conserved in those isolates, and might be involved in regulating a common pathway shared by all isolates. (c) Sequencing the C. burnetii sensor-like gene Sequencing revealed an ORF which could encode a protein of 425 aa (48 kDa; (Fig. 3). As shown by the

187 TABLE

1

Bacterial

strains

and plasmids

Bacteria

or plasmids

Source

Description”

Bacteria E. coli XLl-blue

or reference

t&Al, hsdR17(r;, ml). supE44, thi-1, recA, gyrA96, w/A, (lac-) [F’, proAB+ laclqZAM15,

Stratagene

TnlO(T$)]

endAl, hsdR17(r;,

INVF’

m:),

supE44, &i-l.

recA, gyrA96, relA, (b801acZAM15 A(lacZYA-argF),

deoR, F’

ompThsdS,(r;,

Invitrogen

me), go/, dcm (DE3)

Novagene

BL21(DE3)

F-,

ANCC22

phoR68, phoM

Makino

Acute disease isolate, plasmid QpHl Chronic disease isolate, plasmidless

Samuel et al. (1985) Samuel et al. (1985)

Goat

Samuel et al. (1985)

et al. (1989)

C. hurnetii (Cb) Nine Mile RSA493 ‘S’ Priscilla

4177

abortion

Chronic

‘K’

isolate, plasmid

QpRS

disease isolate. plasmids

Samuel et al. (1985)

QpRS

Plasmids ApR cloning

Stratagene

vector

PSK pCRI1

ApR, KmR, PCR cloning

PET-3a

ApR expression

PLYSS pRS33

CmR carrying

pRS33.C

A 5.4-kb EcoRI fragment

Invitrogen

vector

vector; P T7

Novagen Novagen

T7 lysozyme 115 bp of qrsA

PCR clone carrying

This study

carrying

qrsA in pSK fragment

carrying qrsA in pSK qrsA in pCRI1 under control

pRS33.X pRS33.Y ’ Ap, ampicillin;

This study

A 2.7-kb EcoRI-XhoI

pRS33.1

Same as pRS33.X Cm, chloramphenicol;

Km, kanamycin;

A, deletion;

This study of 1acP

except opposite Tn, transposon;

sequence analysis, potential RBS AAAGA, GAAAG and GGAAG were positioned 2, 8 and 19 nt upstream from the first ATG codon. Interestingly, the first three aa of this ORF were Met. All three Met appeared able to function as the start codons based on the location of the putative RBS (Shine and Dalgarno, 1975). No obvious - 35 and - 10 RNA polymerase-binding sites comparable to Rickettsia (Mallavia, 1991) or E. coli (Hawley and McClure, 1983) were observed. (d) Comparison of the aa sequences

A search of the protein database with the predicted aa sequence of the ORF using the FASTA program detected a number of sensor proteins. Because the ORF shared significant homology with sensor proteins and could have a similar function(s), it was tentatively named QrsA (for Q fever agent regulatory sensor-like protein). For example, QrsA shared overall aa identity of 25.1% (similarity of 50.5%) with CpxA (Weber and Silverman, 1988), 23.9% (similarity of 48.6%) with PhoR (Makino et al., 1989) and 21.7% (similarity of 47.2%) with PhoQ (Miller et al., 1989). The most homology was observed within

This study

orientation p, plasmid;

This study PT7, phage T7.

region III (37 aa). Analysis of the QrsA aa sequence revealed that like a majority of sensor proteins QrsA also has three highly conserved regions (Fig. 4). Thus, region I is characterized by a His, the putative site of phosphorylation, region II is marked by the presence of a conserved Asn and region III spans 27 aa flanked by conserved Asp-Xaa-Gly-Xaa-Gly and Gly-Xaa-Gly. Analysis of the topology of many sensor proteins predicts two membrane-spanning domains, a periplasmic domain containing the sensory receptor and a C-terminal cytoplasmic domain harboring the His protein kinase activity, as well as the autophosphorylated His. A hydropathy plot was performed on QrsA using the algorithms of Kyte and Doolittle (1982), with CpxA as a reference. The resulting hydropathy plot of deduced QrsA sequence was very similar to that of CpxA (not shown). (e) Expression of qm4 in E. coli

To verify this ORF, in vitro transcription and translation (IVTT) was carried out with pRS33.1 using an IVTT kit (Promega). As shown in Fig. 5, a unique protein band

