Genome size and genetic map of Cowdria ruminantium

E. P. de Villiersa,1, K. A. Brayton2, E. Zweygarth1 and B. A. Allsopp1

Onderstepoort Veterinary Institute, Onderstepoort, 0110, Republic of South Africa1
Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, WA 99164, USA2

Author for correspondence: E. P. de Villiers. Tel: +31 30 253 6923. Fax: +31 30 254 0784. e-mail: e.devilliers{at}vet.uu.nl


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cowdria ruminantium is the cause of a serious tick-borne disease of domestic ruminants, known as heartwater or cowdriosis. The organism belongs to the tribe Ehrlichieae, which contains obligate intracellular pathogens, causing several important animal and human diseases. Although a few C. ruminantium genes have been cloned and sequenced, very little is known about the size, gross structure and organization of the genome. This paper presents a complete physical map and a preliminary genetic map for C. ruminantium. Chromosomal C. ruminantium DNA was examined by PFGE and Southern hybridization. PFGE analysis revealed that C. ruminantium has a circular chromosome approximately 1576 kb in size. A physical map was derived by combining the results of PFGE analysis of DNA fragments resulting from digestion of the whole genome with KspI, RsrII and SmaI and Southern hybridization analysis with a series of gene probes and isolated macrorestriction fragments. A genetic map for C. ruminantium with a mean resolution of 290 kb was established, the first for a member of the Ehrlichieae. A total of nine genes or cloned C. ruminantium DNA fragments were mapped to specific KspI, RsrII and SmaI fragments, including the major antigenic protein gene, map-1.

Keywords: Cowdria ruminantium, Rickettsiales, genome size, genetic map

Abbreviations: MACS, magnetic cell separation

a Present address: Division of Parasitology and Tropical Veterinary Medicine, Faculty of Veterinary Medicine, Utrecht University, PO Box 80.165, 3508 TD Utrecht, The Netherlands.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cowdria ruminantium is an obligate intracellular rickettsial pathogen, a member of the group III Ehrlichieae (Allsopp et al., 1996 ). C. ruminantium is transmitted by ticks of the genus Amblyomma and causes heartwater in domestic ruminants throughout sub-Saharan Africa (Uilenberg, 1983 ; Jongejan & Bekker, 1999 ), where it is a serious constraint to animal production. The disease is also present in the Caribbean region, where it poses the threat of spreading to the American mainland (Perreau et al., 1980 ; Barré et al., 1987 ). The existing vaccine is difficult to manufacture, distribute and administer and cannot be used in non-endemic areas, because it contains viable organisms (Oberem & Bezuidenhout, 1987 ). A new, cost-effective vaccine, preferably a recombinant one, is needed for the control of heartwater, and to achieve this goal a better understanding of the molecular genetics of the organism is required.

All Ehrlichia species are parasites of eukaryotic cells and many species cause important diseases of domestic animals and of humans (Rikihisa, 1991 ). Molecular genetic analysis of the Ehrlichieae, with the exception of Anaplasma spp., has been severely impeded because of their obligate intracellular habitat. The organisms are therefore difficult to separate from host-cell components, especially the nuclei, which makes genomic library construction and large-fragment DNA separation difficult.

