Detection of Rochalimaea henselae DNA in Specimens from Cat

2 downloads 0 Views 2MB Size Report
(unfixed) lymph node biopsy specimens and nine aspirates from patients with clinical cat scratch disease. (CSD) to assay for the presence of R. henselae or R.
JOURNAL

OF

CLINICAL MICROBIOLOGY, Apr. 1994,

p.

942-948

Vol. 32, No. 4

0095-1 137/94/$04.00+0 Copyright ©D 1994, American Society for Microbiology

Detection of Rochalimaea henselae DNA in Specimens from Cat Scratch Disease Patients by PCR BURT ANDERSON,' KIMETHA SIMS,1 RUSSELL REGNERY,' LAURA ROBINSON,' MARY JANE SCHMIDT,' SIMIN GORAL,2 CYNTHIA HAGER,2 AND KATHRYN EDWARDS2 Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia 30333,1 and Department of Pediatrics, Vanderbilt University, Nashville, Tennessee 37232-25812 Received 23 September 1993/Returned for modification 18 November 1993/Accepted 12 January 1994

A PCR assay was developed by using degenerate primers that allow amplification of a 414-bp fragment of DNA from the rickettsia-like organisms Rochalimaea henselae and R. quintana. Internal oligonucleotides were used as hybridization probes, permitting rapid differentiation of these two Rochalimaea species. DNAs from 12 different isolates of R. henselae were amplified with the PCR primers, and the resulting 414-bp PCR product hybridized only with the R. henselae-specific probe. DNAs from four different isolates of R. quintana were amplified and produced a PCR product of the same size that hybridized only with the R. quintana-specific probe. DNAs from isolates of R. elizabethae, R. vinsonii, Bartonella bacilliformis, and Afipiafelis failed to amplify the 414-bp fragment in the PCR assay. This two-step assay was applied to DNAs extracted from 16 fresh (unfixed) lymph node biopsy specimens and nine aspirates from patients with clinical cat scratch disease (CSD) to assay for the presence of R. henselae or R. quintana DNA in these samples. Twenty-one (84%) of 25 lymph node samples from CSD patients were positive for R. henselae, while none were positive for R. quintana. The characteristic 414-bp fragment was not amplified from eight lymph node tissue samples from non-CSD cases. These results provide evidence that R. henselae, and not R. quintana, plays the central role in the etiology of CSD.

Identification of the etiologic agent of cat scratch disease (CSD) has been the source of a lengthy controversy. A variety of agents have been associated with CSD, including an organism that was isolated from the lymph nodes of 10 of 19 patients with CSD (9). This organism has subsequently been identified and classified as a new bacterium, Afipia felis (5). However, more recently the rickettsia-like organism Rochalimaea henselae has been isolated from the lymph nodes of two patients with CSD (7) and anti-Rochalimaea antibodies have been detected in the sera of 84 to 88% of patients with clinical CSD (19, 27). Additionally, the R. henselae 16S rRNA gene sequence has been detected in CSD skin test antigens that have been used for diagnosis of this disease for many years (1). Thus, there are now several lines of evidence which support the notion that Rochalimaea spp. play a major role in the etiology of CSD. Both R. henselae and R. quintana have been isolated or detected in humans and have been associated with various clinical syndromes, including bacillary angiomatosis (12, 21), and persistent fever with bacteremia in immunocompetent and immunosuppressed individuals (14, 18, 22, 25). Despite the association of both R. henselae and R. quintana with disease, R. quintana has not been firmly linked to CSD. To assess the role of both R. henselae and R. quintana in CSD, we have developed a PCR assay that specifically allows detection of both of these organisms in clinical samples. The resulting PCR product was used as a target for species-specific oligonucleotide hybridization probes that differentiate R. henselae from R. quintana. These techniques were applied to a set of samples consisting of

fresh lymph node tissue excised from CSD patients or lymph node aspirates from CSD patients to evaluate the PCR assay for rapid diagnosis and as a tool for the study of Rochalimaeaassociated diseases. MATERIALS AND METHODS Bacterial strains. All of the strains of bacteria used to evaluate the specificity of the PCR and hybridization assay are listed in Table 1. Rochalimaea spp. were grown on heart infusion agar plates supplemented with 5% defibrinated rabbit blood (HIA-RB; BBL Microbiology Systems, Cockeysville, Md.) incubated for 3 to 5 days at 34°C in the presence of 5% CO2. Bartonella bacilliformis was cultivated on HIA-RB for 6 to 8 days at 28°C without supplemental CO2. A. felis was grown on charcoal-yeast extract agar plates (Carr-Scarborough Microbiologicals, Decatur, Ga.) for 2 to 3 days at 32°C without

Co,.

