Journal of General Microbiology (1986), 132, 2581-2594.
Printed in Great Britain
2587
Overlapping Deletion in Two Spontaneous Phase Variants of
Coxiella burnetii
By M I C H A E L H. V O D K I N ' * A N D J I M C.
Rickettsia1 Diseases Laboratory, Airborne Diseases Division, US Army Medical Research
Institute of Infectious Diseases, Fort Detrick, Frederick, MD 21701-501I, USA
=Ofice of the Director of Intramural Research Programs, National Institute of Allergy and
Infectious Diseases, Bethesda, MD 20204, USA
(Received 24 February 1986;revised 1 May 1986)
Chromosomal DNA from the Nine Mile phase I strain of Coxiella bumetii (CB9MIC7) was
cloned into the cosmid vector pHC79. The resulting gene library was probed with a
radiolabelled HaeIII fragment present in the parental strain but absent from a spontaneously
derived Nine Mile phase I1 strain (CB9MIIC4). The insert, which includes the missing HaeIII
fragment, was 38.5 kb in length. When DNA from this cosmid clone was hybridized to genomic
DNA of the parental CB9MIC7 and its derivative CB9MIIC4, a number of fragments were
missing or altered in the latter strain. Restriction mapping localized the fragments to a
contiguous portion of the chromosomal DNA fragment. The data were consistent with an 18 kb
deletion in the chromosome of CB9MIIC4. Another intrastrain spontaneous derivative,
CB9MI514, also lacked the sentinel HaeIII fragment and carried a deletion of approximately
29 kb within the same cloned insert. Both deletions appeared to share a common terminus,
within the limits of resolution. In all other strains investigated, both phase I and phase 11, the
DNA represented by the insert seemed intact. The strains examined were representative of
various stages of phase variation. The relationship between the observed deletions and the
mechanism of phase transition in Nine Mile strains is discussed.
INTRODUCTION
Prolonged growth of the obligate intracellular bacterial parasite Coxiella burnetii, the
aetiological agent of Q fever, in the absence of a cellular immune response leads progressively to
a population of micro-organisms in phase I1 (Stoker & Fiset, 1956). Phase I1 cells are thought to
be avirulent and are characterized by alterations in lipopolysaccharide (LPS) chemotype
(Amano & Williams, 1984; Schramek & Mayer, 1982; Williams & Stewart, 1984), a different
spectrum of antigens or epitopes on their surface (Williams et al., 1984) and the absence of
pathogenetic factors synthesized by most phase I C. bumetii (Damrow et al., 1985; J. C.
Williams, unpublished observations). Phase I1 micro-organisms are usually unable to survive in
immunocompetent animals and the analogy has been made that the phase shifts are similar to
the smooth to rough LPS mutational transitions observed in some Enterobacteriaceae (Goldman
& Leive, 1980; Schramek & Mayer, 1982; Amano & Williams, 1984). The virulent phase I C.
bumetii appears to be at a selective biological disadvantage under the cultural conditions of the
yolk sac of embryonated chicken eggs (Kazar et al., 1974)or in eukaryotic cells in culture (Burton
et al., 1978). Because there are probably several functions controlling phase I maintenance,
selection pressure can produce independent phase I1 isolates which have distinguishable
genotypes. The European phase I1 vaccine strain, CBM44II (Johnson et al., 1976), is able to
survive several animal passages, in contrast to the Nine Mile phase I1 clone 4 (CB9MIIC4)
intrastrain derivative of a phase I parent, which is not (Amano & Williams, 1984). Thus, two
interstrain phase I1 C. burnetii provide a contrast for the classification of phase I (virulent) and
phase I1 (avirulent) strains.
0001-3282
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
2588
M. H . V O D K I N A N D J . C . W I L L I A M S
Recently, the molecular diversity of C. burnetii strains has focussed attention on the
chromosomal and plasmid DNA (Vodkin et al., 1986; Samuel et al., 1985; O’Rourke et al.,
1985). Genetic differences were demonstrated among various isolates of C. burnetii by using
restriction endonuclease cleavage patterns of chromosomal and plasmid DNA. A chromosomal
change which correlated with variation in virulence and LPS structure was identified (Vodkin et
al., 1986). Therefore, we have studied a 4.4 kb HaeIII fragment that is present in chromosomal
DNA of phase I strains and of the CBM44II strain, but is deleted in the CB9MIIC4 strain. This
paper reports mapping of the termini of the deletion in CB9MIIC4, and demonstrates that
another independent intrastrain CB9MI isolate, which has a semi-rough LPS structure and
which is completely virulent, carries a larger and overlapping deletion.
