581

Association of Mycoplasma and Human Immunodeficiency Virus Infection: Detection of Amplified Mycoplasma fermentans DNA in Blood Richard E. Hawkins, Leland S. Rickman, Sten H. Vermund, and Mitchell Carl

Division of Infectious Diseases. Department of Internal Medicine. National Naval Medical Center. Infectious Diseases Department. Accelerated Product Development Program. Naval Medical Research Institute. and Epidemiology Branch. Division ofAIDS. National Institute of Allergy and Infectious Diseases. National Institutes of Health. Bethesda. Maryland; Division of Infectious Diseases. University ofCalifornia. San Diego

Mycoplasma fermentans has recently received attention as a potential pathogen in humans and in nonhuman primates. Lo et al. [I] initially identified a virus-like infectious agent in tissues of patients dying of AIDS; however, this agent was subsequently cultured in cell-free medium and classified as a Mycoplasma species. Tentatively named Mycoplasma incognitus because of an apparent ability to elude immune reaction, subsequent genetic and serologic comparison showed that M. incognitus and M. fermentans were speciesrelated [2]. While there is evidence that M. fermentans infection occurs in patients dying of AIDS, its association as an opportunistic infection or as a cofactor in disease progression is undefined. Until recently there were few data showing infection with mycoplasma in living AIDS patients. Dawson et al. [3], from the Armed Forces Institute of Pathology (AFIP, Washington, DC), identified M. fermentans DNA sequences in 10 of 43 urine samples from 40 human immunodeficiency virus

Received 8 August 1991; revised 31 October 1991. Presented in part: 91 st general meeting. American Society for Microbiology. May 1991. Dallas. Informed consent was obtained from study subjects. The study was reviewed by the Committee for the Protection of Human Subjects. National Naval Medical Center. for compliance with guidelines for obtaining informed consent. The opinions or assertions herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Defense or the Department of Health and Human Services. Financial support: Clinical Investigation Program. Navy Bureau ofMedicine and Surgery (89-06-2733): Naval Medical Research and Development Command (62787 A 3M 162787A870.AR-207). Reprints or correspondence (present address): Dr. Richard E. Hawkins. Internal Medicine Department. Uniformed Services University of the Health Sciences. 430 I Jones Bridge Rd .. Bethesda. MD 20814-4799. The Journal of Infectious Diseases 1992;165:581-5 © 1992 by The University of Chicago. All rights reserved. 0022-1899/92/6503-0030$01.00

(HIV)-infected subjects, and Montagnier et al. [4] reported that Mycoplasma pirum, M. fermentans, and Mycoplasma genitalium were isolated from blood samples of AIDS patients and asymptomatic HI V-seropositive patients. Primarily on the basis of in vitro data, Montagnier et al. [4] suggested a cofactor role for mycoplasma in AIDS pathogenesis. Given data from the Lo and Montagnier research teams, expeditious evaluation of this potential is warranted. We assessed the prevalence of M. fermentans infection in HIV-infected and noninfected populations using polymerase chain reaction (PCR) analysis of peripheral blood samples.

Materials and Methods The sample population consisted of 55 HIV-seropositive patients of various clinical stages who were undergoing routine evaluation from June 1990 through September 1990 at the National Naval Medical Center (Bethesda, MD). All HIV-seropositive subjects were confirmed positive by ELISA and Western blot. Twenty-six HIV-seronegative volunteers were recruited from hospital staff, medical residents, and students to assess M. fermentans prevalence in a group at low risk for HIV infection. DNA extraction. Blood samples were obtained in sterile 7-ml tubes containing EDTA. Whole blood (5 ml) was lysed with 2X lysis buffer (Applied Biosystems, Foster City, CA) and treated with 600 ~l of proteinase K (4X; Applied Biosystems) for 1 hat 65°C. Two phenol-water-chloroform extractions were followed by precipitation with 3 M sodium acetate and 95%ethanol. Pellets were resuspended in 500 ~l ofTRIS-EDTA (TE) buffer for 24 h. Specimens were incubated with RNase (final concentration, 40 ~g/ml) for I h at 37°C. After phenol-water-chloroform extraction and precipitation with 3 M sodium acetate and 95% ethanol, pellets were resuspended in 500 JLI of TE buffer. DNA amplification. The targeted sequence was 206 bp long. Each 1OO-~l reaction mixture consisted of 1 ~g of sample DNA in 10 mMTRIS-HCl, pH 8.0,1 mMEDTA, 10 mMNaCI; 10 ~l