Clinical samples. Twenty-five samples from patients clinically diagnosed with CSD were used to evaluate the PCR assay (Table 2). All CSD cases were clinically diagnosed by a physician and had regional lymphadenopathy and cat contact in the absence of another obvious diagnosis. All of the patients whose samples were used met this definition except patient 16 (Table 2), who had no known history of cat contact. Sixteen of these samples were fresh lymph node biopsy specimens, and nine were lymph node aspirates. In addition, five lymph node aspirates from non-CSD patients, from whom other organisms were isolated, were included as negative controls. Likewise, three lymph node biopsies from non-CSD patients were used as additional negative controls. Serologic analysis was performed on some patients (when serum was available) by the indirect fluorescent-antibody test (19). DNA extraction. DNA was extracted from bacterial cells, lymph node tissue, or lymph node aspirates by using modifi-

* Corresponding author. Mailing address: Viral and Rickettsial Zoonoses Branch, Bldg. 15, Room 1611, Mailstop G13, Centers for Disease Control and Prevention, Atlanta, GA 30333. Phone: (404) 639-1082. Fax: (404) 639-3163.

942

PCR FOR ROCHALIMAEA HENSELAE

VOL. 32, 1994

943

TABLE 1. Isolates whose DNAs were used for specificity testing of the PCR primers and oligonucleotide probes Hybridization with probe: Bacterial strain

Source (reference)

PCR result

RHI

R. henselae Houston-lT R. henselae San Ant-I R. henselae San Ant-2 R. henselae San Ant-3 R. henselae San Diego-2 R henselae OK88-64 R. henselae OK88-712 R. henselae OK89-674 R. henselae OK89-675 R. henselae OK90-615 R. henselae OK90-782 R henselae CAL-1 R. quintana FullerT R. quintana OK90-268 R. quintana SH-PERM R. quintana D-PERM R. elizabethae F9251T R. vinsonii RVT B. bacilliformis KC584 A. felis BVT a

b

HIV+a patient (18) HIV- patient (14) CSD patient (7) CSD patient (7) San Diego, HIV+ patient HIV+ patient (25) HIV- patient (25) HIV- patient(25) HIV- patient (25) HIV- patient (25) HIV+ patient (25) San Diego, HIV+ patient ATCC VR358 HIV+ (25) Russia

Russia Heart valve (6) ATCC VR152 ATCC 35686 ATCC 53690 (5)

+ +

+ +

+ + +

+ + +

+

+

+ + + + + +

+ + + + + +

+ + + +

-

_b

_

-

-

HIV+, human immunodeficiency virus positive. The 414-bp PCR product was not observed. A larger product of approximately 1,300 bp (Fig. 2) that failed to hybridize with either the

amplified from R. elizabethae (Fig. 3).

cations of a procedure previously described (4). Briefly, bacterial growth harvested from approximately one-eighth of a standard-size (85-mm-diameter) HIA-RB plate was suspended in 300 p.l of PCR diluent (10 mM Tris, 10 mM NaCl, 1 mM EDTA, pH 8.0). For lymph node tissue, samples (approximately 100 mg) were dispersed with a disposable homogenizer in minimal essential medium (0.5 ml), and 50 to 100 p.l of this homogenate was diluted to 300 p.l with PCR diluent. Lymph node aspirates (50 to 100 ,u1) were suspended and diluted to 300 p.l with PCR diluent. The samples were then made to contain 1.0% sodium dodecyl sulfate (SDS), proteinase K was added to a final concentration of 100 ng/ll, and the samples were incubated for 2 h at 55°C. After incubation, the lysates were extracted three to four times with a 50:50 mixture of buffer-saturated phenol and chloroform-isoamyl alcohol (24: 1). The resulting aqueous supernatant was diluted to 2.0 ml with PCR diluent, filtered through a Centricon 30 filter (Amicon, Danvers, Mass.), and washed twice more with 2.0-ml aliquots of PCR diluent. The subsequent filter retentate (average volume, 40 p.l) was collected and used as a template for the PCR. For every DNA extraction run, a reagent blank was processed exactly as described above to ensure that all of the extraction buffers and reagents were not contaminated with Rochalimaea DNA. PCR primer and hybridization probe design. A library of R henselae DNA was constructed in the vector lambda ZAPII (2). A clone expressing a 60-kDa antigen reactive with rabbit anti-R henselae serum has been isolated, and the gene has been sequenced (3) (GenBank accession no. L20127). The deduced amino acid sequence was shown to have 37% sequence homology (over the entire sequence) with the htrA locus described from Escherichia coli (13). Primer pair CAT1 5'-GA7ITCAATTGGI'TTGAA(G and A)GAGGCT-3' and CAT2 5'-TCACATCACCAGG(A and G)CGTATTC-3' (Fig. 1), which defines a 414-bp fragment from both R henselae and R quintana, was used for PCR amplification. Twenty-base-pair oligonucleotide probes RH1 and RQ1 (Fig. 1) were used as species-specific hybridization probes.

