JOURNAL OF CLINICAL MICROBIOLOGY, Mar. 1991, p. 444-448
Vol. 29, No. 3
0095-1137/91/030444-05$02.00/0
Specific Immunofluorescence Staining of Treponema pallidum in Smears and Tissues F. ITO,' E. F. HUNTER,2 R. W. GEORGE,2 B. L. SWISHER,3 AND S. A. LARSEN2* Department of Dermatology, Second Hospital of Nippon Medical School, Kanagawa, Japan,' and Division of Sexually Transmitted Disease Laboratory Research2 and Division of Immunologic, Oncologic, and Hematologic Diseases,3 Center for Infectious Diseases, Centers for Disease Control, Atlanta, Georgia 30333 Received 25 June 1990/Accepted 27 November 1990
To date, tissue sections prepared from Formalin-fixed tissues have not been successfully stained with Treponema pallidum subspecies-specific antibody in a direct fluorescent-antibody assay. While current methods stain T. pallidum, they do not distinguish T. pallidum from other spirochetes such as Borrelia burgdorferi (E. F. Hunter, P. W. Greer, B. L. Swisher, A. R. Simons, C. E. Farshy, J. A. Crawford, and K. R. Sulzer, Arch. Pathol. Lab. Med. 108:878-880, 1984). Because trypsin pretreatment of tissue sections has enhanced other immunofluorescent-antibody (IFA) applications, we compared the use of the trypsin digestion method with the current 1% ammonium hydroxide (NH40H) method as a means to obtain specific staining of T. pallidum in tissues by both direct and indirect IFA techniques. Pretreated T. pallidum-infected tissues sections from rabbits, hamsters, and humans were quantitatively examined with the direct fluorescent-antibody-T. pallidum test conjugate absorbed with Treponema phagedenis, the Reiter treponeme. For indirect staining, a serum specimen from a patient with syphilis absorbed by affinity chromatography with T. phagedenis was used as the primary reagent, and a fluorescein isothiocyanate-labeled rabbit anti-human globulin was used as the secondary reagent. Serum specificity was established first by examining antigen smears of T. pallidum subsp. pallidum, T. pallidum subsp. pertenue, B. burgdorferi, T. phagedenis, and Treponema denticola MRB and then by examining tissues infected with these pathogens plus those infected with four Leptospira serovars. When we stained tissue using the direct IFA method that is currently a standard method for the examination of chancre smears, we found it to be unsuitable for use with tissue. Trypsin digestion did not offer an improvement over the NH40H pretreatment method in the specific identification of T. pallidum by direct IFA. However, specific identification of T. pallidum in tissue sections was obtained by the indirect IFA technique after either trypsin or NH40H pretreatment.
presumably because interference with the fixed tissues prevents the antibody from reaching the antigen. Enzyme digestion for staining antigen in Formalin-fixed sections has been described for other immunohistochemical applications (3, 9), including T. pallidum (12, 20). The trypsin digestion method supposedly unmasks immunoreactive sites of the antigen (9), and as a result, brilliant staining is obtained. We examined the use of trypsin digestion in the specific identification of T. pallidum in tissues by both direct and indirect staining methods and compared these results with the staining results obtained after using the NH4OH pretreatment method.