Downloaded from http://jid.oxfordjournals.org/ at UB Giessen on August 11, 2015

A cross-sectional study was undertaken to determine the prevalence of Mycoplasma fermentans infection in patients with human immunodeficiency virus (HIV) infection using polymerase chain reaction methodology. Targeted M. fermentans DNA sequences could be amplified from the DNA extracted from the blood of 6 (11%) of 55 HIV -seropositive patients but from none of 26 HIV-seronegative subjects at low risk for HIV infection (P = .17). There was no correlation between Mi fermentans infection and HIV clinical stage. There was a nonsignificant trend toward an association between M. fermentans infection and a history of syphilis. Infection with M. fermentans may occur more commonly in HIV-infected patients; however, a role as a copathogen or opportunistic infection was not established in this study.

JID 1992;165 (March)

ConciseCommunications

582

of lOX reaction buffer (100 mM TRIS-HCI, 500 mM KCI, 15 mM MgCI 2 , O. 1% (wt/vol) gelatin); each deoxyribonucleoside triphosphate at 200 ILM; 0.5 ILl of Amplitaq DNA polymerase (2.5 units; Cetus, Norwalk, CT); and 50 pmol ofeach oligonucleotide primer. Each primer (Synthecell, Rockville, MD) was 24 bp long (5'-GGACTATTGTCTAAACAATTTCCC-3', 5'-GGTTATTCGATTTCTAAATCGCCT-3'). Primers were based on DNA sequences provided by S.-C. Lo (AFIP). This primer pair has been shown to selectively amplify M. fermentans DNA sequences but not sequences from other species of human mycoplasmas, common tissue culture-contaminating mycoplasmas, selected bacteria, or human, mouse, or monkey tissue [5] (Lo SC, personal communication). The samples were overlaid with 100 ILl of mineral oil and heated to 95°C for 2 min 15 s. Thermal cycling consisted of 45 cycles: 95°C for 35 s, 56°C for 45 s, 72°C for I min, automatic segment extension of 56°C cycle at Is/cycle. Southern blot analysis. Amplified samples (12 ILl) were subjected to electrophoresis on 1.0% agarose gel in TRIS-borateEDTA buffer. Transfer to nylon filters was done using 0.4 M NaOH. After UV cross-linking, filters were washed with 2X SSC (standard saline citrate). Filters were then pre hybridized for 2 h at 68°C. The hybridization solution consisted of5X SSC; blocking reagent (Boehringer Mannheim, Indianapolis), 0.5% (wt/ vol): N-Iaurylsarcosine, Na salt, 0.1 % (wt/vol); SDS, 0.02% (wt/vol), Filters were then incubated for 24 h with 2.5 mi/IOO em? hybridization solution containing 150 ng of labeled DNA probe and 450 ng of salmon sperm DNA. A 73-bp probe was generated using the PCR reaction with primers (5'-GATGAGTGTATTGTCATCC-3', 5'-AACGTAGAAGAGAATGGC3'), which were internal to those used for the targeted DNA sequence described above. The probe was labeled by adding 0.6 mM ofdigoxigenin-dUTP (Boehringer Mannheim) to each PCR tube. After hybridization, filters were washed with 2X SSC; SDS 0.1 % (2X 5 min at room temperature) followed by 0.1 X SSC; SDS 0.1 % (2X 15 min at 68°C). Hybrids were detected by

ELISA using an antidigoxigenin alkaline phosphatase conjugate according to the manufacturer's instructions (DNA labeling and detection kit, nonradioactive; Boehringer Mannheim). Interpretation of results. Gel electrophoresis/Southern blot reactions were positive if an ethidium bromide-stained band was noted on gel and hybridization occurred on nylon filter paper. Positive samples were confirmed in duplicate tests. Laboratory technicians and investigators were blinded with regard to HIV status of samples. Standard statistical techniques were used to assess significance, including Fisher's exact test. No statistical correction for multiple comparisons was made.

Results

615

492 369 246 1 23

2

3

4

5

6

7

8

2

3

4

5

6

7

8

Figure 1. A, Ethidium bromide-stained agarose gel of polymerase chain reaction-amplified DNA. By lane: I, molecular weight markers (base pairs); 2, control Mycoplasma fermentans DNA; 3, 5, 7, representative negative samples; 4, 6, 8, representative positive samples. B, Southern blot analysis of gel shown in A. By lane: 2, control M. fermentans DNA; 3, 5, 7, representative negative samples; 4, 6, 8, representative positive samples.

