Biology at Washington & Jefferson College
 

Joshua W. Courtney
Patrick J. Ford
Emily K. Nease
Biology 311
20 November 2000 

Within the last decade, significant progress has been made toward understanding the way that Bartonella (formerly Rochalimaea) species are maintained and transmitted. Several species of Bartonella that exist in nature are associated with several human diseases. Bartonella quintana, historically the causative agent of trench fever, has recently been isolated in homeless persons throughout the United States and France (Kosoy, Regnery, Kosaya, and Childs, 1999). Recent studies have identified Bartonella henselae to be the agent responsible for causing cat-scratch disease (CSD) (Sander, Posselt, Bohm, Ruess, and Altwegg, 1999). The majority of pathogenic B. henselae have been isolated in the United States (Arvand, Wendt, Regnathm, Ullrich, and Hahn, 1998). Evidence has suggested that both B. quintana and B. henselae produce opportunistic infections in patients with acquired immunodeficiency syndrome (AIDS) and other immunoincompetent persons (Comer, Flynn, Regnery, Vlahov, and Childs, 1996). In addition, Bartonella elizabethae, which has recently been isolated from a patient suffering from endocarditis, is thought to also have an urban transmission cycle. Little is known, however, about the natural history and transmission of these agents (Kosoy et al., 1999).

Within the last four years, combined laboratory and field investigations have revealed that multiple Bartonella spp. exist in wildlife populations (Kosoy et al., 1999). Recent research dealing with the natural occurrence of Bartonella spp. in rodent communities throughout the southeastern United States has also indicated that various species of Bartonella are widely distributed and highly prevalent in rodent species (Handley and Regnery, 2000).

Previous attempts have been made to develop rapid detection assays to detect Bartonella species prior to the onset of sub-clinical infection. However, Bartonella spp. present slow, fastidious growth characteristics that make diagnosis of Bartonella-associated illnesses very difficult.

Previous rapid detection methods include: PCR amplification of the 16S-23S rRNA intergenic region with species-specific primers; restriction fragment length polymorphism (RFLP) analysis of PCR-amplified 16S genes; RFLP analysis of the PCR-amplified citrate-synthase gene (Jensen, Fall, Rooney, Kordick, and Breitschwerdt, 2000). These detection methods are tedious because they require multiple PCR amplification reactions and/or additional sample-processing steps beyond the primary PCR amplification (Kosoy et al., 1999).

Other Bartonella detection methods that do not rely on multi-step PCR amplifications have been developed such as enterobacterial repetitive intergenic consensus PCR (ERIC-PCR), repetitive extragenic palindromic PCR (REP-PCR), and arbitrarily primed PCR (AP-PCR) (Jensen et al., 2000). Although these assays have been successfully implemented in the detection of Bartonella spp., they are very sensitive to experimental variation and make reproducibility and standardization difficult.

A new typing method known as infrequent-restriction site PCR (IRS-PCR) has been proposed that may become a universal tool for molecular analysis of pathogen species (Yoo, Choi, Shin, Huh, Cho, Kim, Kin, and Kang, 1999). The main strategy of IRS-PCR is the selective amplification of the DNA sequence located between both frequently and infrequently occurring restriction sites by using adaptors and primers based on these two enzymes (Sambrook, Fritsch, and Maniatis, 1989). The discriminatory power of IRS-PCR has been shown to be equal to that of pulse-field gel electrophoresis (PFGE), the method currently used to distinguish Bartonella spp. Yoo et al., recently applied IRS-PCR to clinical isolates of Actinobacter baumannii and Serratia marcescens and found that IRS-PCR and PFGE are equally discriminatory; however, IRS-PCR is less tedious and less laborious (Yoo et al., 1999).

We plan to compare the effectiveness and usefulness of both PFGE and IRS-PCR to identify virulent and non-virulent Bartonella spp. at the species and subspecies level. Previous evidence suggests that IRS-PCR will provide a means to identify Bartonella spp. that is more effective than previous detection methods. We will compare both PFGE and IRS-PCR detection assays in an attempt to identify pathogenic Bartonella spp. from both samples identified through previous detection methods and samples that have never been tested. We will then evaluate each detection method based on time and cost-effectiveness, as well as, the effectiveness of identification.