Treponema pallidum subsp. pallidum infections have increased in number. In 1989 there were nearly 43,000 confirmed cases reported in the United States; this exceeded the number reported for any year since World War 11 (2). Recent reports have alerted us that syphilis may be easily overlooked in teenagers, in whom the disease often presents with atypical characteristics (24, 25). Generally, the manifestations of syphilis are recognizable, but in some cases the clinical disease mimics other infections, and serologic tests may not be able to resolve the coinfection. For example, up to 22.5% of serum specimens from patients with Lyme disease have shown cross-reactivity by the fluorescent treponemal antibody-absorption test (11, 23). In addition, the standard tests for syphilis may fail to react in early primary syphilis in humans with or without human immunodeficiency virus infection. Although these tests have rarely, if ever, been reported to be falsely nonreactive in human immunodeficiency virus-seronegative patients with secondary syphilis (14), delayed development of an expected serologic reaction for syphilis in an human immunodeficiency virus-seroreactive person was reported by Hicks et al. (8). For these reasons, a rapid and specific test for identifying T. pallidum in Formalin-fixed tissues is needed. T. pallidum in chancre smears (15) can easily be stained with specific fluorescein-labeled anti- T.pallidum conjugate absorbed with Treponema phagedenis. However, this conjugate has not been used successfully with Formalin-fixed tissues (10), *
MATERIALS AND METHODS Antigen and tissue preparation. Adult male New Zealand White rabbits were inoculated intratesticularly with 0.5 ml of either T. pallidum subsp. pallidum (Nichols strain) or T. pallidum subsp. pertenue (Gauthier strain) suspended in passage medium (1 part 0.85% NaCl and 1 part heatinactivated normal rabbit serum). Orchitis developed in 10 to 11 days in rabbits that were inoculated with T. pallidum subsp. pallidum and in 13 to 18 days in rabbits that were inoculated with T. pallidum subsp. pertenue. Infected testes were fixed in 10% neutral-buffered Formalin for at least 24 h at room temperature, dehydrated, cleared, and infiltrated with paraffin on an automatic tissue processor. To obtain skin lesion material, adult male New Zealand White rabbits were inoculated intradermally on the back with 0.1 ml of T.
pallidum subsp. pallidum suspended in 50% normal rabbit
Corresponding author. 444
VOL. 29,
1991
IMMUNOFLUORESCENCE STAINING OF T. PALLIDUM
serum. Twenty-six days later, infected skin lesions were excised and placed in 10% neutral-buffered Formalin and processed as described above. Borrelia burgdorferi-infected hamster kidney, bladder, and spleen were received from R. C. Johnson, University of Minnesota (Minneapolis). Leptospira-infected hamster kidneys (four serovars: canicola, grippotyphosa, icterohaemorragiae, and pomona) were received from the National Veterinary Service Laboratories, Animal and Plant Health Inspection Service, U.S. Department of Agriculture. Paraffin-embedded sections of Formalin-fixed human skin lesions from patients with diagnosed primary and secondary syphilis were examined; the patients were from First Hospital of Nippon Medical School, Tokyo, Japan, and the University of Virginia Medical Center, Charlottesville. For smears, T. phagedenis Reiter and Treponema denticola MRB were grown in National Institutes of Health thioglycolate broth containing 10% normal rabbit serum that was heated to 60°C for 2 h. Cultures were harvested after 4 days of incubation at 37°C. The spirochetes were sedimented by centrifugation at 755 x g and washed three times in phosphate-buffered saline (PBS). B. burgdorferi organisms were grown at 33°C for 5 to 7 days in a modified BarbourStoener-Kelly (BSK) medium (1, 26). Leptospira interrogans serovar pomona was grown at 28 to 30°C for 7 days (7) in polysorbate liquid medium-5 liquid (Armour Pharmaceuticals, Kankakee, Ill.). These spirochetes were sedimented by centrifugation at 12,000 x g and then washed three times in PBS. The fluorescent treponemal antibody-absorption test (28) antigen was used for the T. pallidum subsp. pallidum smears, whereas T. pallidum subsp. pertenue antigen was extracted from infected rabbit testicular tissue. Slide preparation. Antigens for direct and indirect staining were placed on 12-well microscopic slides, air dried, and then fixed with reagent-grade acetone for 10 min. Formalinfixed paraffin blocks were cut approximately 3-km thick, floated on a 38°C water bath, and collected on aminoalkylsilane-coated microscope slides (22). Sections were dried in a 60°C oven for 30 min, deparaffinized through two changes of Shandon Xylene Substitute (Shandon, Inc., Pittsburgh, Pa.) combined with xylene, and rehydrated in absolute ethanol-95% ethanol-80% ethanol and placed in a phosphate-buffered solution (pH 7.60) until they were ready to be