RQ1

+ + + +

RHI or the RQ1 probe was

PCR assays. DNAs prepared from bacteria, fresh lymph node tissue, or lymph node aspirates were used as templates for the PCR assays. Five-microliter portions of the template DNAs (undiluted and diluted 1:10) extracted from the clinical samples were used for the PCR assays. The approximate concentration of DNA extracted from each bacterial isolate was determined by agarose gel electrophoresis next to known quantities of standard DNA. Diluted bacterial DNA (approximately 1 ng) was used for initial determination of primer specificity. For subsequent PCR on clinical samples, 10 pg (in 10 p.l) of DNA extracted from either R. henselae or R. quintana was used as a positive control, and the DNA extraction blank and water (10 p.l of each) were used as negative controls. The GeneAmp reagent kit (Perkin-Elmer Cetus, Norwalk, Conn.) was used for all PCR assays. Degenerate primer pair CAT1 and CAT2 was used to prime the polymerization reactions. Amplification was accomplished by predenaturing for 5 min at 94°C followed by 35 cycles of 94°C for 30 s, 50°C for 60 s, and 70°C for 45 s in a 9600 thermal cycler (Perkin Elmer). Ten microliters from each PCR assay was electrophoresed through a 1.2% agarose gel, stained with ethidium bromide, and photographed. The presence of a 414-bp band was considered positive. Each sample of DNA extracted from the clinical specimens was also amplified with primer pair GHPCR1 and GHPCR2 (26). This primer pair amplifies a 422-bp fragment from a conserved region of the human growth hormone gene and served as a positive control for successful extraction of amplifiable DNA. DNA extracts from clinical samples that failed to be amplified with primer pair GHPCR1 and GHPCR2 were excluded from further study. Dot blot hybridizations. To confirm the identity of the PCR products and to allow differentiation of products amplified from R. henselae and R. quintana templates, a dot blot hybridization assay was used. Oligonucleotide probes RHi and RQ1 were used for this purpose. RHi and RQ1 were nonisotopically labeled by transfer of a digoxigenin-ddUTP nucleotide to the 3' end of each oligonucleotide by means of terminal transferase with a Genius 5 labeling kit (Boehringer Mann-

944

ANDERSON ET AL.

J. CLIN. MICROBIOL.

TABLE 2. Information and PCR-dot blot results for CSD patients Hybridization

Patient no.

State

Sample

PCRofdetection CSD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Mass. Mass. Mo. Fla. Fla. Ohio S.C. N.J. Va. N.J. N.J. Pa. Mass. W.Va. Maine N.C. Wash. Mass. Ga. Tenn. Tenn. Tenn. Fla. Tenn. Va. Tenn. Tenn.

Aspirate Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Biopsy Aspirate Biopsy Biopsy Biopsy Biopsy Aspirate Biopsy Aspirate Aspirate Aspirate Aspirate Aspirate Aspirate Aspirate Aspirate

+ + + + + + +

+ + + + + + +

-

-

+ + + + + + + +

+ + + + + + + +

+ + + + + +

+ + + + + +

-C' -C

-

with probe: RHI RQO

Serologic result"

-

+ + + + + + + + + + + + + + + + + +

-

NDb ND ND ND ND

cSerologic analysis was performed by the indirect fluorescent-antibody test as

previously described (19). An anti-Rochalimaea titer of 1:64 or higher was

heim, Indianapolis, Ind.). For the dot blot hybridization assays, 5 RI of each PCR product was denatured for 10 min by addition of 0.5 RI of 4 M NaOH containing 100 mM EDTA. One-microliter aliquots were spotted onto each of two nylon membranes (Boehringer Mannheim), and the DNA was crosslinked to the nylon by UV irradiation (Stratalinker; Stratagene, La Jolla, Calif.). The nylon membranes were then blocked for 1 h at 62°C by using standard prehybridization solution from the Genius 7 luminescence detection kit (Boehringer Mannheim). The standard hybridization solution was 5 x SSC

R.

1eamelee

R.

quintana

R.

-lizab.tha

R.

vienonii

RHI/RQ1

CAT1 G

_AT

CT

GGTGCGTTAATTACCGATCC

. . ............... A T...... ....T..G..

(C)

FIG.

2.

Agarose gel electrophoresis

of PCR

products

obtained with

primers CAT_ and CAT2 and template DNAs from different bacteria.

purified from each of the following organisms was used as the for amplification: lane B, R. henselae; lane C, R. quintana; lane D, R. elizabethae; lane E, R. vinsonii; lane F, B. baciiformis; lane G, A. felis; lane H, no DNA (negative control). Molecular size standards, 4X174 phage DNA digested with Haelll (1,353, 1,078, 872, 603, 310, 281 plus 271, 234, 194, 118, and 72 bp, from top to bottom), DNA

template

are

shown in lane A.