Twelve separate species of Bartonella have been isolated. The genus Bartonella now includes four species that may infect humans: B. bacilliformis; B. quintana; B. henselae; and B. elizabethae (Matar, Koehler, Malcolm, Lambert-Fair, Tappero, Hunter, and Swaminathan, 1999).

B. bacilliformis is the agent of Carrion's disease (Massei, Massimetti, Messina, Macchia, and Maggiore, 2000). In 1993, B. bacilliformis (originally Rochalimaea species) was removed from the genus Rochalimaea and included in the genus Bartonella, within the family Bartonellaceae (Koehler, Glaser, and Tappero, 1994).

Bartonella quintana is the microbe responsible for the disease known as trench disease and the syndrome bacillary angiomatosis (Comer et al., 1996). Trench disease gained notoriety during World War I because it was found to plague soldiers confined to trenches. During that time, the louse was found to be the principal vector species responsible for the transmission of trench disease. The disease had not been identified in the United States until ten cases were reported among the homeless population of Seattle, Washington in 1992. Trench disease has since been identified in immunoincompetent persons, such as AIDS patients (Koehler, Sanchez, Garrido, Whitfeld, Chen, Berger, Rodriguez-Barradas, LeBoit, and Tapper, 1997).

Bacillary angiomatosis can lead to serious complications in immune-compromised individuals. These complications include skin lesions, anemia, and weight loss. The most serious complication is the growth of non-malignant tumors, which can cause obstruction of airways and damage cardiac valves. This results in the reduction of blood flow in infected persons (Koehler et al., 1997).

The other major disease caused by Bartonella is commonly referred to as CSD. Bartonella henselae has been determined to be the causative agent of this disease (Kosoy et al., 1999). In addition to CSD, B. henselae has been found to present through fever of unknown origin, granulomatous hepatitis, encephalitis, and osteomyelitis (Karem, Dubois, McGill, and Regnery, 1999). While the cat flea is the vector of B. henselae among cats, the domestic cat appears to be a major vector (by scratch or bite) from cat to humans. CSD can be transmitted by being licked, scratched, or bitten by a kitten. This disease is also common in AIDS patients who are immunodeficient (Comer et al., 1996). The disease is characterized by regional lymphadenopathy, fever, and fatigue and is often resistant to antibiotic treatment (Koehler et al., 1997).

Less information is known about B. elizabethae. A single isolate of B. elizabethae was recovered from the blood of a patient with endocarditis (Comer et al., 1996). The spectrum of clinical manifestations related to Bartonella species has expanded since 1990. This expansion is partly due to newly available molecular biological techniques. However, some aspects of Bartonella-related diseases remain unsettled, including epidemiology, physiopathology, and optimum therapy to be administered (Koehler et al., 1997).

The greatest dilemma that exists with Bartonella infection is proper diagnosis. Bacillary angiomatosis, for example, is curable with common antibiotics if it is diagnosed correctly. This disease, though, is often misdiagnosed because the clinical signs, such as the skin lesions, resemble another disease that affects AIDS patients, Kaposi’s sarcoma. There is a significant need for an effective means to test for the individual species of Bartonella (Opavsky, 1997).

 III. Experimental Design and Methods:

A. Bacterial strains and growth conditions:

Bartonella species that will be used in this study include B. quintana, B. elizabethae, B. bacilliformis, B. henselae strain Houston-1 and B. henselae strain Berlin-1. These strains will be employed because of their pathogenic effects on humans. All strains will be obtained from a Centers for Disease Control and Prevention (CDC) collection (Atlanta, GA). Each individual strain will be cultivated directly on two agar plates. Commercially available rabbit blood heart infusion agar will be obtained from Becton Dickinson Microbiology Systems (Cockeysville, MD). Within a humidified CO₂ environment, the agar plates will be incubated at 32 ° C. Based upon the growth characteristics of the different species, the cultures will be allowed to grow for varying lengths of time (Handley and Regnery, 2000).