stained. Ammonium hydroxide pretreatment of tissue sections. Deparaffinized tissue sections on microscopic slides were flooded with 1% NH40H for 3 min and then rinsed in 3% polysorbate 80 (Tween 80) in PBS, followed by rinsing for 10 min in PBS (4, 10, 18). Trypsin pretreatment of tissue sections. Deparaffinized tissue sections on aminoalkylsilane-coated slides were covered with PBS (pH 7.2) containing 0.25% trypsin (1:250 trypsin; ICN Biomedicals, Cleveland, Ohio) in a moist chamber and incubated at 37°C for 20 to 30 min. Slides were rinsed for 5 min in each of three changes of PBS (12). Conjugates and serum for immunofluorescent-antibody (IFA) staining. Fluorescein isothiocyanate (FITC)-labeled human anti-T. pallidum immunoglobulin G (IgG), which was unabsorbed and absorbed with T. phagedenis, and rabbit FITC-labeled anti-human IgG were reference reagents obtained from the Inventory Control Activity, Technical Service Branch, Scientific Resources Program, Center for Infectious Disease, Centers for Disease Control, Atlanta, Ga. FITC-labeled rabbit anti-L. interrogans and anti-B. burgdorferi were obtained from Katherine Sulzer, Leptospirosis Reference Laboratory, Division of Bacterial Diseases, Cen-
445
TABLE 1. Direct IFA staining of T. pallidum-infected tissues stained with FITC-labeled anti-T. pallidum globulin absorbed with T. phagedenis following trypsin and NH40H pretreatment IFA staining intensity after pretreatment with the following at the indicated conjugate dilution: Trypsin NH40H
Type of tissue
1:2
Rabbit testicular Rabbit skin Human skin Antigen smear
2+ 1+
1:4
1:8
2+ -
1:16
1:2
1:4
1:8
1+
-
1+
2+
2+
-
-
-
-
-
-
-
+
+
-
-
-
-
3-4+
3+
3+
2+
1:16
ter for Infectious Disease, Centers for Disease Control. Human serum specimens from patients with secondary syphilis were obtained from the Serum Bank maintained in
the Treponemal Pathogenesis and Immunobiology Branch, Division of Sexually Transmitted Diseases Laboratory Research, Center for Infectious Disease, Centers for Disease Control. Cross-reacting treponemal antibody was removed by passing the serum through a triethylaminoethyl cellulose column (19) containing T. phagedenis and was concentrated with an Amicon PM-10 filter to 11 mg of protein per ml. Direct IFA procedure. For antigen slides, labeled antibody was diluted serially in PBS containing 2% Tween 80. For tissue slides, conjugates were diluted in PBS containing 2% Tween 80 and 1:20,000 Evans blue dye. After conjugate was added, the slides were placed in a moist chamber and incubated for 30 min at 37°C. Slides were rinsed with a stream of PBS, soaked in PBS for two 5-min periods, rinsed in distilled water, and then blotted to remove excess water. Slides were mounted in glycerol-saline (9 parts glycerol plus 1 part PBS [pH 7.2]) and covered with a cover slip. Indirect IFA procedure. For antigen and tissue slides, the T. phagedenis-absorbed human serum was serially diluted in PBS, beginning at a 1:2 dilution for antigen smears and a 1:5 dilution for tissues. Slides were covered with the appropriate serum dilution and placed in a moist chamber. Slides were incubated at 37°C and rinsed as described above for the direct IFA procedure. FITC-labeled rabbit anti-human IgG diluted 1:5 in PBS containing 2% Tween 80 and 1:20,000 Evans blue dye for tissue sections was added to the slides. Slides were placed in a moist chamber and incubated as described above. Washing steps were repeated, and the slides were mounted with the same medium described above and covered with a cover slip. Observation of antigen smears and tissue sections. After staining, antigen smears and tissue sections were examined
TABLE 2. Indirect IFA staining of T. pallidum-infected human tissues stained with anti-T. pallidum serum absorbed with T. phagedenis following trypsin or NH40H pretreatment
Type of tissue
Rabbit testicular Rabbit skin Human skin
IFA staining intensity after pretreatment with the following at the indicated serum dilution: Trypsin NH40H 1:5
1:10
1:20
1:40
1:5
1:10
1:20
1:40
4+ 3+ 3+
3+ 2+ 4+
3+ 2+ 3+
2+ 2+ 1+
3+ 2+ 3+
4+ 1+ 2+
4+ 2+ 3+
2+ 1+
2+
446
J. CLIN. MICROBIOL.
ITO ET AL.