Although a few genes from C. ruminantium have been identified and cloned (Mahan et al., 1994 ; Van Vliet et al., 1994 ; Lally et al., 1995 ; Allsopp et al., 1997 ; Brayton et al., 1997 ), nothing was previously known about the structure and organization of the genome. Recent advances in the purification of C. ruminantium (De Villiers et al., 1998 ), using a combination of Percoll density-gradient centrifugation (Tamura et al., 1982 ) and high-gradient magnetic cell separation (MACS) (Miltenyi et al., 1990 ), have allowed us to study some of the basic molecular genetics of this organism. The objectives of this study were to determine the genome size of the Welgevonden isolate of C. ruminantium by PFGE analysis and restriction endonuclease digestion, and to construct a combined physical and genetic map by integrating PFGE and Southern blot analytical data.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Preparation of chromosomal DNA plugs.
The Welgevonden isolate of C. ruminantium (Du Plessis, 1985 ) was harvested from heavily infected bovine endothelial cell cultures (Bezuidenhout et al., 1985 ). The organisms were tested for Mycoplasma contamination by three separate tests: (i) amplification by PCR using primers described by Kuppeveld et al. (1994) , (ii) DNA fluorochrome staining (Del Giudice & Hopps, 1978 ) and (iii) direct culturing in MycoVito medium (International Mycoplasma). Cultures that were positive for Mycoplasma were decontaminated using the following methods. Mycoplasma contamination was removed by first of all treating the bovine endothelial cells with 10–50 µg ciprofloxacin hydrochloride ml-1 (Bayer). C. ruminantium was cleaned by infection of mice and subsequent re-isolation into clean feeder bovine endothelial cell monolayers (Eremeeva et al., 1994 ). The Welgevonden isolate has been in culture in our laboratory for several years and is used as our reference stock. Although the isolate has not been cloned, we have evidence, as discussed below, that it does not contain a mixture of genotypes. Endothelial cells were disrupted to release C. ruminantium organisms in all stages of development and host-cell debris was separated from the organisms on a discontinuous Percoll (Sigma) density gradient (Tamura et al., 1982 ; De Villiers et al., 1998 ). Purified organisms were stained with acridine orange and counted (Knight et al., 1989 ), resuspended at a concentration of 1·2 x 109 cells ml-1 in Dulbecco’s phosphate-buffered saline (pH 7·1) and embedded in 0·5% agarose by mixing the suspension 1:1 (v/v) with FMC InCert Agarose [1 % (w/v) in PBS] (FMC Bioproducts). The agarose mixture was cast into 100 µl moulds to yield plugs containing approximately 1 µg DNA per plug. Plugs were treated with lysis buffer [6 mM Tris/HCl, 1 M NaCl, 0·01 mM EDTA, 0·2% sodium deoxycholate (v/v) and 0·5% (v/v) Sarkosyl, pH 7·6] for 16 h at 37 °C. Lysis buffer was replaced with reaction buffer [10 mM Tris/HCl (pH 8·0), 0·01 mM EDTA, 1 % (v/v) Sarkosyl] containing 15 units Pretaq ml-1 (Roche Molecular Biochemicals) and the plugs were incubated for 2 h at 60 °C to digest all proteins. Plugs were washed twice for 20 min at room temperature with 10 mM Tris/HCl (pH 8·0), 0·1 mM EDTA and stored at 4 °C in the same buffer. Pretaq is inactive at 37 °C and therefore does not require specific inactivation.

Restriction endonuclease digestion.
Individual plugs containing intact C. ruminantium DNA were digested with restriction endonucleases (Roche Molecular Biochemicals). Agarose plugs were equilibrated with 100 µl suitable restriction endonuclease buffer at 4 °C for 20 min. The buffer was exchanged with 100 µl fresh buffer containing 10 units of restriction endonuclease and the mixture incubated for 4 h at the appropriate temperature. Reactions were stopped by the addition of 50 µl 0·5 M EDTA (pH 8·0).

PFGE.
DNA plugs were loaded into wells of a 1% (w/v) FastLane agarose gel (FMC Bioproducts) in 0·025 M TBE (2·25 mM Tris base, 2·25 mM boric acid, 0·05 mM EDTA, pH 8·5). Electrophoresis was carried out at 13 °C in a Rotaphor Type 5 PFGE apparatus (Biometra) using the following separation conditions: voltage 120–125 V (logarithmic ramp); pulse rate 5–90 s (linear ramp); rotation angle 110–125° (logarithmic ramp) with a duration of 37 h. Saccharomyces cerevisiae chromosomal DNA markers (New England Biolabs) and Low Range PFG markers (New England Biolabs) were used as molecular size markers. Gels were stained with ethidium bromide and restriction fragments photographed and analysed with a Lumi-Imager F1 Workstation (Roche Molecular Biochemicals). Band sizes were calculated from at least three independent separations of restriction fragments.

Two-dimensional electrophoresis was carried out according to the method of Römling & Tümmler (1991) . A lane containing restriction endonuclease digested DNA was cut from a PFGE gel, incubated with 200 U of a second restriction endonuclease in 3 ml of appropriate reaction buffer and loaded across the width of a second PFGE gel. This gel was then electrophoresed at right angles to the sample strip. DNA spots were visualized with ethidium bromide staining and subsequently photographed and analysed with the Lumi-Imager F1 Workstation.