(1 x SSC is 0.15 M NaCl and 0.015 M sodium citrate) containing 0.1% N-laurylsarcosine, 0.02% SDS, and 1.0% blocking reagent (Boehringer Mannheim). Hybridization was then performed at 62°C (Ta, - 8°C for both probes) for 1 h in fresh prehybridization solution containing either probe RH1 or RQ1 at a concentration of 2 pmol/ml. The hybridized membrane was washed twice for 15 min each time in 2 x SSC containing 0.1% SDS at room temperature and then washed twice for 15 min each time at 52°C in 0.5 x SSC-0.1% SDS. The hybridized filter was then blocked, reacted with alkaline phosphatase-conjugated antibody, washed, and soaked in Lumigen PPD chemiluminescent substrate in accordance with the manufacturer's directions (Genius 7 kit; Boehringer Mannheim). The resulting filter was exposed to X-ray film for 5 to 20 min, and the film was developed.

GAATACGTCCTGGTGATGTGA

T.. ... .......

C. .AA... ...T.NGG.G ..T

ND

A........ A....AC...T .....CT

ND

G. .A . ..C

CAT2

bp--

+ + +

considered positive. b ND, not done. ' Representative negative control (non-+ case from which other bacteria were isolated).

Species

414

.......C

FIG. 1. Nucleotide sequence alignment for the three regions of the antigen gene corresponding to primers CAT1 and CAT2 and oligonucleotide probes RH1 and RQ1. The antigen gene sequences were aligned for maximal homology by using only the portion corresponding to the primer and probe sequences. The sequences for R. henselae and corresponding base substitutions (from the R. henselae sequence) for other species are shown, and conserved positions are indicated with periods. The complement of PCR primer CAT2 [CAT2 (C)] is shown. ND, not done.

RESULTS Specificity of the PCR assay. Approximately 1 ng of DNA from each of the bacteria listed in Table 1 was used as a template for PCR assays with the CAT1 and CAT2 primer pair to evaluate specificity. Under the optimal conditions described in Materials and Methods and with purified template DNA, all 12 R. henselae isolates and all 4 R. quintana isolates yielded the predicted 414-bp fragment of amplified DNA (Table 1). Figure 2 shows the PCR products obtained from the representative type strains of R. henselae (lane B), R. quintana (lane C), and R. elizabethae (lane D). No amplification products were observed for R. vinsonii (lane E), B. bacilliformis (lane F), and A. felis (lane G). R. elizabethae was amplified with this primer pair,

VOL. 32, 1994

A

PCR FOR ROCHALIMAEA HENSELAE

2 3

4 5

A *.a*

a .

B .

0

1

.

0

6

7

8

.

a

C

B

945

414 bp-

1

2

3 4 5

6

7

8

A A B CDE FG H I J K L

B C

*

FIG. 3. Dot blot hybridization of PCR products amplified from different bacteria with digoxigenin-labeled oligonucleotide probes RHI (A) and RQl (B). PCR products amplified by using the following bacterial DNAs as templates were spotted onto the following locations on a nylon membrane for hybridization: Al, R. henselae Houston-1; A2, R. henselae San Diego-2; A3, R. henselae San Ant-2; A4, R. henselae OK89-674; A5, R. henselae OK87-66; A6, R. henselae OK90615; A7, R. henselae OK88-64; A8, R. henselae San Ant-1; Bi, R. henselae Cal-1; B2, R henselae OK88-712; B3, R. henselae OK89-675; B4, R. henselae OK90-782; B5, A. felis BV; B6, B. bacilliformis KC584; B7, R. elizabethae F9251; B8, R. vinsonii RV; Ci, R. quintana Fuller; C2, R. quintana SH-Perm; C3, R. quintana OK90-268; C4, R quintana D-Perm. Well C5 contained the products of a no-DNA control amplification.

but the product was much larger (approximately 1,300 bp) than the 414 bp predicted for R henselae and R. quintana. Thus, the 414-bp PCR product appears to be specific for R. henselae and R. quintana. A PCR product of approximately 50 to 60 bp was occasionally observed in the no-DNA control (lane H) and presumably corresponds to a primer dimer. Specificity of the dot blot hybridization assay. To confirm the identity of the PCR product and to differentiate between PCR products amplified from R. henselae and R. quintana, a dot blot hybridization assay was performed on the PCR products amplified from the bacteria listed in Table 1. PCR products amplified from all 12 isolates of R. henselae hybridized with probe RH1 (Fig. 3A, dots Al to B4) and not to probe RQ1 (Fig. 3B, dots Al to B4). PCR products from all four isolates of R quintana hybridized only with probe RQi (Fig. 3B, dots Ci to C4). Neither probe RH1 nor RQ1 hybridized to the PCR products from R. elizabethae (Fig. 3, dot B7), R.