188 R Pc DR333.C

’

N L

”

pRS33.1

xI

qrsA

PI R ,

1 2 3

4

5

6

7

8

9

10

M

r 200 bP

pRS33.X

-23

PAS33

Fig. 1. Restriction translation

is a PCR

clone

obtained

by using

Abbreviations

the degenerate

S-GARGAYRAYGGNCCNGG

5’-AYNGCNARNCCNACNCC,

and

where N=A

or G and Y =C or T (DNA

International,

After separation

in 2% agarose

isolated,

using GeneClean

purified

strand,

or C or G or T, R=A

gels, PCR products

OR, USA).

(100-300

bp) were

II (Bio 101, La Jolla, CA. USA) and

on XGal plates were grown and plasmids

white, transformant extracted

colonies

primer as appropriate (Yanisch-Perron was isolated using [c(-32P]dCTP-labeled screening.

EcoRI-digested packaged library

et al., 1985). Plasmid pRS33.C pRS33 as a probe for plaque

A Cb (Nine Mile phase I) genomic

by ligating

EcoRI

partially

h arms (ilZAPI1,

into h heads using Gigapack was grown

membrane particles

at 37C.

in alkaline

Plus (Stratagene).

Plaques

(Schleicher&Schuell, lysed

Keene,

solution

library

genomic

DNA

was to

La Jolla, CA, USA) and

on E. coli XLl-blue

overnight

top layer of soft agar

digested

Stratagene,

The hZAPI1

on LB plates with a

were transferred NH,

USA)

to a Nytran

and

(0.5 M NaOH/1.5

the phage

M NaCl)

as

previously described (Sambrook et al., 1989). Viral DNAs from positive plaques were excised in vivo according to the supplier’s instructions and confirmed by Southern hybridization (Southern, 1975). Construction of pRS33.X is as follows: a 1310-bp DNA fragment encompassing was amplified expression shown

the potential by PCR

and

RBS and translation cloned

of qrsA is under control

here) carried

into

-0.5

by the ‘easy prepara-

tion’ method (Berghammer and Auer, 1993), which allowed direct determination of the nt sequence of inserts using the M 13 universal or reverse

hybridization

-2.2 -2.0

primers:

antisense

Lake Oswego,

cloned into the vector pCR II. Individual

constructed

-9.6 -6.7 -4.3

The predicted

C, ClaI; N, NheI; P, PstI; R, EcoRI; X, XhoI.

enzymes:

sense strand,

and its subclones.

of qrsA is shown by an arrowhead.

direction

for restriction pRS33

map of pRS33.C

pCRI1

terminator in such

of QrsA a way that

Fig. 2. Southern

analysis

of Ch DNA using the [cc-32P]dCTP-labeled

DNA insert of pRS33 as a probe. The probe was prepared pRS33 with EcoRI, separated

in agarose

by digesting

gels. purified using GeneClean

II and labeled with [rx-3ZP]dCTP using Prime It II. Prehybridization was carried out in 6 x SSC (1 x SSC is 150 mM NaCI/lS mM Na,citrate,

pH 7.6), 2 x Denhardt’s

(US Biochemical,

Cleveland,

OH.

USA)/O.l% SDS/100 ug per ml sheared salmon sperm DNA for 4 h at 65°C. Hybridization was done in the same solution with the added probe,

under

the same conditions

for 16 h. Membranes

twice for 30 min each in 2 x SSC/O.l%

twice for 30 min each in 0.1 x SSC/O.l% DNAs were digested

with restriction

were washed

SDS at room temperature,

and

SDS at 68 C. Ch genomic

endonucleases

and loaded

onto a

0.7% agarose gel at 2 ug DNA/lane. Lane 1, E. coli DNA digested with HindHI; lanes 2-5 and 669 are CluI and HindIII-digested DNAs. Lanes: 2 and 6, Nine Mile; 3 and 7, ‘S’; 4 and 8, Priscilla: 5 and 9, ‘K’. Lane 10 contains M, h/Hind111

pRS33.C molecular

DNA digested mass markers

with EroRI indicated

(5 ng.!lane) and lane in kb.

of lacP on the vector. pRS33.Y (not

the same insert but oriented

opposite

to the lacP.