Downloaded from http://jid.oxfordjournals.org/ at UB Giessen on August 11, 2015

Assays were consistently able to detect peR-amplified products from M. fermentans control DNA in ethidium bromide-stained agarose gels and by Southern blot analysis using as little as 100 fg of template DNA (data not shown). The targeted M. fermentans DNA sequence could be amplified from the DNA extracted from the peripheral blood of 6 (11 %) of the 55 HIV-seropositive patients. No specific M. fermentans sequences, however, could be amplified from DNA extracted from the peripheral blood of any of the 26 seronegatives (P = .17 by two-tailed Fisher's exact test comparing HIV-seropositive and -seronegative subjects). A representative agarose gel and Southern blot are shown in figure 1. Demographic, historical, clinical, and laboratory variables for the HIV-seropositive patients are summarized in table I. The six patients who had M. fermentans DNA sequences identified in peripheral blood samples were similar to the remaining HI v-positive patients with regard to age, race, and sex. All six patients whose peripheral blood contained M. fermentans-specific DNA sequences were classified as having early-stage disease (Walter Reed stages I-IV, table I). M.

lID 1992;165 (March)

Concise Communications

Table 1. Human immunodeficiency virus-infected patients: comparison of Mycoplasmafermentans-positive and -negative patients. M. fermentans-

Patient characteristics

NOTE.

32.3 ± 3.9

M. fermentans-

negative = 49)

(n

33.2 ± 8.2

3 2 1

27 18 4

6 0

48

3 1 1 1 0 0

14* 9 7 3 9 7

2 0

15 7

3 0 0 1 0 2 0 0

9t 15 4 9 4 11 4 3

1 2 3

13 25 11

2

21

who tested negative for M. fer men tans infection had received doxycycline within 3 months of having blood samples drawn. The mean age (± SD) of the seronegative subjects was 33.9 ± 9 years, similar to that of the seropositive subjects (table 1). Those seronegative were more likely to be women (14 [54%] of 26 compared with only 1 [2%] of 55 of those seropositive) and white (24 [92%] of 26 compared with 26 [47%] of 55). One member of the seronegative group was transfused in the 1978-1985 period; another had had heterosexual contact with an HIV-seropositive patient.

I

Discussion

2/5

12/46

4

30

2 0

11/48 6/48 8/47 8/47 3/47

All P> .2, except where shown otherwise.

* P = .17 (two-tailed Fisher's exact test comparing stages 1-4 to 5-6). Syphilis, herpes simplex infection, anal condyloma. P = .08. I P = .055. II Includes hepatitis B surface antigen, surface antibody, or core antibody. t t

fermentans-positive and -negative subjects were similar with regard to skin test results, presence of adenopathy and the use of antiviral, antiparasitic, antifungal, and antibacterial therapy (data not shown). Two HIV-seropositive patients

The present study identifies M. fermentans DNA sequences in PCR-amplified whole blood samples from 11% of HIV-seropositive subjects. No association with HIV disease severity was noted. Thus these initial data do not support the hypothesis that M. fermentans is a cofactor for HIV pathogenesis, It is possible that PCR amplification may be more sensitive using peripheral blood mononuclear cells or other tissue rather than whole blood. Also, primers specific for other mycoplasma were not used. It is possible that other mycoplasma are present; M. fermentans-specific primers would then detect only a subset. Infection with genital mycoplasmas other than M ycoplasma hominis or Ureaplasma urealyticum is relatively uncommon. M. genitalium has been detected using a DNA probe or isolated from patients with nongonococcal urethritis [6, 7] and is seen somewhat more commonly in recurrent or persistent cases and in homosexual men [6]. M. fermentans is also an uncommon isolate and of questionable pathogenicity. Initial isolates were from patients with balanitis and vulvovaginitis; one was obtained in pure culture from the uterine tube of a patient with subacute salpingitis [8]. M. fermentans has also been inconsistently isolated from synovia of patients with arthritis [8] and from blood and bone marrow from patients with leukemia [9]. In determining the clinical significance of a mycoplasma isolated from a clinical specimen, the culture and isolation methods are of critical importance. Mycoplasma are ubiquitous in nature and are frequent tissue culture contaminants. In the present study, contamination was extremely unlikely since direct patient specimens were analyzed with a minimum of manipulation, without passage through tissue culture. DNA extraction and PCR were done in separate laboratories, neither of which were used for tissue culture. Because of the demonstrated colonization of sexually active adults [10], it may be difficult to discern the clinical significance of a mycoplasma isolated from a urogenital source. Isolation of a mycoplasma from a sterile body fluid, such as blood, may carry more clinical significance. However, even documentation of bloodstream invasion may be difficult to correlate with clinical manifestations [11].