FIG. 4. Agarose gel electrophoresis of representative PCR products amplified from DNAs extracted from lymph nodes of CSD patients. The PCR products were amplified from DNAs extracted from the lymph node tissue of the following patients (Table 2): lane B, 1; lane C, 2; lane D, 3; lane E, 11; lane F, 8 (undiluted DNA); lane G, 8 (DNA diluted 1:10); lane H, 25; lane I, 27; lane J, no-DNA control. The amplification products from R. quintana (lane K) and R. henselae (lane L) purified DNAs are shown as positive controls. Molecular size standards, XX174 phage DNA digested with HaeIII (1,353, 1,078, 872, 603, 310, 281 plus 271, 234, 194, 118, and 72 bp, from top to bottom), are shown in lane A.

vinsonii (Fig. 3, dot B8), B. bacilliformis (Fig. 3, dot B6), and A. felis (Fig. 3, dot B5). Thus, the dot blot hybridization assay allows differentiation between PCR products amplified from R. henselae and R. quintana. Application of the PCR and dot blot assays to clinical samples. To evaluate the PCR and dot blot assays for detection of R. henselae and R. quintana in clinical samples, we applied these techniques to 16 samples of fresh lymph node tissue and nine aspirates from CSD cases. Twenty-one of 25 samples produced the 414-bp product that is characteristic of R. henselae or R. quintana (Table 2). Figure 4 shows representative PCR products obtained from lymph node biopsy samples (lanes C to E and G) and lymph node aspirates (lanes B and H) of patients suspected of having CSD. Two samples produced the characteristic 414-bp band only when the template DNA was diluted 1:10 before amplification. Typical of these samples was no. 9, the amplification of which appeared to be inhibited by large amounts of leukocyte DNA, as indicated by a blur on an agarose gel (Fig. 4, lane F). When the sample containing the template DNA was diluted 1:10 prior to amplification, the 414-bp band was clearly produced (Fig. 4, lane G). The characteristic 414-bp fragment was not amplified from any of the eight lymph node tissue samples from non-CSD cases (typified by sample 27 [Fig. 4, lane I]) or from DNA extraction blanks (lane J). To confirm the identity of the PCR products amplified from

946

J. CLIN. MICROBIOL.

ANDERSON ET AL.