Southern hybridization.
Gels were depurinated for 15 min in 0·25 M HCl, washed twice for 30 min in denaturing solution (0·5 M NaOH, 1·5 M NaCl) and for 1 h in neutralizing solution (1·5 M NaCl, 0·5 M Tris/HCl, 1 mM EDTA, pH 7·5) and then transferred to MagnaCharge nylon transfer membrane (MSI) with 20x SSC [1x SSC is 150 mM NaCl plus 75 mM sodium citrate (pH 7·0)] (Sambrook et al., 1989 ). After transfer, DNA was fixed to the membrane by baking at 80 °C for 2 h. DNA probes were labelled using a Megaprime labelling kit (Amersham) and [{alpha}-32P]dATP (3·7 x 108 Bq ml-1, Amersham) according to the instructions of the manufacturer. Partially purified organisms were further purified by high-gradient MACS (Miltenyi et al., 1990 ) as previously described (De Villiers et al., 1998 ) to obtain genomic DNA that contained <10% bovine DNA for use as probes. Cloned genes (see Table 4) and gel-purified restriction endonuclease fragments were also used as probes. Filters were prehybridized in 0·5 M sodium phosphate (pH 7·4), 7% SDS, for at least 1 h and hybridized overnight at 64 °C in the same buffer following addition of probe. Excess probe was removed by washing at room temperature (1x SSC, 0·5% SDS; 2 x10 min) followed by washing in the same buffer (1x10 min at 64 °C). Blots were stripped with boiling 0·5% SDS and stored at 4 °C until re-use. Hybridization results were visualized by autoradiography and analysed with the Lumi-Imager F1 Workstation.


View this table:
[in this window]
[in a new window]
 
Table 4. Restriction fragments resulting from KspI, RsrII and SmaI digestion which gave a hybridization signal with several DNA probes

 

   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The genome of C. ruminantium consists of a circular chromosome
PFGE electrophoresis of intact, undigested C. ruminantium genomic DNA, under conditions designed to separate linear fragments in the 3–2000 kb size range, revealed that most of the sample DNA remained in the wells (Fig. 1a, lane 3), with a small amount of DNA migrating at a size of 1600 kb. This fragment was apparently linearized chromosomal DNA that was sheared or nicked during the isolation of the organisms or the preparation of genomic DNA. The intensity of this fragment varied from DNA preparation to DNA preparation, which supports this conclusion (data not shown). Genomic DNA from uninfected bovine endothelial cells only hybridized to lane 2 (Fig. 1b) containing bovine DNA. DNA in the wells, the compression front at 1900 kb, and the individual C. ruminantium restriction fragments gave a strong hybridization signal with genomic DNA from purified C. ruminantium organisms (Fig. 1c). This probe also hybridized with bovine DNA in lane 2 (Fig. 1c), because C. ruminantium chromosomal DNA purified on Percoll density gradients still contains 5–10% bovine DNA (De Villiers et al., 1998 ). Southern hybridization of the same blot with a specific C. ruminantium gene probe, map-1, gave a strong signal with specific restriction fragments on which map-1 is located, as well as undigested C. ruminantium DNA (Fig. 1d). The data showed that the majority of C. ruminantium chromosomal DNA remained in the wells during PFGE, which is indicative of a circular genome. Plasmids were not detected either by PFGE or alkaline lysis extraction.



View larger version (59K):
[in this window]
[in a new window]
 
Fig. 1. PFGE and Southern hybridization analysis of undigested and digested C. ruminantium genomic DNA. (a) Ethidium bromide stained gel. Corresponding Southern hybridization with, (b) bovine (endothelial cell) genomic DNA, (c) purified C. ruminantium DNA and (d) map-1 gene DNA as probe. Lanes: 1, lambda ladder and S. cerevisiae molecular mass markers; 2, undigested genomic bovine DNA; 3, undigested genomic C. ruminantium (Welgevonden isolate) DNA; 4, RsrII-digested, 5, KspI-digested and 6, SmaI-digested C. ruminantium DNA. PFGE running conditions were for 3–2000 kb fragment separation. Sizes are indicated on the left. Images were captured on the Lumi-Imager F1 Workstation and quantified and analysed with the Lumi Analyst Version 3.0 software.

 
Genome size determination
Eight restriction endonucleases that recognize GC-rich cleavage sites (ApaI, KspI, MluI, NaeI, NotI, RsrII, SfiI and SmaI) were tested for their suitability for mapping the genome. We expected that these enzymes would recognize only a few sites in the C. ruminantium genome due to its low G+C content (26–33 mol%) (Brayton et al., 1997 ). Three restriction endonucleases, KspI, RsrII and SmaI, were found to digest the genome into a small number of DNA fragments (2–6) which could easily be distinguished and sized by PFGE (Fig. 1a). Table 1 summarizes the mean sizes of the fragments obtained from at least three separate gels with the restriction endonucleases selected for this study.