METHODS
C . burnetii strains. Most of the strains and their passage history have been described previously (Vodkin et al.,
1986). Another Nine Mile isolate in phase I was obtained as RSA514 from M. G. Peacock (NIH, NIAID, Rocky
Mountain Laboratories, Hamilton, Mont., USA). This i olate was derived from the placental tissue of guinea pigs
343 d post-infection with the parental phase I Nine Mildstrain (CB9MI) (Williams & Cantrell, 1982). It was then
passaged five times in the yolk sac of embryonated chilcken eggs and purified (Williams et al., 1981). A phase 11
isolate from Australia (CBAUSTII) was also obtained from M.G. Peacock.
Constructionof gene library. DNA extractions, restriction endonuclease digestions, gene analyses and Southern
hybridizations have been described previously (Vodkin et al., 1986). Purified chromosomal (linear or open
circular) DNA from CB9MIC7 was digested by BamHI under conditions which reduced the size of the bulk DNA
to fragments ranging in size from 15 to 30 kb. The chromosomal fragments were ligated to the cosmid vector
pHC79, also digested with BamHI, in a 2 : 1 molar ratio. The ligation mixture was packaged in oitro with a 1
packaging extract (Amersham) and then transfected into Escherichia coli TB1, obtained from BRL. Putative
recombinants were selected on L-agar plates supplemented with 20 mg ampicillin ml-l. About 200 colonies, which
comprised the pJB series, were picked and amplified.
Colony hybridization. Colony hybridization was done essentially according to the protocol of Grunstein &
Hogness (1975). DNA was radiolabelled in oitro by nick translation (Rigby et al., 1977).
Genetic mapping. A restriction map for the cloned DNA was deduced by double digestions, while taking into
account the map published for pHC79 (Collins & Hohn, 1978). Localization of the specific Hue111 fragments was
done by hybridizing them to single and double digests of cloned DNA. The assignment of the deletion termini was
based on single and double digestions of genomic DNA, followed by hybridizations with cloned DNA.
RESULTS
Screening and identifying clones covering the deletion in CB9MIIC4
We constructed a library (designated pJB) of about 200 clones using a partial BamHI digest of
CB9MIC7 chromosomal DNA. This DNA was then inserted into the BamHI site of the cosmid
pHC79. Since C. burnetii was propagated in and then purified from the yolk sacs of embryonated
chicken eggs, contamination of bacterial DNA by chicken DNA could represent a serious
problem. However, the cloned DNA used in this study hybridized to C. burnetii DNA but not to
chicken DNA. The largest (4-4 kb) HaeIII fragment of CB9MIC7 DNA (Fig. 1, lane A) was
purified, radiolabelled and then used to probe the pJB bank. Three of the colonies (pJB16,
pJB15O and pJB153) scored positive (data not shown); their cosmid DNA was extracted,
digested with HaeIII and analysed by gel electrophoresis. DNA from two (pJB16 and pJB153) of
the three clones displayed an upper band that co-migrated with the 4.4 kb fragment from
CB9MIC7 (Fig. 1). Clone pJBl50 (Fig. 1, lane C) contained a band with a greater molecular size
than the 4.4 kb fragment, and did not contain the third band generated by HaeIII digestion of
pJBl6 and pJB153 DNA. However, all of the other bands appeared to co-migrate with those of
the other two cosmid clones.