However, these same primers have been used by others without success to detect Rochalimaea spp. in lymph nodes from CSD patients (10). For these reasons, a different pair of 9 A primers was used for the PCR assay described in this report. 0 For the PCR primers and hybridization probes described in this report, we chose sequences from the putative R. henselae * * B * a analog of the htrA gene of E. coli. The htrA gene product of E. a a * coli belongs to a group of heat shock proteins that confer the ability to grow at elevated temperatures. E. coli htrA mutants c *Sdo not grow at elevated temperatures (13). A gene encoding an * * antigenic protein of R. henselae has been sequenced, and the resulting deduced amino acid sequence shares 37% identity 10 with the E. coli HtrA protein over the entire coding region (3). 8 9 2 3 4 5 6 7 1 A partial nucleotide sequence (approximately 300 nucleotides) of the same gene from the other three Rochalimaea species was A obtained by using conserved PCR primers and was found to be 85 to 93% conserved with the R. henselae sequence. No evidence of this gene was found in B. bacilliformis, an organism B phylogenetically closely related to Rochalimaea spp. (16, 18, 20). This is interesting, since B. bacilliformis does not grow at elevated temperatures, a trait which may be due in part to the C lack of a functional htrA gene product. The results of the PCR with degenerate primers CAT1 and CAT2 indicate that these primers are well conserved within the isolates of R. henselae F IG. 5. Dot blot hybridization of PCR products amplified from and R. quintana tested. Likewise, the oligonucleotide probes lymiph node extracts of suspect CSD cases. Panel A was hybridized with (RH1 and RQ1), while being species specific, appear to be well oligronucleotide probe RH1, and panel B was hybridized with probe conserved within the species. RQ!1. PCR products amplified from DNAs extracted from lymph node By using PCR primers CAT1 and CAT2, the PCR assay was tissiues of the following patients (listed in Table 2) are shown at the applied to 16 fresh lymph node biopsy specimens and nine folkowing positions: A4, 2; A5, 1; A6, 3; A7, 7; A8, 4; A9, 5; A10, 6; Bi, 18; B2, 10; B3, 1 ; B4, 12; B5, 9; B6, 13; B7, 15; B8, 16; B9, 17; BlO, aspirates from patients clinically diagnosed with CSD, and 21 8; C'1, 14; C2, 19; C3, 20; C4, 21; C5, 25; C6, 26; C7, 27; C8, 23; C9, 24; (84%) were amplified to produce the 414-bp fragment that is cio), 22. The PCR products from a DNA extraction blank (Al), R. characteristic of either R. henselae or R. quintana. The resulthen selae (A2), and R. quintana (A3) are also shown. ing PCR products were hybridized to species-specific oligonucleotide probes RQ1 and RH1, and all 21 of the PCR-positive samples hybridized to probe RH1. Thus, all of the PCRpositive samples contained only R. henselae DNA and no the clinical samples and to sort infections caused by R. henselae evidence of R. quintana-infected lymph nodes was found. was hybridization blot a dot by R. quintana, m those caused fro] These results suggest that unlike bacillary angiomatosis and performed with species-specific probes RHi and RQ1. The other opportunistic infections seen among AIDS patients that PC]R products from all 21 samples that were amplified to may be caused by either R. henselae or R. quintana, CSD pro duce the characteristic 414-bp fragment hybridized with R. appears to be caused primarily (or perhaps exclusively) by R selae-specific probe RHi (Fig. SA). Conversely, none of henselae. the se samples hybridized with R. quintana-specific probe RQ1 Fresh frozen tissue samples from both lymph node biopsy (Fi1 g. 5B). Thus, all of the PCR-positive samples studied here, specimens and lymph node aspirates were suitable. Greater fro]m patients in 11 different states suspected of having CSD success was obtained by using aspirates (nine of nine [100%] (Tatble 2), appear to be associated with R. henselae and not R. were positive) than by using biopsies (12 of 16 [75%] were qui)ntana. None of the samples that failed to be amplified positive), probably because of the inherent difference between yielIding the 414-bp fragment as determined by agarose gel nodes which are fluctuant and thus can be aspirated and those ele etrophoresis hybridized with either probe. which are not. Difficulty was encountered in standardizing the or aspieither lymph both nodesRNA amount of DNA extracted DISCUSSION and a totalfrom procedure, lysate rates. Since we utilized iJiven the controversy surrounding the etiology of CSD and DNA were obtained and measurement of lysate absorbance would be a poor indicator of the DNA concentration. Accordassociation of both R. henselae and R. quintana with human ingly, we used each template sample undiluted and at a 1:10 ease, we developed a method to detect each of these dilution for amplification. Some samples were positive only orgranisms directly in lymph node tissue from CSD patients. when the starting template was diluted prior to amplification. e PCR assay described here was designed to amplify DNAs These samples typically contained large amounts of cellular froim both R. henselae and R. quintana and was coupled with a DNA that may have inhibited amplification (Fig. 4, lane F) and spe cies-specific, dot blot hybridization assay for differentiation upon dilution yielded the 414-bp band (Fig. 4, lane G). Only tthese two species. Although PCR primers derived from the DNA extracts that were amplified to produce the 422-bp 16S rRNA gene sequence that allow detection of Rochalimaea fragment of the human growth hormone gene were used for in our these primers study 4A have been described (21), DA the specific PCR assay. Although this control serves to allow re not specific, amplifying target DNAs from all four we] unamplifiable extracts to be discarded, it does not require the Rotchalimaea species and B. bacilliformis. This lack of primer sensitivity needed to amplify small amounts of DNA from spe cificity was also noted in the report of Tappero et al. (23), bacterial pathogens. The DNA extraction procedure described o used these same primers successfully to amplify DNAs here is rather lengthy and may be impractical for use in a fro m patients with bacillary angiomatosis and bacillary peliosis.