View this table:
[in this window]
[in a new window]
 
Table 1. DNA fragments from C. ruminantium (Welgevonden isolate) produced by digestion with three restriction endonucleases

 
Digestion of chromosomal C. ruminantium DNA with the restriction endonuclease RsrII yielded two fragments (Fig. 1a, lane 4), fragment 1 (R1) which migrated lower than the compression front at approximately 1453 kb and another smaller fragment 2 (R2) at 181 kb. Addition of these two fragments gave a genome size estimate of 1633±15 kb (Table 1). Southern hybridization with a C. ruminantium genomic DNA probe yielded strong signals with residual DNA in the well and with the R1 fragment and a faint signal with the R2 fragment (Fig. 1c, lane 4).

The restriction endonuclease KspI produced five fragments (Fig. 1a, lane 5), which added up to 1551±114 kb (Table 1). SmaI produced six fragments (Fig. 1a, lane 6) with a total length of 1545±143 kb (Table 1). Based on the data obtained by these three enzyme digestions, we estimated the C. ruminantium genome size at 1576±91 kb.

Construction of a physical map
Both physical methods and hybridization analysis were used in the construction of a physical map for C. ruminantium (Welgevonden). To supplement the data obtained with restriction endonuclease digestion of whole genomic DNA with KspI, RsrII and SmaI, partial digestions with these endonucleases were obtained either by limiting the amount of restriction endonuclease added to a reaction or by limiting the reaction time of the digestions (Fig. 2a).



View larger version (64K):
[in this window]
[in a new window]
 
Fig. 2. PFGE of restriction fragments obtained from (a) partially digested C. ruminantium DNA. Lanes: 1, KspI; 2, SmaI. (b) PFGE of restriction fragments obtained from double digestion of C. ruminantium DNA with KspI and RsrII. Lanes: 1, KspI; 2, KspI and RsrII; 3, RsrII only. (c) PFGE of restriction fragments obtained from double digestion of C. ruminantium DNA with KspI and SmaI, and SmaI and RsrII. Lanes: 1, RsrII; 2, SmaI; 3, KspI; 4, KspI and SmaI; 5, SmaI and RsrII. PFGE running conditions for all gels were for 3–2000 kb fragment separation. Sizes are indicated on the left of each panel. Images were captured on the Lumi-Imager F1 Workstation.

 
To provide further information about the linkage of fragments, double digestion of total genomic DNA was performed with combinations of KspI and RsrII, SmaI and RsrII, and KspI and SmaI (Table 2). Fragments K1, K3, K4 and K5 were not digested by RsrII (Fig. 2b). K2 was digested at two sites by RsrII, yielding the R2 fragment and two additional fragments, KR1 and KR2, of 253 and 18 kb, respectively (Table 2, Fig. 2b, lane 2). Fragments S1, S2, S3, S5 and S6 were not digested by RsrII. S4 was digested at two sites by RsrII, yielding the R2 fragment and two additional fragments, SR1 and SR2, of 70 and 6 kb, respectively (Table 2, Fig. 2c, lane 5). The results with KspI and SmaI were more complex. Fragments K4 and K5 were not digested by SmaI, and fragments S1, S5 and S6 were not digested by KspI. SmaI digested fragment K1 into three fragments, S1, SK4 and SK5 (Table 2, Fig. 2c, lane 4).


View this table:
[in this window]
[in a new window]
 
Table 2. Fragment sizes resulting from double digestion of C. ruminantium (Welgevonden isolate) DNA

 
These results were confirmed with two-dimensional restriction fragment analysis. Chromosomal DNA was digested with KspI and the resulting fragments (K1 through K5) were separated by PFGE. The lane containing the DNA bands was cut out, digested with SmaI, and separated in the second dimension. As a control, KspI- and SmaI-digested chromosomal DNA samples were also separated on the same gel (Fig. 3). The results showed that fragment K1 had two SmaI sites, yielding three fragments: a 525 kb fragment (Fig. 3, lane 3 fragment S1), a 40 kb fragment (Fig. 3, lane 3 fragment SK4) and a smaller fragment of 11 kb (Fig. 3, lane 3 fragment SK5). Fragment K2 also had two SmaI sites, yielding three fragments, a doublet consisting of S5 and SK2, and SK3 (Fig. 3, lane 3). A small amount of undigested K2 was also visible. Fragment K3 had a single SmaI site, yielding two fragments, SK1 and S6 (Fig. 3, lane 3). With data obtained from the partial digests, double digests and two-dimensional restriction fragment analysis, a physical map was constructed which was confirmed by hybridization analysis.