of approx. 48 kDa was produced by pRS33.1, but not by pSK. In addition, we made attempts to overexpress the qrsA in pET3-a (Novagen, Madison, WI, USA). Unexpectedly, the amount of QrsA using the T7 expression vector was very low. We detected a protein band with the predicted size when labeled with [35S]Met, but not in Coomassie-brilliant-blue-stained gels (data not shown). Nevertheless, these results suggested that the size of the produced protein is in good agreement with that of QrsA predicted from the sequence. Why qrsA cannot be overexpressed by the PET system is not clear. In general, low levels of expression are attributed to toxicity of the cloned gene or instability of the recombinant plasmid (Studier et al., 1990). In the case of QrsA, the E. coli cells carrying qrsA grew normally, and introduction of pLysS did not improve the expression. Further, counting viable cells on antibiotic plates indicated that the plasmid was stably maintained. In addition, nt sequences of the fused DNA fragment in the recombinant plasmid were confirmed by sequencing; Northern blotting detected a fair amount of qrsA mRNA.

Taken together, the low protein levels observed appears not to result from the above-mentioned factors. Previous studies found a distinguishable codon bias in the htpAB operon of Cb (Vodkin and Williams, 1988) when compared to the codon usage of highly expressed proteins of E. coli. A similar bias was evident in the codon usage of qrsA. For instance, the use of GGA (Gly), CGG (Arg) and CCC (Pro) by E. coli is extremely rare, but these codons were commonly used in qrsA at frequencies of 0.29, 0.19 and 0.50, respectively, suggesting that low expression levels may reflect a limited availability of certain tRNA species. Whether this is the real explanation, however, remains to be determined. To investigate the function of QrsA, we did a complementation analysis for qrsA in E. coli based on the fact that similar sensor proteins can cross-talk to some degree (Ninfa et al., 1988; Aiba et al., 1993). Strain ANCC22 exhibits a phenotype of PhoA- due to mutations in the phoR and phoM genes, both of which encode a sensor protein required for phosphorylation of the response regulator (PhoB) (Makino et al., 1989). When pRS33.X (carrying qrsA) was introduced into ANCC22, a pale blue color was observed on XP indicator plates which were

189 .

.

.

.

.

.

.

.

.

.

.

.

TATTCAAATTAGCCGGCTGCGGCACA~GTT~~~GATCCT~TCCCC~TCATC~CCATCCGTGGCGGT~TATGTGTTAGCGG~TTGTTC~~CT

120

MMHCIVDKFARHVSGTSAFAILCVIIVAHLITMTFYINDS ATGATGATGTGCATTGTCGATAAATTTT~GCCACGTCACGTCAGTG~ACTAGC~TTTTGCCATTTTATGCGTCATTATTGTTGC~CCTTAT~CCATGACGTTTTATATT~CGATAGC

40 240

RHARRAADRDAMIQKIINVINLLEATPVEERAKAIAAMND CGTCATGCACGACGCGCAGCCGATCGCGACGCGATGATT CAAAAAATTATTAACGTGATTAATTTTAGAAGCGACGCT

80 360

PDMKASISSKPQFALQFQDISFWQINGALRKQLGAFAVSI CCCGATATGAAAGCGTCCATCTCAAGCAAACCCCAATTTGTT

120 480

LLTKDQWLNINATLYTHFLLRQLMFFVFEIVAFGFILILI CTGTTAACCAAAGATCAATGT~TATCAATGCCACTTTATACACTCATTTTCTTTT~GACAGTT~TGTTCTTTGTATTC~TCGTC~TTTT~TTTTATTCT~TTTT~TC

160 600

WSINRFTRPIENFRRAAERLSMDLNTQPVPIDGPSVVQEA TGGTCCATTAATCGCTTTACCCGTCCGATTGAAAATTTTTCGGC~GC~CC~GCGATT~GCAT~ATTT~TACCC~CCTGTGCCGATTGAT~TCCGTC~TGGTGC~G~~C

200 720

ARAMNRMQQRIQDLIRNRTQMLAAISHDLRTPITRMKLRA GCAAGGGCGATGAATCGGATGCAGCAGCGGATTC~GATTT~TTCG~TCGCAC~~T~TAGC~GATTTC~ATGATTT~~C~CCATCACCCG~TG~TTACGT~T