A

B

hen

the disi

Th

of

wh

1

2

3

4

5

6

7

8

9

10

VOL. 32, 1994

clinical setting; the development or adaptation of a simpler method of DNA extraction from lymph node tissue will encourage use of the assay described here in nonresearch laboratories. The 84% PCR assay positivity among samples from patients suspected of having CSD described here is virtually identical to the 84 and 88% positivity observed by serologic means in two separate groups of samples from patients suspected of having CSD (19, 27). Twenty-two of the samples from CSD cases tested in this report by PCR were also tested serologically (Table 2). Twenty-one (95%) of these had an immunofluorescent-antibody titer of 1:64 or greater. Thus, three samples collected from seropositive individuals were negative by the PCR assay. This apparent discrepancy may be due in part to the lack of intact organisms in the lymph nodes from patients in the later stages of CSD. In fact, Gerber et al. have postulated that the lingering cell-mediated immune response and resulting granulomatous reaction, rather than bacterial invasion, may be the major pathogenic mechanism of CSD (11). Alternatively, PCR inhibitors that preclude the attainment of optimal assay sensitivity may be present in lymph node tissue. The PCR assay offers the advantage of early diagnosis, since it is not dependent on the mounting of a detectable humoral immune response by the patient. In addition, the PCR assay differentiates R. henselae from R. quintana infections. Available serologic tests (17, 19) demonstrate antibodies that are cross-reactive between R. henselae and R. quintana. Both genetic and biochemical methods for differentiation of R. henselae from R. quintana isolates have been reported (8, 18, 24); however, these methods are dependent on the timeconsuming procedure of isolation, which may require 3 to 4 weeks or longer. A rapid and specific test for CSD would afford clinicians an alternative to culture or serologic analysis for laboratory confirmation of the diagnosis, thereby permitting them to rule out malignancies such as lymphoma and to consider antibiotic therapy. Although the efficacy of antibiotics for the treatment of CSD remains uncertain, successful treatment with four antibiotics (rifampin, ciprofloxacin, trimethoprim-sulfamethoxazole, and gentamicin sulfate) has been reported (15), and in vitro, R henselae is susceptible to many common antibiotics (7). In summary, the results of the present study are significant for several reasons. (i) A PCR-dot blot assay that allows detection of, and differentiation between, R. henselae and R quintana directly in clinical samples is described. This assay should be a rapid and suitable alternative to either isolation or serologic analysis for diagnosis of CSD. (ii) Despite recent evidence of R. quintana infections among immunocompromised patients, no evidence of R. quintana-mediated CSD was found. (iii) The detection of R. henselae in 84% of the tested lymph node samples from patients suspected of having CSD supports the notion of a major role of R. henselae in the etiology of CSD. REFERENCES 1. Anderson, B., C. Kelly, R. Threlkel, and K. Edwards. 1993. Detection of Rochalimaea henselae in cat-scratch disease skin test antigens. J. Infect. Dis. 168:1034-1036. 2. Anderson, B., and G. McDonald. 1993. Construction of DNA libraries of A-T rich organisms using EcoRI star activity. Anal. Biochem. 211:325-327. 3. Anderson, B., K. Sims, D. Jones, W. Dewitt, and W. Bibb. 1993. Molecular cloning of Rochalimaea henselae antigens, abstr. D-90, p. 111. Program Abstr. 93rd Annu. Meet. Am. Soc. Microbiol. 1993, Washington D.C.