Identification of chromosomal defects in CB9MIIC4 and CB9MI.514
The cosmid DNA from pJB153 was radiolabelled in vitro by nick translation and utilized to
probe HaeIII-digested chromosomal DNA of several C. burnetii isolates for fragments that can
be resolved in this gel system. The hybridization patterns were identical except for those of three
strains, CBPAVI (Fig. 2, lane C), CB9MIIC4 (Fig. 2, lane D) and CB9MI514 (Fig. 3, lane D). In
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
Deletions of DNA in Coxiella burnetii
A
B
C
2589
D
kb
6.6
4.4
2.3
2.0
0.6
Fig. 1 . Hue111 digestion patterns of (A) ch.romosoma1 DNA of C. burnetii CB9MIC7 and If cosmid DNA of three clones, (B) pJB16, (C) pJBlSO
and (D) pJB153, from a library of C. bunietii.
previous studies, we showed that CBPAVI lacked the 4.4 kb fragment, since it was distributed
in two discrete segments (Vodkin et al., 1986).CB9MIIC4 and CB9MI514 also lacked the 4.4 kb
fragment, in addition to several other HaeIII fragments (Fig. 3). Some bands were absent in
both strains (1,4,6 and 7), while others were present in both strains (3 and 5). Band 2 was present
in CB9MIIC4 but absent in CB9MI514, and a novel band, with an apparent molecular size of
2.0 kb, was present in CB9MI514.
Restriction map of the C. burnetii insert of pJBl53
The insert of pJB153 was mapped by using single and double restriction digestions (Fig. 4).
HaeIII fragments 1 and 2 were positioned on this map by hybridizing the respective purified
DNA to both single and double digestions of pJB153 and of chromosomal DNA from three
strains. The data suggest an 18 kb deletion, in the phase I1 derivative CB9MIIC4, which lies
entirely within a single 38-5kb insert cloned from CB9MIC7 chromosomal DNA. The deletion
of CB9MI514 was 29 kb.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
2590
M . H. VODKIN A N D J. C. WILLIAMS
(a)
A
B
C
D
E
F
G
H
@)A
B
C
D
E
F
G
H
Fig. 2. Hybridization of pJB153 to HaeIII-digested chromosomal DNA from various strains of C.
burnetii. Purified chromosomal DNA preparations were digested with Hue111 and transferred to
nitrocellulose.The blot was probed with DNA of pJB153 radiolabelledwith [32P]dATPin v i m by nick
translation. A and H, HindIII-digestedR. DNA (included as a marker and as a probe); B, CB9MIC7; C,
CBPAVI; D, CB9MIIC4; E, CBHENI; F, CBM44II: G, CBOI. (a) D N A visualized by ethidium
bromide staining. (6) Pattern of hybridization as revealed by autoradiography.
The left terminus of both deletions appeared to coincide, within the limits of resolution of this
agarose gel system. Neither deletion strain showed any sequences of fragment I (Vodkin et aZ.,
1986; unpublished observations) or the 4.5 and 2-0 kb CZaI fragments. HaeIII fragment 5
contained a PstI site (Fig. 4), and was present in all strains. Therefore, the left terminus from
each deletion strain had to be included within the cloned insert (see Fig. 5).
The right end of the deletion of CB9MIIC4 was fixed by hybridizing HueIII fragment 2 to
restricted pJB153 DNA. Fragment 2 contained one of the termini (Figs 2-4), and a PstI site
located slightly more distal to one of the HueIII junctions (Fig. 4). The sequences of fragment 2
partition between the 6 and 2.5 kb ClaI fragments (not shown). Hybridizations of pJB153 to
restricted chromosomal DNA were consistent with this placement. Neither of the two 6.5 kb
KpnI fragments was present in CB9MIIC4 (Fig. 6). Following PstI digestion, one or more of the
three large fragments predicted was apparently lost (Figs 5 and 6), and a different 7 kb fragment
appeared in CB9MIIC4 (Fig. 6). Consistent with these data was the location of the right
terminus of the deletion 17.5 kb from the right end of the insert (Fig. 5).
The right end of the deletion in CB9MI514 was mapped with hybridizations of pJB153 to
restricted genomic DNA. Both 6.5 kb KpnI fragments were missing but an 8.5 kb KpnI
fragment, not present in the parental DNA, appeared (data not shown). In a PstI digest, a novel
7-5 kb fragment was present. These data are consistent with a location of the right terminus of
the deletion about 6 kb from the right terminus of the cloned insert (Fig. 5).
Failure to detect hybridization on a Southern blot with purified band 2 suggested that it was
completely deleted in CB9MI514. The novel band present in CB9MI514 probably represented a
truncated HueIII fragment 4, since it was the only unaccounted fragment of sufficient size.