View larger version (82K):
[in this window]
[in a new window]
 
Fig. 3. Determination of SmaI restriction endonuclease sites in KspI fragments by two-dimensional PFGE. A lane containing KspI digestion fragments was excised from a PFGE gel, digested with SmaI and loaded lengthwise and separated on a second PFGE gel together with KspI- and SmaI-digested DNA. Lanes: 1, SmaI; 2, KspI; 3, two-dimensional separation of KspI fragments digested with SmaI. Sizes are indicated on the left. Images were captured on the Lumi-Imager F1 Workstation (part of the same gel as in Fig. 2c).

 
Southern blots of single- and/or double-digested C. ruminantium genomic DNA were probed with individual restriction fragments (Table 3) and with cloned or PCR-amplified genes of C. ruminantium obtained from various sources (Table 4). Individually labelled restriction fragments were hybridized to PFGE-separated single- and/or double-digested chromosomal DNA fragments. For example, the K1 fragment hybridized to fragments S1, S2 and S4, indicating the linkage between these three fragments (Table 3). The rpoC gene probe hybridized to fragments K1 and S2 and to a 45 kb fragment resulting from the KspI/SmaI double digest, indicating that fragments K1 and S2 were linked with a 45 kb common fragment on which the rpoC gene was located (Table 4, Fig. 4). In a similar fashion, most of the KspI, SmaI and RsrII fragments were located on the C. ruminantium physical map thus confirming its topology (Fig. 5).


View this table:
[in this window]
[in a new window]
 
Table 3. Hybridization with genomic restriction fragments isolated from agarose gels

 


View larger version (44K):
[in this window]
[in a new window]
 
Fig. 4. PFGE of restriction fragments obtained from double digestion of C. ruminantium DNA with KspI and SmaI and probed with the rpoC gene. (a) Ethidium bromide stained gel. (b) Corresponding Southern hybridization with rpoC gene. Lanes: 1, SmaI; 2, KspI; 3, KspI and SmaI. PFGE running conditions for all gels were for 3–2000 kb fragment separation. Sizes are indicated on the left. Images were captured on the Lumi-Imager F1 Workstation (part of the same gel as in Fig. 2c).

 


View larger version (33K):
[in this window]
[in a new window]
 
Fig. 5. Physical and preliminary genetic map of the C. ruminantium chromosome. The map was constructed from partial and double digestions obtained with KspI, RsrII and SmaI followed by PFGE separation and subsequent Southern hybridization of 32P-labelled fragments cut out from restriction endonuclease digested genomic DNA separated by PFGE. DNA probes used for genetic mapping are shown in Table 4.

 
Construction of a genetic map
Nine genes or cloned DNA fragments were located on the C. ruminantium physical map by hybridization to restriction endonuclease digested genomic DNA separated by PFGE. The probes used in this study, and the results obtained, are listed in Table 4. For example, the map-1 gene probe hybridized to the R1 fragment and not to the R2 fragment. It further hybridized only to the K1 and S1 fragments (Fig. 1d, Table 4), indicating that it was located on fragment 1 of RsrII, KspI and SmaI. In the same manner, all the genes and cloned DNA fragments were located on the physical map as indicated in Table 4 and Fig. 5, with position 0/1576 chosen at random.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We have determined a complete physical map of the genome of the Welgevonden isolate of C. ruminantium. This is the first such map for a member of the tribe Ehrlichieae. The Welgevonden isolate has not been somatically cloned, but the polymorphic map-1 gene of this isolate has been cloned and sequenced on different occasions by different laboratories over a period of more than 6 years (Brayton et al., 1997 ; A. Bensaid & D. Martinez, personal communication, 1999; Sulsona et al., 1999 ). In all but one unusual case (Sulsona et al., 1999 ) the sequence of this gene has remained identical, which provides good evidence that this is not a mixed isolate. More significantly, recent macrorestriction profile analysis by PFGE of the Welgevonden isolate also gave a consistent and clear pattern, further indicating that the isolate is genetically stable and does not contain a mixture of genotypes (De Villiers et al., 2000 ).