PCR FOR ROCHALIMAEA HENSELAE

947

4. Anderson, B. E., J. E. Dawson, D. C. Jones, and K. H. Wilson. 1991. Ehrlichia chaffeensis, a new species associated with human ehrlichiosis. J. Clin. Microbiol. 29:2838-2842. 5. Brenner, D. J., D. G. Hollis, C. W. Moss, C. K. English, G. S. Hall, J. V. Vincent, J. Radosevic, K. A. Birkness, W. F. Bibb, F. D. Quinn, B. Swaminathan, R. E. Weaver, M. W. Reeves, S. P. O'Connor, P. S. Hayes, F. C. Tenover, A. G. Steigerwalt, B. A. Perkins, M. I. Daneshvar, B. C. Hill, J. A. Washington, T. C. Woods, S. B. Hunter, T. L. Hadfield, G. W. Ajello, A. F. Kaufnann, D. J. Wear, and J. D. Wenger. 1991. Proposal ofAfipia, gen. nov., with Afipia felis sp. nov. (formerly the cat scratch disease bacillus), Afipia clevelandensis sp. nov. (formerly the Clevland Clinic Foundation strain), Afipia broomeae sp. nov., and three unnamed genospecies. J. Clin. Microbiol. 29:2450-2460. 6. Daly, J. S., M. G. Worthington, D. J. Brenner, C. W. Moss, D. G. Hollis, R. S. Weyant, A. G. Steigerwalt, R. E. Weaver, M. I. Daneshvar, and S. P. O'Connor. 1993. Rochalimaea elizabethae sp. nov. isolated from a patient with endocarditis. J. Clin. Microbiol. 31:872-881. 7. Dolan, M. J., M. T. Wong, R. L. Regnery, J. H. Jorgensen, M. Garcia, J. Peters, and D. Drehner. 1993. Syndrome of Rochalimaea henselae suggesting cat scratch disease. Ann. Intern. Med. 118:331-336. 8. Drancourt, M., and D. Raoult. 1992. Abstr. Tenth Sesquiannual Meeting of the American Society for Rickettsiology and Rickettsial Diseases, Hamilton, Mont. 9. English, C. K., D. J. Wear, A. M. Margileth, C. R. Lissner, and G. P. Walsh. 1988. Cat-scratch disease: isolation and culture of the bacterial agent. JAMA 259:1347-1352. 10. Erler, B. S., A. M. Jiminez, M. L. Gedebou, J. W. Said, and W. S. Nichols. 1993. Absence of Rochalimaea henselae sequences in cat scratch disease lymph nodes using a polymerase chain reaction assay. Mod. Pathol. 6:105A. 11. Gerber, M. A., P. Rapacz, S. S. Kalter, and M. Ballow. 1986. Cell-mediated immunity in cat-scratch disease. J. Allergy Clin. Immunol. 78:887-890. 12. Koehler, J. E., F. D. Quinn, T. G. Berger, P. E. LeBoit, and J. W. Tappero. 1992. Isolation of Rochalimaea species from cutaneous and osseous lesions of bacillary angiomatosis. N. Engl. J. Med. 327:1625-1631. 13. Lipinska, B., S. Sharma, and C. Georgopoulos. 1988. Sequence analysis and regulation of the htrA gene of Escherichia coli: a CF32-independent mechanism of heat-inducible transcription. Nucleic Acids Res. 16:10053-10067. 14. Lucey, D., M. J. Dolan, C. W. Moss, M. Garcia, D. G. Hollis, S. Wenger, G. Morgan, R. Almeida, D. Leong, K. S. Greisen, D. F. Welch, and L. N. Slater. 1992. Relapsing illness due to Rochalimaea henselae in immunocompetent hosts: implication for therapy and new epidemiological associations. Clin. Infect. Dis. 14:683688. 15. Margileth, A. M. 1992. Antibiotic therapy for cat-scratch disease: clinical study of therapeutic outcome in 268 patients and a review of the literature. Pediatr. Infect. Dis. 11:474-478. 16. O'Connor, S. P., M. Dorsch, A. G. Steigerwalt, D. J. Brenner, and E. Stackebrandt. 1991. 16S rRNA sequences of Bartonella bacilliformis and cat scratch disease bacillus reveal phylogenetic relationships with the alpha-2 subgroup of the class Proteobacteria. J. Clin. Microbiol. 29:2144-2150. 17. Patnaik, M., W. A. Schwartzman, N. E. Barka, and J. B. Peter. 1992. Lancet 340:971. (Letter.) 18. Regnery, R. L., B. E. Anderson, J. E. Clarridge III, M. C. Rodriquez-Barradas, D. C. Jones, and J. H. Carr. 1992. Characterization of a novel Rochalimaea species, R henselae sp. nov., isolated from blood of a febrile, human immunodeficiency viruspositive patient. J. Clin. Microbiol. 30:265-274. 19. Regnery, R. L., J. G. Olson, B. A. Perkins, and W. Bibb. 1992. Serological response to "Rochalimaea henselae" antigen in suspected cat-scratch disease. Lancet 339:1443-1445. 20. Relman, D. A., P. W. Lepp, K. N. Sadler, and T. M. Schmidt. 1992. Phylogenetic relationships among the agent of bacillary angiomatosis, Bartonella bacillifornis, and other alpha-proteobacteria. Mol. Microbiol. 6:1801-1807. 21. Relman, D. A., J. S. Loutit, T. M. Schmidt, S. Falkow, and L. S.

948

ANDERSON ET AL. Tompkins. 1990. The agent of bacillary angiomatosis. N. Engl. J.

Med. 323:1573-1580. 22. Slater, L. N., D. F. Welch, D. Hensel, and D. W. Coody. 1990. A newly recognized fastidious gram-negative pathogen as a cause of fever and bacteremia. N. Engl. J. Med. 323:1587-1593. 23. Tappero, J. W., J. Mohle-Boetani, J. E. Koehler, B. Swaminathan, T. G. Berger, P. E. LeBoit, L. L. Smith, J. D. Wenger, R. W. Pinner, C. A. Kemper, and A. L. Reingold. 1993. The epidemiology of bacillary angiomatosig and bacillary peliosis. JAMA 269:770775. 24. Welch, D. F., D. M. Hensel, D. A. Pickett, V. H. San Joaquin, A. Robinson, and L. N. Slater. 1993. Bacteremia due to Rochalimaea henselae in a child: practical identification of isolates in the clinical

J. CLIN. MICROBIOL. laboratory. J. Clin. Microbiol. 31:2381-2386. 25. Welch, D. F., D. A. Pickett, L. N. Slater, A. G. Steigerwalt, and D. J. Brenner. 1992. Rochalimaea henselae sp. nov., a cause of septicemia, bacillary angiomatosis, and parenchymal bacillary peliosis. J. Clin. Microbiol. 30:275-280. 26. Wu, D. Y., L. Ugozzoli, B. K. Pal, and R. B. Wallace. 1989. Allele-specific enzymatic amplification of ,-globin genomic DNA for diagnosis of sickle cell anemia. Proc. Natl. Acad. Sci. USA 86:2757-2760. 27. Zangwill, K. M., D. H. Hamilton, B. A. Perkins, R. L. Regnery, B. D. Plikaytis, J. L. Hadler, M. L. Cartter, and J. D. Wenger. 1993. Cat scratch disease in Connecticut. N. Engl. J. Med. 329:8-13.