DISCUSSION
Initial restriction analysis of the isogenic pair CB9MIC7 and CB9MIIC4 revealed a missing
fragment of chromosomal DNA in the phase I1 strain. This DNA fragment was purified from
the phase I strain and used as a probe to show that sequences homologous to the probe were not
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
259 1
Deletions of DNA in Coxiella burnetii
( a ) A
B
C
D
(b)A
B
C
D
Fig. 3. Hybridization of pJB153 to HueIII-digested chromosomal DNA from strains of C. burnefii that
carry deletions in DNA included in the probe. A, HindIII-digested L D N A ; B, CB9MIC7; C,
CB9MIIC4; D, CB9MI514. (a) DNA visualized by ethidium bromide staining. (b) Pattern of
hybridization as revealed by autoradiography.
detectable in the phase I1 strain as band(s) with altered electrophoretic mobility (Vodkin et al.,
1986). In the current study, we used the probe to screen cosmid clones in a library derived from
chromosomal DNA of CB9MIC7. At least two of the three clones (pJB16 and pJB153)
contained the intact probe. Restriction mapping and hybridization of pJB 153 to chromosomal
DNA revealed that an 18 kb deletion had occurred in the phase I1 strain. The deletion in the
phase II strain included the entire HaeIII fragment 1 and a small fraction of HueIII fragment 2.
Double digestions of DNA suggested that in the HaeIII pattern of CB9MIIC4 (see Figs 1 and 3),
this partially deleted fragment now co-migrated with the fourth band.
There was evidence of heterogeneity of restriction sites in the cosmid clones, since pJB15O
differed with respect to two HueIII fragments from pJB16 and pJB153. The larger fragment in
pJB15O could not account for the sum of the masses of the two missing fragments. Thus, the
segment cloned in pJB 150 may represent a polymorphic deletion or an aberration that occurred
during or subsequent to the cloning. Although cosmid clones are known to undergo
rearrangements in recA+ hosts (Williams & Blattner, 1980), these results may also reveal
heterogeneity among the phase I cells used to prepare the clone. Heterogeneity at other sites was
evident since there was a CfuI site in genomic DNA and in pJB150, but not in pJB153.
Another independently derived intrastrain isolate (CB9MI5 14) from the same Nine Mile
phase I parental stock was identified as having a similar deletion of chromosomal DNA. This
deletion shared a terminus with the CB9MIIC4 deletion, and hybridization data were consistent
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
2592
M. H. V O D K I N A N D J . C. WILLIAMS
(a)A
B
C
D
E
@ ) A
B
D
C
E
Fig. 4. Iybridization of pJB153 to single and double digests of chromosomal DNA preparations. A,
HindIIIdigested 1 DNA ;B, HueIII-digested CB9MIC7 DNA : C, HueIII-digested CB9MIIC4 DNA;
D, HueIII/PstI-digested CB9MIC7 DNA; E, HueIII/PstI-digested CB9MIIC4 DNA. (a) DNA
visualized by ethidium bromide staining. (6) Pattern of hybridization as revealed by autoradiography.
S
K
P
c c
A
A
1
I
8
C
C
10
c c
A
20
C
C
A
30
40 kb
Fig. 5. Restriction map of the insert of pJB153. The DNA of pJB153 was cut with a number of
restriction enzymes in single and several pairwise combinations. The position of the largest HueIII
fragment (1) was determined by purifying it and then hybridizing it to some of the above digests. The
placement of the other HueIII fragments was deduced from double-digestionsof pJBl53 with HueIII as
one of the endonucleases. The location of the deletion termini of CB9MIIC4 u) and CB9MI514ca> is
explained in the text.
with a loss of 29 kb of chromosomal DNA in CB9MI514. The phenotypes of the two strains are
distinguishable; CB9MI514 still retains some characteristics of phase I, including an LPS with a
structural complexity intermediate between phase I and I1 (Hackstadt et al., 1985; K.-I. Amano
and J. C. Williams, unpublished observations). In preliminary tests, CB9MI514 appeared to be
as virulent as the CB9MI parent (J. C. Williams and D. M. Waag, unpublished observations),
whereas the phase I1 derivative was avirulent and did not survive in immunocompetent animals
(Amano & Williams, 1984).
The deletions represent about 1 to 2 % of the genome of C. burnetii (Myers et al., 1980).