240 840

QFLDNSTTRNALVKDLNEMEVMINETLSFARDDFADNAKV CAATTTCTCGATAATTCCACGACGCGCAATGCGTTffiT~GATTT~T~T~AGGTTATGATT~T~GACGTTATCTTTTGCTCG~ATGATTTTGCT~T~TGC~GTT

280 960

NLDLVSLVCSLVESMQDMGYNIQFHSHPQRIKILGRASAL AACCTTGATTTAGTCTCCCTCGTTTGTTCGTTGGTTG~TC~T~A~ACATGGGATAC~CATTC~TTC~CA~CATCCC~~~ATT~TATT~CCGAGCCAGCGCATTA

320 1080

KRAFTNLLNNAIRYAKNVNVRIQWRQNRVKVLIEDDGPGI AAACGAGCGTTTACTAATTTATTAAATAATGCTATTCGTTAT~G~~TGT~CGTTCGTATTC~TGGC~CA~TCGAGTG~GGTTCTGATT~

360 1200

AEKELEQVFEPFYRGEHSRSRDTGGVGLGLAVRRDIISDH TAGGCGGGATATTATTTCTGACCAC NG KV I L T N R P N GG L CA T V E L L S EV H * 425 AATGGAAAAGTAATTTTAACGAATCGACCCAATGGTGGTTTGT~GC~CGGT~~TTACTTT~GA~TT~TTAGAGCTGTTATG~T~TGTTATTCGTTCTATT~~TTTT~G GTTGGGCGCTCTCGGCGATATTTGTCTTACCGTTCWLTTGCTC 1508

Fig. 3. Nucleotide product

(pRS33)

by dideoxy needed.

chain-termination

A total of 1508-bp

for the primer

regions.

the chromosome. VADMS laboratory databases

of qrsA and deduced

sequence corresponding

aa sequence.

to nt 1179-1294 (Sanger

et al., 1977) using Sequenase

DNA were sequenced

on both strands.

There were 8 nt in the sense primer

The nt sequence of Washington

under accession

Potential

and deduced

.

by italicized

library

data were analyzed

The sequence data presented

.

.

.

by an underline.

Customized

sequencing

of the insert in pRS33 was identical

using CCC

primer

region different

Sequence

1440

letters with an underline.

is indicated

Version 2.0 (US Biochemical). The nt sequence

region and 3 nt in the antisense

aa sequence

State University.

No. UO7186.

RBS are noted

that was used to screen the genomic

Analysis

primers

(Devereux

.

PCR

was carried

were used when

to that of pRS33.1

in this study will appear in the EMBL, GenBank

.

The original

Sequencing

from the original

Package

400 1320

except

nt sequence

of

et al., 1984) at

and DDBJ sequence

12M

RLQLGTALLRRRS 269 CpxA PELEAGPQEFLAAGASFNQMVTALERMNTSQQRLLSB I II II I Ill IIII II I QrsA PVPIDGPSVVQE'AAUMNFMQQRIQDLIRNRTQML ~LRAQFL.... 243 I I I II PhoR VNPYTHKQLLMVA.....RDVTQMHQLEGARRNFF ~VLQGYLEMMNEQP 233

CpxA GES....KELERIETEAQRLDSMINDLLVMSRNQQKNALVSETIKANE 325 I III I I II QrsA DNS....TTRNALVKDLNEMEVMINETLSFARDD.... .FADNAKVNLDLVSLVCSLVES 294 I II I PhoR LEGAVREKALHTMREQTQRGLVKQLLTLSKXEAAPTHLLNEKVDVPMNLRWEREAQT 293

CpxA AEQMGRSLTVNFPPGPNPLYGNLRYSH..TKIEVGFAVDKDGITI II I I I II II QrsA MQDMGYNIQFHSHPQRIKILG RA&&&um IRYA...KNVNVRIQWRQNRVKV II I PhoR LSQKKQTFTFEIDNG.LKVSGN ~~HTPSGTHIT~W~RVPHGASF II CpxA T~EDREQIFRPFYRTDEARDRES II I IllI IIIII cmn BKELSQVFEPFYRGEHSF%T II III II I III PhoR SmEHIPRLTERFYRVDKARSRQTIII