One agar plate of each Bartonella strain will be obtained for typing by PFGE. Each bacterial culture will be washed from the plates and washed three times in ice-cold phosphate-buffered saline (PBS). We will resuspend the cells in molten, low-melting temperature agarose. We will then allow the molten suspension to solidify into blocks. The size of the blocks will coincide with the thickness of the loading slots of the polyacrylamide gel. The solid blocks will then be incubated in fresh TE of pH 7.6 (10mM Tris-HCl and 1mM EDTA) for 30 min, and then transferred to individual microfuge tubes (Sambrook, Fritsch, and Maniatis, 1989). A volume of 1X HhaI restriction enzyme buffer, which is a medium salt buffer (50-100 mM of NaCl), will be added to each tube. This buffer will be removed after incubating for 30 min. at 4 ° C (Sambrook et al., 1989). Fresh 1X HhaI buffer will then be added to the tubes. HhaI, an infrequently cutting endonuclease, (10 U) will be added to each tube and allowed to incubate for 12-16 hr. at 37 ° C. The enzyme will be obtained from New England Biolabs (Beverly, MA). In order to remove any salt in the restriction buffer, the blocks will then be soaked in TE (pH 7.6) at 4 ° C (Sambrook et al., 1989).

Electrophoresis will be performed using the contour-clamped homogeneous electric field DRII system (Bio-Rad laboratories, Hercules, Calif.) (Riffard, Presti, Vandenesch, Forey, Reyrolle, and Etienne, 1998). The agarose blocks will be directly loaded into the gel slots. If we encounter difficulties loading the blocks into the slots, blocks will be melted by heating at 65 ° C. The molten agarose will then be pipetted into appropriate well (Sambrook et al., 1989). The initial pulse time of 25 sec. for 11 hr. will be utilized for separation, followed by increased pulse times of 35-60 sec. for 11 hr. PFGE markers will be obtained from Boehringer Mannheim for use as size markers (Riffard et al., 1998).

Gels will be stained for 10 min. with ethidium bromide obtained from Bioprobe Systems (Paris, France) (Riffard et al., 1998). The gel will be destained in water for several hours before photographing to facilitate the detection of minor species of DNA (Sambrook et al., 1989). The stained gels will then be photographed using a UV illumination camera available at Washington & Jefferson College, (Washington, PA). An exposure time of 30 sec. will be utilized to obtain good images of the bands (Sambrook et al., 1989).

Bacterial cultures will be washed from the remaining plates with 10 mM Tris buffer (pH 8.0) containing 1mM EDTA. The cells of each strain will be separately vortexed and pelleted by centrifugation. The Easy-DNA kit (Invitrogen, Carlsbad, CA) will be used to extract genomic DNA from the bacterial pellets. The final DNA concentration will be set at 50 ng/m l as assessed by UV spectrophotometry (Handley and Regnery, 2000).

Double digestion of genomic DNA

All enzymes will be obtained from New England Biolabs (Beverly, MA). Genomic DNA (1 m l) will be digested with 10 U of HhaI and 10 U of EagI, SmaI, or XbaI in the appropriate digestion buffer (Figure 1A). The combinations of enzymes include a frequently cutting endonuclease, HhaI, and an infrequently cutting endonuclease (EagI, SmaI, or XbaI). After digesting for 2 hr. at 37 ° C, a ligation mix will be prepared combining T4 DNA ligase, ATP, 10X ligation buffer, the HhaI adapter, either the EagI, SmaI, or XbaI adapter and water (total volume 7.5 m l) (Handley and Regnery, 2000).

The genomic DNA double-digested mixture (12.5 m l) will be mixed with 7.5 m l of the ligation mix. Ligation will occur while the solution is incubated at 16 ° C for 2 hr. The solution will then incubate at 37 ° C for 30 min. to ensure cleavage of religated fragments and at 65 ° C for 20 min. to inactivate the remaining enzymes (Handley and Regnery, 2000).

Figure 1. (A) Bartonella genomic DNA is digested with a combination of enzymes that includes a frequently cutting enzyme (HhaI) and an infrequently cutting enzyme (EagI, SmaI, or XbaI). (B) Adapter pairs ligate to corresponding cleaved ends produced by restriction endonuclease digestion. As shown, the HhaI cleaved end of the genomic fragment (green) complements the HhaI cleaved end of the prepared adapter (green). The EagI cleaved ends are shown in pink. Denaturation occurs when the double-stranded DNA fragment is heated, producing two complementary strands. (C) Four primers (blue), each differing by a single nucleotide (A-adenine, T-thymine, C-cytosine, G-guanine), are constructed to complement each adapter. As shown, the EagI Primer (pink/blue) with the correct complementary nucleotide (yellow) will base-pair to the EagI adapter. Likewise, the HhaI primer (AH1) (blue/green) with the correct complementary nucleotide (yellow) will base-pair to the HhaI adapter. This primer combination will allow the selective amplication of fragments flanked by HhaI and EagI sites only when the primer with the correct nucleotide is present (Handley and Regnery, 2000).