Screening the DNA by HaeIII digestion and analysis on agarose gels is probably effective for the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
Deletions of DNA in Coxiella burnetii
A
B
C
D
2593
E
kb
23.1
-
9.4
-
6.6 4-4-
2.3
2.0
-
-
0.6 -
Fig. 6. Hybridization of pJB153 to chromosomal
DNA preparations. A, HindIIIdigested 1 DNA;
B, KpnIdigested CB9MIC7 DNA; C, KpnIb
digested CB9MIIC4 DNA; D, PstIdigested
CB9MIC7 DNA : PstI-digested CB9MIIC4
DNA.
larger HaeIII fragments that comprise about 2 % of the genome. Therefore, if other deletions
have occurred in CB9MIIC4, they would be missed. Mapping, of course, would not, detect
missense mutations unless the restriction site also changed. Although the two Nine Mile
derivatives are considered isogenic, perhaps 100 passages (Table 1 in Vodkin et al., 1986) or
lo00 generations had ensued since the original plaque purification. One of the deletions in
CB9MIIC4 covered an 18 kb fragment. Another deletion in CB9MI514 covered a 29 kb
fragment, but it overlapped and had one end outside the region under study in CB9MIIC4.
We have characterized a set of deletions useful for mapping particular functions of C. burnetii
phase variants among spontaneous intrastrain mutants. Since knowledge of C. burnetii genetics
is non-existent, we have taken the approach of correlating specific functions in phase conversion
to absence of DNA fragments. Phase variation in C. burnetii involves a mechanism that converts
virulent cells (phase I) to avirulent cells (phase 11). In previous studies of this phenomenon, the
Nine Mile strains were used as a model to describe the serological, biochemical and
immunological consequences of this attenuation. Attenuation of phase I cells correlates with the
inability to synthesize a complete LPS structure (Amano & Williams, 1984) and to induce
pathological reactions in C57BL/10 ScN mice (Damrow et al., 1985). The 38.5 kb cloned insert
may be useful for mapping certain functions of C. burnetii phase variants among mutants
isolated spontaneously. It is interesting to speculate that the shared portion of the deletion from
12.5 to 213 kb (Fig. 4) may control some of the functions that convert the chemotype of the LPS
of a ‘semi-rough’ to that of a ‘smooth’, and/or that are involved in suppression of lymphocyte
proliferation.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52
2594
M. H. VODKIN A N D J. C. WILLIAMS
In our studies we have identified C. burnetii variants that do not conform to the generally held
view that phase variation is related only to LPS structure (Amano & Williams, 1984; Schramek
& Mayer, 1982). The lack of correlation of any deletion in the CBM44II and CBAUSTII strains
may be limited by our ability to detect point mutations or may suggest mutations in loci that
have not yet been screened. The possibility that additional genes control the biosynthesis of LPS
and the virulence components responsible for the induction of immunosuppression cannot be
eliminated.
We appreciate the excellent technical assistance from Chris Bolt, Robert Stockman and the Radiology
Department of USAMRIID. We also thank Captain Peter Markiewicz and Captain Frank Niagro for their
helpful criticisms, and Colonel E. H. Stephenson for his suggestions and encouragement.
The views of the authors do not purport to reflect the positions of the Department of the Army or the
Department of Defence.
REFERENCES
AMANO,
K.-I. & WILLIAMS,
J. C. (1984). Chemical and
immunological characterization of lipopolysaccharides from phase I and phase I1 Coxiella burnetii.
Journal of Bacteriology 160, 994- 1002.
BURTON,P. R., STUECKEMANN,
J., WELSH,R. M. &
PARETSKY,
D. (1978). Some ultrastructural effects of
persistent infections by the rickettsia Coxiella burnetii in mouse L cells and green monkey kidney (Vero)
cells. Infection and Immunity 21, 556-566.
COLLINS,J. & HOHN,B. (1978). Cosmids: a type of
plasmid gene-cloning vector that is packageable in
vitro in bacteriophage 1 heads. Proceedings of the
National Academy of Sciences of the United States of
America 75, 4242-4246.
DAMROW,T. A., WILLIAMS,J. C. & WAAG,D. M.
(1985). Suppression of in vitro lymphocyte proliferation in C57BL/10 ScN mice vaccinated with phase I
Coxiella burnetii. Infection and Immunity 47, 149156.