Fig. 4

III

~STAIQQHRGWVKAEDSPL II lllll I III I ~~IISDHNGKVILTNRPN III IIIII I KHAVNHHESRLNIESTVG

383 351

-29

352

443 411 412

III

-18 Fig. 5

Fig. 4. An alignment of the deduced aa sequence at the C terminus of QrsA with that of CpxA and PhoR. The conserved regions are underlined. The most conserved aa His in region I, the presumed site of autophosphorylation, Asn in region II and Asp, two Gly at N terminus of region III, two Gly at the C terminus of region III are indicated in bold letters. Fig. 5. Synthesis of QrsA by IVTT using an E. coli extract. DNA (2 pg) of pSK or pRS33.1 was individually mixed with 10 ~1 of premix minus Met, 7.5 ~1 of S30 extract (Promega, Madison, WI, USA) and 1 ~1 of C3’S]Met (1200 Ci/mmol) in a final volume of 25 ~1. The mixture was incubated for 30 min at 37”C, precipitated with four volumes of acetone and finally dissolved in 100 ~1 of 1 x Laemmli buffer (Laemmli, 1970). The labeled proteins were separated by 0.1% SDS-12.5% PAGE. The gel was dried and autoradiographed using X-OMAT film (Eastman Kodak, Rochester, NY, USA). Molecular masses of protein standards (in kDa) are indicated on the right of the gel. The arrow indicates the band corresponding to QrsA. Lanes: 1, pSK; 2, pRS33.1; M, protein markers.

190 set at 4°C for l-2 days following overnight incubation at 37°C suggesting partial complementation.

structure

and regulation

teins. J. Bacterial. Devereux,

J., Haeberli,

sequence

of synthesis

of the OmpR

and EnvZ pro-

164 (1985) 5788584.

analysis

P. and Smithies, programs

0.:

A comprehensive

for the VAX. Nucleic

set of

Acids Res. 12

(1984) 3877395.

(f ) Conclusions

(I) Evidence is provided for the isolation of a gene coding for a sensor-like protein in Cb. This is, to our knowledge, the first report of a gene encoding a sensor protein from any member of the order Rickettsiales. The identification of QrsA in Cb supports the view that sensor proteins exist universally in prokaryotic organisms. (2) QrsA shares significant aa homology with wellcharacterized sensor proteins such as CpxA, PhoR and PhoQ, especially at three conserved regions. A hydropathy plot of QrsA reveals similarities to many sensor proteins, particularly CpxA, (3) The qrsA gene cloned in a high-copy-number vector is able to partially complement phoR and phoM mutations in E. coli. (4) IVTT identifies a protein predicted to be Cb QrsA. (5) The qrsA gene appears to be conserved among different isolates of Cb.

Fields,

PI., Groisman,

Hawley,

D.K.

and

McClure,

W.R.:

Escherichia co/i promoter (1983) 223772255. Kyte,

J. and

Doolittle,

hydropathic Laemmli,

DNA

cells.

R.F.:

A simple

of a protein.

U.K.: Cleavage

of structural

K., Shinagawa,

regulator.

Compilation

sequences.

character

the head of bacteriophage Makino,

locus that

from phagocytic

murium phoP virulence gene is a transcriptional Nat!. Acad. Sci. USA 86 ( 1989) 707777081.

and

Nucleic

method

Proc.

analysis

of

Acids Res. 11

for displaying

the

J. Mol. Biol. 157 (1982) 1055132. proteins

T4. Nature

during

the assembly

of

227 (1970) 680-685.

H., Amemura,

M., Kawamoto,

T., Yamada,

M. and Nakata, A.: Signal transduction in the phosphate regulon of Escherichia coli involves phosphotransfer between PhoR and PhoB proteins.

J. Mol. Biol. 210 (1989) 551-559.

Mallavia, L.P.: Genetics 2133221. Miller, S.I., Kukral,

of Rickettsiae.

Eur. J. Epidemiol.

A.M. and Mekalanos,

7 (1991)

J.J.: A two-component

regu-

latory system (phoP phoQ) controls Salmonella typhimurium virulence. Proc. Nat!. Acad. Sci. USA 86 (1989) 505445058. A.J., Ninfa,

E.G., Lupas,

Stock, J.: Crosstalk tion proteins Evidence

between

A.N., Stock, bacterial

and the regulator

that nitrogen

by a common

We thank Shirley Schmitt for her excellent technical assistance, Raymond A. Larsen for critical review of this manuscript and James E. Samuel for the hZAPI1 Cb genebank. The software cited is part of the VADMS Center, a campus-wide computer resource at Washington State University. This work was supported by grant AI20190 from the National Institutes of Health (NIAID).