In the exception noted above, Sulsona et al. (1999) found that the map-1 sequence of the Welgevonden isolate (GenBank accession number AF125274) grown in their lab for a number of years was 86·88% identical to the earlier Welgevonden sequence (GenBank accession number U49843). The sequence they obtained was reported to be identical to a new Zimbabwean isolate called Lemco (GenBank accession number AF125277). However, examination of the sequences reported in GenBank show these sequences to share only 99·6% identity. This is the only known report of two different pathogenic isolates of C. ruminantium having identical map-1 sequences. All other laboratories that have sequenced the map-1 gene from the Welgevonden isolate have obtained the same sequence as U49843 (Brayton et al., 1997 ; A. Bensaid & D. Martinez, personal communication, 1999).

PFGE separation of intact C. ruminantium chromosomal DNA indicated that the chromosome was circular, in common with most other prokaryotic organisms, with an estimated size of 1576 kb. A physical map was constructed by a combination of complementary techniques. Single, double and partial digests separated with normal PFGE and two-dimensional PFGE in combination with hybridization experiments with genetic markers and isolated restriction fragments. A total of 13 restriction fragment sites were positioned on the map (5 KspI, 2 RsrII and 6 SmaI sites) with a mean resolution of 290 kb. Similar methods were used to construct the Haemophilus influenzae (Lee et al., 1989 ) and Mycoplasma mycoides (Pyle & Finch, 1988 ) physical maps.

Relatively few genes of C. ruminantium have been identified and therefore only a limited genetic map could be constructed by Southern hybridization with nine probes (Table 4). Interestingly, the majority of these probes were located on one half of the map with the three stable cosmids from a SuperCos1 library clustered on the 271 kb KspI–SmaI fragment together with the 16S rRNA and the 28G ORF.

Previously, using PFGE of uncut circular DNA, we estimated the C. ruminantium chromosome size at 1900 kb (De Villiers et al., 1998 ). The discrepancy is probably due to the fact that circular DNA does not migrate predictably during PFGE (Birkelund & Stephens, 1992 ) and the large molecular mass markers behaved erratically under the PFGE conditions used (New England Biolabs, Yeast Chromosome PFG Marker technical bulletin). In our earlier study, a 815 kb extrachromosomal element was also identified, which upon sequencing of the PCR-amplified product turned out to be an unknown Mycoplasma sp., most closely related to Mycoplasma indiense. All cultures used in this study were tested and shown to be free of mycoplasma contamination by three different methodologies as described earlier.

Several close relatives of C. ruminantium genomes have been determined with PFGE. The closest relative of C. ruminantium is the group III Ehrlichia sp. Ehrlichia chaffeensis with a genome size of 1225·8 kb (Rydkina et al., 1999 ). Very close in size to the C. ruminantium chromosome is the unnamed agent of human granulocytic ehrlichiosis (HGE), which has a genome size of 1494 kb (Rydkina et al., 1999 ). Anaplasma marginale has a slightly smaller genome, between 1200 and 1260 kb (Alleman et al., 1993 ). Ehrlichia sennetsu and Ehrlichia risticii, on the other hand, have substantially smaller genomes of 878 kb and 880 kb, respectively (Rydkina et al., 1999 ). Unfortunately, no physical or genetic maps are available for these organisms. Rickettsia prowazekii, the causative agent of epidemic typhus and only distantly related to the Ehrlichieae, is the only rickettsial organism to have its genome completely mapped and sequenced (Andersson et al., 1998 ). This parasite has a genome size of 1111523 bp. The size of the C. ruminantium genome, therefore, appears to be close to that of its relatives, whose small genomes are consistent with their obligate intracellular lifestyles. Due to the limited amount of genetic data available for C. ruminantium and the fact that only a small number of genes can be compared with this method, it is impossible to make further comparisons between our map and that of R. prowazekii.

However, the C. ruminantium physical map has already been shown to be valuable for comparative genome analyses of C. ruminantium isolates. The restriction endonucleases identified for the construction of the physical map were applied in macrorestriction profile analysis by PFGE to distinguish seven isolates of C. ruminantium from geographically different areas (De Villiers et al., 2000 ). Statistical analysis of the macrorestriction profiles indicated that all isolates were indeed distinct from each other. These data will contribute to a better understanding of the molecular epidemiology of this pathogen and may be further exploited for the identification of genotype-specific DNA probes. In addition, the estimated genome sizes for the seven isolates ranged from 1546 to 1675 kb, in agreement with the genome size calculated in this paper.

The physical and genetic map of the C. ruminantium genome constitutes a significant step forward in the study of the molecular biology of this organism, and will considerably aid the completion of the genome sequencing project which is currently under way.