Oligonucleotides will be supplied by Biotechnology Core Facility (CDC, Atlanta, GA). Adapter pairs will be designed to ligate to the cohesive ends produced by restriction endonuclease digestion by EagI, SmaI, XbaI, and HhaI (Figure 1B). Adapters will be prepared with amounts of individual oligonucleotide adapter pairs combined with 1X PCR buffer. By heating at 90 ° C for 5 min., the oligonucleotides will anneal and will then slowly cool for 30 min. (Handley and Regnery, 2000).

Primers will be constructed to complement the SmaI adapter, the EagI adapter, and the AbaI adapter. Four forward reaction primers, that differ by a single nucleotide extension at the 3' terminus, will be designed for each individual adapter (Figure 1C) (Handley and Regnery, 2000).

PCR mixtures will consist of 1 m l restricted-ligated genomic DNA, HotStarTaq DNA polymerase (Qiagen, Valencia, CA), deoxynucleotide triphosphates (200 m l each), and appropriate primers (1.0 m l) in 1X PCR buffer. Four different reactions will be performed for each combination of the restriction endonucleases. For example, if the enzymes HhaI and EagI are used in the double digestion of genomic DNA, then the primer EagI-A (or EagI-T, EagI-G, or EagI-C) will be combined in individual reaction tubes with the HhaI primer, AH1. This method will allow for selective amplification of the fragments flanked by EagI and HhaI sites when nucleotide directly downstream from the EagI site is a thymine (complementary base-pairing) (Figure 1). Four PCR machines (Perkin Elmer, Norwalk, CT) available at Washington and Jefferson College (Washington, PA) will be used for all PCR assays. Conditions will vary according to the specific primer combinations that are used (Handley and Regnery, 2000). For the primer combination EagI-HhaI, the conditions will include initial denaturing at 95 ° C for 15 min., 25 cycles of denaturing at 94 ° C / 1 min., primer annealing at 67 ° C / 30 sec., extension at 72 ° C / 2 min., and final extension at 72 ° C / 10 min. For SmaI-HhaI, the conditions will include initial denaturing at 95 ° C for 15 min., 25 cycles of denaturing at 94 ° C / 1 min., primer annealing at 72 ° C / 2 min., extension at 72 ° C / 2 min., and final extension at 72 ° C / 10 min. For SbaI-HhaI, the conditions will include initial denaturing at 95 ° C for 15 min., 25 cycles of denaturing at 94 ° C / 1 min., primer annealing at 61 ° C / 30 sec., extension at 72 ° C / 2 min., and final extension at 72 ° C / 10 min.

Each amplified reaction mixture will be electrophoresed on a 10% polyacrylamide gel (NOVEX, San Diego, CA). Buffer, 1X Tris-borate-EDTA will be used for the electrophoresis, which will run at 180 V for 1 to 1 ˝ hr. The gel will then be stained for 30 min. with ethidium bromide obtained from Bioprobe Systems (Paris, France). The gel will be destained in water for several hours before photographing to facilitate the detection of minor species of DNA (Sambrook et al., 1989). The stained gels will then be photographed using a UV illumination camera available at Washington & Jefferson College, (Washington, PA). An exposure time of 30 sec. will be utilized to obtain good images of the bands (Sambrook et al., 1989).

4. DNA Sequencing

The amplified DNA fragments will then be sequenced using the dideoxy-mediated chain termination method. Individual bands of DNA, separated by electrophoresis, will be cut from the polyacrylamide gel. DNA extracted from the gel will then be sequenced in the laboratory utilizing chain-terminating dideoxynucleoside triphosphates (ddNTPs) (Sambrook et al., 1989). Alternately, the DNA samples will be sent to the lab of Dr. Russell L. Regnery, Senior Researcher at the Viral and Reckettsial Zoonoses Branch (VRZB) of the CDC to be sequenced using a Perkin Elmer ABI 377 DNA sequencer.

Since the discriminatory power of IRS-PCR has been shown to be equal to that of PFGE, we anticipate that both techniques will yield well-resolved, easily comparable restriction fragment patterns. IRS-PCR, however, possesses several beneficial characteristics. IRS-PCR is less tedious and reproducible because additional sample-processing steps beyond the primary PCR amplification are not required. We also anticipate that IRS-PCR techniques will be less time consuming than PFGE. If, however, we find that the IRS-PCR techniques are not comparable in discriminatory power, time-efficient, or reproducible, we will resort to past methods of detection that have been shown to be effective despite being time-consuming.