Downloaded from http://jid.oxfordjournals.org/ at UB Giessen on August 11, 2015

Mean age (± SO) Race White, non-Hispanic Black, non-Hispanic Hispanic Sex Male Female Walter Reed status 1 2 3 4 5 6 Medical history Urethritis, gonococcal Urethritis, nongonococcal Other sexually transmitted disease" Otolaryngeal/oral disease Candidal esophagitis Oermatologic disease Neurologic disease Gastrointestinal disease Pulmonary disease Other disease Laboratory findings CD4 count (X 106/1) 500 Positive rapid plasma reagin Positive hepatitis A virus IgG Positive hepatitis B virus serology" White blood cell count 13% of all deaths that occur between the ages of 15 and 74 years may be attributable to Chagas' disease [3]. T. cruzi has one of the most complex life cycles among the trypanosomes found in humans [1]. Trypomastigotes circulate in the blood of vertebrate hosts and are usually transmitted by bloodsucking triatomid bugs. The disease can also be spread by blood transfusion and congenital transmission. Commonly, diagnosis is by identification of parasites in the blood, cerebrospinal fluid, fixed tissue, or lymph during periods of high fever; the organisms may be difficult to find at other times or in the chronic stages of infection. Xenodiagnosis, the procedure in which vectors (insects) take a blood meal from the patient and then have their intestinal contents Received 13 August 1991; revised 6 November 1991. Reprints or correspondence: Dr. Alfred A. Pan, Abbott Laboratories, Department 93B, Bldg. RI, 1401 Sheridan Rd., North Chicago, IL 600644000. * Present addresses: ZymoGenetics, Seattle (G.B.R.); Du Pont, Newark, Delaware rv.r.c.i The Journal of Infectious Diseases 1992;165:585-8 © 1992 by The University of Chicago. All rights reserved. 0022-1899/92/6503-0031 $01.00

examined several weeks later, is laborious and lacks sensitivity [4]. Because of these problems, an accepted alternative is serologic testing in blood banks. Specific IgG responses are detected soon after infection and typically the titers of such antibodies remain high for life [5-7]. Common serologic assays used are indirect hemagglutination (HA), [8-10], indirect immunofluorescence (IFA) [8-11], and complement fixation [8, 9]. We investigated a new EIA for Chagas' disease.

Materials and Methods EIA. The EIA developed by Abbott Laboratories is supplied with 100 0.635-cm polystyrene antigen-coated beads (the antigen is from a derivative ofT. cruziy: one vial (30 ml) of specimen diluent; three vials (I ml each) of anti-human IgG conjugate concentrate (goat) peroxidase (horseradish) (minimum concentration: 0.0 I J,Lg/ml in TRIS buffer); three vials (19 ml each) of conjugate diluent (goat and bovine sera in buffer); one vial (0.4 ml) each of positive and negative controls; one bottle (10 tablets) ofo-phenylenediamine- 2 HCl (OPD) tablets; and one bottle (55 ml) of diluent for OPD (citrate-phosphate buffer containing 0.02% hydrogen peroxide). The system is compatible with other Abbott diagnostic test formats, and required peripherals include 20- and/or 60-well trays, assay tubes, cover seals, and sulfuric acid. The EIA is done as follows: 10 J,LI ofa serum sample is diluted into 200 III of specimen diluent in a tray and then incubated with the antigen-coated bead for 30 min at room temperature.

Downloaded from http://jid.oxfordjournals.org/ at UB Giessen on August 11, 2015

A commercial EIA for the detection of antibody to Trypanosoma cruz; was clinically evaluated. The primary use of this test is in the diagnosis and screening of donated blood in Latin America. When compared with sera positive by xenodiagnosis, the assay had a clinical sensitivity of 100%. When tested against matched hemagglutination (HA) and immunofluorescence (IFA) results (i.e., when both tests gave negative results) the EIA had a specificity of 99.03% (305/308). The cross-reactivity of this test was determined using sera from malaria and leishmaniasis patients (obtained from Africa, ensuring that the sera did not contain Chagasic antibodies) and from schistosomiasis, toxoplasmosis, tuberculosis, syphilis, and systemic lupus erythematosus samples. The EIA was 100% specific whereas IFA or commercially available HA kits from Latin America cross-reacted with several of the samples. In this investigation, the EIA appeared to be at least as sensitive and more specific than IFA or HA in the serodiagnosis of Chagas' disease.