F.: A Salmonella

proteins

E.A., Chiao, E., Lipps, C.J. and Heffron, F.: Sdmone~~a typhi-

Groisman,

Ninfa,

ACKNOWLEDGEMENTS

E.A. and Heffron,

controls resistance to microbiocidal Science 243 (1989) 1059-1062.

A., Magasanik,

chemotaxis

of transcription

assimilation

phosphotransfer

of the Ntr regulon:

and chemotaxis

mechanism.

B. and

signal transducare controlled

Proc. Nat!. Acad. Sci.

USA 85 (1988) 5492-5496. Sambrook. J., Fritsch, E.F. and Maniatis, T: Molecular Cloning. A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press. Cold Spring Harbor, NY, 1989. Samuel, J.E., Frazier, M.E., Kahn,

M.L.,

Thomashow,

L.S.

Mallavia, L.P.: Isolation and characterization of a plasmid phase I Coxiella burnetii. Infect, Immun. 41 (1983) 4888493. Samuel,

J.E., Frazier,

M.E. and Mallavia,

type and disease

caused

by Coxiella

L.P.: Correlation burnetii.

Infect.

and from

of plasmid Immun.

49

(1985) 7755779. Sanger,

F., Nicklen,

terminating

S. and Coulson,

inhibitors.

Proc.

A.R.: DNA sequencing Nat!.

Acad.

with chain-

Sci. USA

74 (1977)

5463-5467. REFERENCES

Shine, J. and Dalgarno,

Aiba, H., Nagaya, M. and Mizuno, T.: Sensor and regulator proteins from the cyanobacterium Synechococcus species PCC7942 that belong to the bacterial

signal-transduction

protein

families: implica-

tion in the adaptive response to phosphate limitation. Mol. Microbial. 8 (1993) 81-91. Baca, O.G. and Paretsky, D.: Q-fever and Coxiella burnetii: a model for host-parasite interactions. Microbial. Rev. 47 (1983) 127-149. Baca, O.G., Roman, M.J., Glew, R.H., Christner, R.F., Buhler, J.E. and Aragon. AS.: Acid phosphatase activity in Coxiella burnetii: a possible virulence factor. Infect. Immun. 61 (1993) 4232-4239 Bayles, K.W.: The use of degenerate, sensor gene-specific, oligodeoxyribonucleotide primers to amplify DNA fragments from Staphylococcus aureus. Gene 123 (1993) 99-103. Berghammer, H. and Auer, B.: “Easypreps”: fast and easy plasmid minipreparation for analysis of recombinant clones in E. coli. BioTechniques 14 (1993) 524-528. Comeau, D.E., Ikenaka, K., Tsung, K. and Inouye, M.: Primary characterization of the protein products of the Escherichia coli ompB locus:

L.: Determinant

rial ribosomes. Nature Southern, E.M.: Detection

of cistron

254 (1975) 34-38. of specific sequences

specificity

among

in bacte-

DNA fragments

separated by gel electrophoresis. J. Mol. Biol. 98 (1975) 5033517. Stock, J.B., Ninfa, A.J. and Stock, A.M.: Protein phosphorylation and regulation

of adaptive

responses

in bacteria.

Microbial.

Rev. 53

(1989) 450-490. Studier F.W., Rosenberg, A.H., Dunn, J.J. and Dubendorff, J.W.: Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185 (1990) 60-89. Vodkin, M.H. and Williams, J.C.: A heat shock operon in Coxiellu burnetii produces a major antigen mycobacteria and Escherichia

homologous to a protein in both coli. J. Bacterial. 170 (1988)

122771234. Weber, R.F. and Silverman, P.M.: The Cpx proteins of Escherichia cofi K 12: structure of the CpxA polypeptide as an inner membrane component. J. Mol. Biol. 203 (1988) 467-478. Yanisch-Perron, C., Vieira, J. and Messing, J.: Improved Ml3 cloning vectors and host strains: nucleotide sequences of the Ml3 mp18 and pUC19 vectors. Gene 33 (1985) 103-119.