Literature Cited

Arvand, M., C. Wendt, T. Regnath, R. Ullrich, and H. Hahn.
      1998. Characterization of Bartonella henselae isolated
      from bacillary angiomatosis lesions in a Human
      Immunodeficiency Virus – infected patient in Germany.
      Clinical Infectious Diseases 26(1):1296-1299.

Comer, J.A., C. Flynn, R.L. Regnery, D. Vlahov, and J.E.
      Childs. 1996. Antibodies to Bartonella species in inner-
      city intravenous drug users in Baltimore, Md. Archives of
      Internal Medicine 156(1):2491-2495.

Handley, S.A., and R.L. Regnery. 2000. Differentiation of
      pathogenic Bartonella species by infrequent restriction
      site PCR. J Clin Microbiol 38(8):3010-3015.

Jensen, W.A., M.Z. Fall, J. Rooney, D.L. Kordick, and E.B.
      Breitschwerdt. 2000. Rapid Identification and
      differentiation of Bartonella species using single-step
      PCR assay. J Clin Microbiol 38(5):1717-1722.

Karem, K.L., K.A. Dubois, S.L. McGill, and R.L. Regnery.
      1999. Characterization of Bartonella henselae – specific
      immunity in BALB/c mice. Immunology 97(1):352-358.

Koehler, J.E., C.A. Glaser, and J.W. Tappero. 1994.
       Rochalimaea henselae infection: a new zoonosis with
       the domestic cat as reservoir. JAMA 271(7):531.

Koehler, J.E., M.A. Sanchez, C.S. Garrido, M.J. Whitfeld,
      F.M. Chen, T.G. Berger, M.C. Todriguez-Barradas,
      B.E. LeBoit, J.W. Tapper. 1997. Molecular
      epidemiology of Bartonella infections in patients with
      bacillary angiomatosis – peliosis. New England J. of
      Medicine 337(26):1876-83.

Kosoy, M.Y., R.L. Regnery, O.I. Kosaya, and J.E. Childs.
      1999. Experimental infection of cotton rats with three
      naturally occurring Bartonella species. J. of Wildlife
      Diseases 35(2):275-284.

Massei, F., M. Massimetti, F. Messina, P. Macchia, G.
      Maggiore. 2000. Bartonella henselae and inflammatory
      bowel disease. Lancet 356(9237):1245-1247.

Matar, G.M., J.E. Koehler, D. Malcolm, M.A. Lambert-Fair,
      J. Tappero, S.B. Hunter, and B. Swaminathan. 1999.
      Identification of Bartonella species directly in clinical
      specimens by PCR-restriction fragment length
      polymorphism analysis of a 16s rRNA gene fragment. J.
      Clin Microbiol 37(12):4045-4047.

Opavsky, M.A. 1997. Cat Scratch Disease: the story
      continues. Canadian J. of Infectious Diseases 8(1):43-49.

Riffard, S., F.L. Presti, F. Vandenesch, F. Forey, M.
      Reyrolle, and J. Etienne. 1998. Comparative analysis of
      infrequent-restriction-site PCR and pulsed-field gel
      electrophoresis for epidemiological typing of Legionella
      pneumophila serogroup 1 strains. J Clin Microbiol
     36(1):161-167.

Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular
      cloning: a laboratory manual. 2^nd Ed. Cold Spring Harbor
      Publishing, New York.

Sander A., M. Posselt, V. Bohm, M. Ruess, and M.
      Altwegg. 1999. Detection of Bartonella henselae DNA
      by two different PCR assays and determination of the
      genotypes of strains involved in histologically defined Cat
      Scratch Disease. J Clin Microbiol 37(4):993-997.

Yoo, J.H., J.H. Choi, W.S. Shin, D.H. Huh, Y.K. Cho,
      K.M.Kim, M.Y. Kim, and M.W. Kang. 1999.
      Application of infrequent-restriction-site PCR to clinical
      isolates of Acinetobacter baumannii and Serratia
      marcescens. J Clin Microbiol 37(10): 3108-3112.

Back to Previous Page

Biology Homepage