   ACKNOWLEDGEMENTS
 
The authors wish to thank E. Horn and A. I. Josemans for in vitro cultivation of C. ruminantium, M. Venter and A. Pretorius for help in the preparation of gene probes and Dr F. Jongejan for critically reading the manuscript. We thank CIRAD-EMVT for providing the 28G RAPD product probe. This work was funded by the Agricultural Research Council of South Africa. Additional support was provided by the INCO-DC program of the European Union under contract #ICI8-CT95-0008, entitled ‘‘Integrated Control of Cowdriosis'.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Alleman, A. R., Kamper, S. M., Viseshakul, N. & Barbet, A. F. (1993). Analysis of the Anaplasma marginale genome by pulsed-field electrophoresis. J Gen Microbiol 139, 2439-2444.[Medline]

Allsopp, B. A., Allsopp, M. T. E. P., Du Plessis, J. L. & Visser, E. S. (1996). Uncharacterized Ehrlichia spp. may contribute to clinical heartwater. Ann N Y Acad Sci 791, 17-23.[Medline]

Allsopp, M. T. E. P., Visser, E. S., Du Plessis, J. L., Vogel, S. W. & Allsopp, B. A. (1997). Different organisms associated with heartwater as shown by analysis of 16S ribosomal RNA gene sequences. Vet Parasitol 71, 283-300.[Medline]

Andersson, S. G. E., Zomorodipour, A., Andersson, J. A. & 7 other authors (1998). The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396, 133–140.[Medline]

Barré, N., Uilenberg, G., Morel, P. C. & Camus, E. (1987). Danger of introducing heartwater onto the American mainland: potential role of indigenous and exotic Amblyomma ticks. Onderstepoort J Vet Res 54, 405-417.[Medline]

Bezuidenhout, J. D., Patterson, C. L. & Barnard, B. J. H. (1985). In vitro cultivation of Cowdria ruminantium. Onderstepoort J Vet Res 52, 113-120.[Medline]

Birkelund, S. & Stephens, R. S. (1992). Construction of physical and genetic maps of Chlamydia trachomatis serovar L2 by pulsed-field gel electrophoresis. J Bacteriol 174, 2742-2747.[Abstract]

Brayton, K. A., Fehrsen, J., de Villiers, E. P., van Kleef, M. & Allsopp, B. A. (1997). Construction and initial analysis of a representative {lambda}ZAPII expression library of the intracellular rickettsia Cowdria ruminantium: cloning of map-1 and three other Cowdria genes. Vet Parasitol 72, 185-199.[Medline]

Brayton, K. A., de Villiers, E. P., Fehrsen, J., Nxomani, C., Collins, N. E. & Allsopp, B. A. (1999). Cowdria ruminantium DNA is unstable in a SuperCos1 library. Onderstepoort J Vet Res 66, 111-117.[Medline]

Del Giudice, R. A. & Hopps, H. E. (1978). Microbiological methods and fluorescent microscopy for the direct demonstration of mycoplasma infection of cell culture. In Mycoplasma Infection of Cell Cultures , pp. 57-69. Edited by G. J. McGarrity, D. G. Murphy & W. W. Nichols. New York:Plenum.

De Villiers, E. P., Brayton, K. A., Zweygarth, E. & Allsopp, B. A. (1998). Purification of Cowdria ruminantium organisms for use in genome analysis by pulsed-field gel electrophoresis. Ann N Y Acad Sci 849, 313-320.[Abstract/Free Full Text]

De Villiers, E. P., Brayton, K. A., Zweygarth, E. & Allsopp, B. A. (2000). Macrorestriction fragment profiles reveal genetic variation of Cowdria ruminantium isolates. J Clin Microbiol 38, 1967-1970.[Abstract/Free Full Text]

Du Plessis, J. L. (1985). A method for determining the Cowdria ruminantium infection rate of Amblyomma hebraeum: effects in mice injected with tick homogenates. Onderstepoort J Vet Res 52, 55-61.[Medline]

Eremeeva, M. E., Balayeva, N. M. & Raoult, D. (1994). Purification of rickettsial cultures contaminated by mycoplasmas. Acta Virol 38, 231-233.[Medline]

Jongejan, F. & Bekker, P. J. (1999). Cowdria ruminantium: recent developments in diagnostic methods, molecular characterization and vaccines. In Rickettsiae and Rickettsial Diseases at the Turn of the Third Millennium , pp. 373-386. Edited by D. Raoult & P. Brouqui. Paris:Elsevier.