GOLDMAN,
R. C. & LEIVE,L. (1980). Heterogeneity of
antigens - side chain length in lipopolysaccharide
from Escherichia coli 01 1 1 and Salmonella typhimurium LT2. European Journal of Biochemistry 107, 145153.
GRUNSTEIN,M. & HOGNESS, D. (1975). Colony
hybridization: a method for the isolation of cloned
DNAs that contain a specific gene. Proceedings of the
National Academy of Sciences of the United States of
America 72, 3691-3695.
HACKSTADT,
T., PEACOCK,M. G., HITCHCOCK,
P. J. &
COLE,R. L. (1985). Lipopolysaccharide variation in
Coxiella burnetii: intrastrain heterogeneity in structure and antigenicity. Infection and Immunity 48,
359-365.
JOHNSON,
J. W., EDDY,G. A. & PEDERSEN,
C. E., JR
(1976). Biological properties of the M-44 strain of
Coxiella burnetii. Journal of Infectious Diseases 133,
334-3 38.
KAZAR,J., BREZINA,
R., SCHRAMEK,
S.,URVOLGYI,
J.,
POSPISIL,V. & KOVACOVA,
E. (1974). Virulence,
antigenic properties and physico-chemical characteristics of Coxiella burnetii strains with different
chick embryo yolk sac passage history. Acta viroIogica 10, 434-442.
MYERS,W. F.,BACA,0. G. & WISSEMAN,
C. L., JR
(1980). Genome size of the rickettsia Coxiella
hurnetii. Journal of Bacteriology 144, 460461.
O’ROURKE,A. T., PEACOCK,M., SAMUEL,J. E.,
FRAZIER,
M. E., NATVIG,D. O., MALLAVIA,
L. P. &
BACA,0. (1985). Genomic analysis of phase I and I1
Coxiella burnetii with restriction endonucleases.
Journal of General Microbiology 131, 1543-1 546.
RIGBY,P.W. J., DIECKMANN,
M., NODES,
C. &BERG,
P. (1977). Labeling deoxyribonucleic acid to high
specific activity in vitro by nick translation with
DNA polymerase I. Journal of Molecular Biology
113, 237-25 1.
SAMUEL,
J. E., FRAZIER,
M. E. & MALLAVIA,
L. P.
(1985). Correlation of plasmid type and disease
caused by Coxiella burnetii. Infection and Immunity
49, 775-779.
SCHRAMEK,
S. & MAYER,H. (1982). Different sugar
compositions of lipopolysaccharides isolated from
phase I and phase I1 cells of Coxiella burnetii.
Infection and Immunity 38, 53-57.
STOKER,
M. G. P. & FISET,
P. (1956). Phase variation of
the Nine Mile and other strains of Rickettsia burnetii.
Canadian Journal of Microbiology 2, 310-321.
VODKIN,M. H., WILLIAMS,
J. C. & STEPHENSON,
E. H.
(1986). Genetic heterogeneity among isolates of
Coxiellaburnetii. Journal of General Microbiology 132,
455-463.
WILLIAMS,
B. G. & BLATTNER,
F. R. (1980). Bacteriophage 1 vectors for DNA cloning. In Genetic
Engineering, vol. 2, p. 201. Edited by J. K. Setlow &
A. Hollaender. New York: Plenum Press.
WILLIAMS,J. C. & CANTRELL,
J. L. (1982). Biological
and immunological properties of Coxiella burnetii
vaccines in C57BL/ 10 ScN endotoxin-nonresponder
mice. Infection and Immunity 35, 1091-1 102.
WILLIAMS,
J. C. &STEWART,S.(1984). Identification of
immunogenic proteins of Coxiella burnetii phase
variants. In Microbiology - 1984, pp. 257-262.
Edited by D. Schlessinger. Washington, DC : American Society for Microbiology.
WILLIAM,J. C., PEACOCK,M. G. & MCCAUL,T. F.
( 1981). Immunological and biological characterization of Coxiella burnetii, phase I and 11, separated
from host components. Infection and Immunity 32,
840-85 1.
WILLIAMS,J. C., JOHNSTON,
M. R., PEACOCK,M. G.,
THOMAS,L. A., STEWART,S.& PORTIS,J. L. (1984).
Monoclonal antibodies distinguish phase variants of
Coxiellaburnetii. Infection and Immunity 43,421-428.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 14 Jun 2017 15:32:52