Knight, S. T., Neece, V. R. & Witt, D. J. (1989). Rapid culture-independent techniques for quantitation of Chlamydia trachomatis elementary bodies. J Microbiol Methods 10, 255-263.

Kuppeveld, F. J. M., Johansson, K. E., Galama, J. M. D., Kissing, J., Bölske, G., van der Logt, J. T. M. & Melchers, W. J. G. (1994). Detection of mycoplasma contamination in cell cultures by a mycoplasma group-specific PCR. Appl Environ Microbiol 60, 149-152.[Abstract]

Lally, N. C., Nicoll, S., Paxton, E. A., Cary, C. M. & Sumption, K. J. (1995). The Cowdria ruminantium groE operon. Microbiology 141, 2091-2100.[Abstract]

Lee, J. J., Smith, H. O. & Redfield, R. J. (1989). Organization of the Haemophilus influenzae Rd genome. J Bacteriol 171, 3016-3024.[Medline]

Mahan, S. M., McGuire, T. C., Semu, S. M., Bowie, M. V., Jongejan, F., Rurangirwa, F. R. & Barbet, A. F. (1994). Molecular cloning of a gene encoding the immunogenic 21 kDa protein of Cowdria ruminantium. Microbiology 140, 2135-2142.[Abstract]

Miltenyi, S., Müller, W., Weichel, W. & Radbruch, A. (1990). High gradient magnetic cell separation with MACS. Cytometry 11, 231-238.[Medline]

Oberem, P. T. & Bezuidenhout, J. D. (1987). The production of heartwater vaccine. Onderstepoort J Vet Res 54, 485-488.[Medline]

Perez, J., Martinez, D., Debus, A., Sheikboudou, C. & Bensaid, A. (1997). Detection of genomic polymorphisms among isolates of the intracellular bacterium Cowdria ruminantium by random amplified polymorphic DNA and Southern blotting. FEMS Microbiol Lett 154, 73-79.[Medline]

Perreau, P., Morel, P. C., Barré, N. & Durand, P. (1980). Existence de la cowdriose à Cowdria ruminantium, chez les petits ruminants des Antilles Françaises (La Guadeloupe) et des Mascareignes (La Réunion et Ile Maurice). Rev Élev Méd Vét Pays Trop 33, 21-22.[Medline]

Pyle, L. E. & Finch, L. R. (1988). A physical map of the genome of Mycoplasma mycoides subspecies mycoides Y with some functional loci. Nucleic Acids Res 16, 6027-6039.[Abstract]

Rikihisa, Y. (1991). The tribe Ehrlichieae and ehrlichial diseases. Clin Microbiol Rev 4, 286-308.[Medline]

Römling, U. & Tümmler, B. (1991). The impact of two-dimensional pulsed field gel electrophoresis techniques for the consistent and complete mapping of bacterial genomes: refined physical map of Pseudomonas aeruginosa PAO. Nucleic Acids Res 19, 3199-3206.[Abstract]

Rydkina, E., Roux, V. & Raoult, D. (1999). Determination of the genome size of Ehrlichia spp., using pulsed field gel electrophoresis. FEMS Microbiol Lett 176, 73-78.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis T. (1989). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sulsona, C. R., Mahan, S. M. & Barbet, A. F. (1999). The map-1 gene of Cowdria ruminantium is a member of a multigene family containing both conserved and variable genes. Biochem Biophys Res Commun 257, 300-305.[Medline]

Tamura, A., Urakami, H. & Tsuruhara, T. (1982). Purification of Rickettsia tsutsugamushi by Percoll density gradient centrifugation. Microbiol Immunol 26, 321-328.[Medline]

Uilenberg, G. (1983). Heartwater (Cowdria ruminantium infection): current status. Adv Vet Sci Comp Med 27, 427-480.[Medline]

Van Vliet, A. H. M., Jongejan, F., van Kleef, M. & van der Zeijst, B. A. M. (1994). Molecular cloning, sequence analysis, and expression of the gene encoding the immunodominant 32-kilodalton protein of Cowdria ruminantium. Infect Immun 62, 1451-1456.[Abstract]

Waghela, S. D., Rurangirwa, F. R., Mahan, S. M., Yunker, C. E., Crawford, T. B., Barbet, A. F., Burridge, M. J. & McGuire, T. C. (1991). A cloned DNA probe identifies Cowdria ruminantium in Amblyomma variegatum ticks. J Clin Microbiol 29, 2571-2577.[Medline]

Received 22 March 2000; revised 26 June 2000; accepted 12 July 2000.