TABLE 3. Specificity of the indirect staining procedure with anti-T. pallidum human serum absorbed with T. phagedenis in antigen smears of homologous and heterologous spirochetes Staining intensity at the following
Antigen
T. pallidum subsp. pallidum T. pallidum subsp. pertenue T. phagedenis T. denticola B. burgdorferi
serum dilution:
1:2
1:4
1:8
1:16
1:32
1:64
1:128
1:256
1:512
1:1,024
4+ 4+ 1+ + ±
4+ 4+
4+ 3+
± +
± -
3+ 3+ -
3+ 2+ -
3+ 2+ -
2+ 2+ -
2+ 1+ -
1+ 1+ -
1+ ± -
on a Leitz Ortholux II microscope equipped with incident illumination with a mercury lamp (HBO-50). Observations were made by using x40/1.30 oil or x 100/1.32-0.60 oil objectives and x10 oculars. A K filter cube containing BP (band pass) 470 to 490, RKP (Reflexions-Kurz pass) 510, and LP (long pass) 515 filters was used for reading FITC fluorescence. The reader was blinded to the results for the tissue sections.
-
treated slide, but the intensity
RESULTS When rabbit and human tissues were pretreated with trypsin or NH40H and stained with unabsorbed FITClabeled anti-T. pallidum globulin, T. pallidum stained brilliantly (3+ to 4+) from the lowest dilution (1:5) through a 1:80 dilution. Slight variations in staining intensities were observed when the antibody was examined quantitatively and the reader was blinded to the results. These variations were considered minor differences in visual interpretation since all treponemes stained well. In most tissue sections, there was a slight but unquantifiable increase in the intensity of T. pallidum staining when slides were pretreated with trypsin. When tissues were stained with specific absorbed human anti-T. pallidum conjugate, the T. pallidum staining intensity was reduced from that obtained with the unabsorbed conjugate: staining could be observed only with rabbit tissues, and staining could be observed only at a low dilution (Table 1). There was essentially no improvement of staining of rabbit tissue sections when slides were pretreated with trypsin and compared with NH40H-treated slides. Furthermore, the use of trypsin did not increase the permeability of human tissue to staining by the direct method. T. pallidum in rabbit skin tissue was observed at a 1:2 conjugate dilution on a trypsin-
was
too weak for diagnostic
use.
In contrast, specific staining of T. pallidum was demonstrated in rabbit and human tissue sections by an indirect IFA procedure when the reactive human serum absorbed with T. phagedenis by affinity chromatography and an FITC-labeled anti-human antibody were used. Staining intensity and antibody titer were only slightly improved when slides were pretreated with trypsin (Table 2) and, again, may have been only minor differences in visual interpretation. The specificity of the indirect staining procedure was demonstrated quantitatively on antigen smears of homologous and heterologous spirochetes. As seen in Table 3, only minor cross-reactive antibody was observed with T. phagedenis, T. denticola, and B. burgdorferi. Based on the results given in Tables 2 and 3, a 1:8 or a 1:16 dilution of absorbed human serum was selected for further tissue evaluation. The specificity of the selected dilution (1:16) in the indirect IFA procedure for tissue is shown in Table 4. No difference in staining intensity between tissues infected with T. pallidum subsp. pallidum and T. pallidum subsp. pertenue was observed. B. burgdorferi and four Leptospira serovars in infected hamster tissues were not stained with indirect anti-T. pallidum serum but were detected with control homologous anti-B. burgdorferi or Leptospira serovar globulin. DISCUSSION With the reported increase in the incidence of syphilis, clinicians are more likely to encounter patients with syphilis. Usually, these patients are screened by a nontreponemal blood test for syphilis (28) plus a dark-field examination (28) or the Parker ink method (12) when facilities are available.
TABLE 4. Specificity of indirect staining procedure with anti-T. pallidum serum absorbed with T. phagedenis in tissues infected with pathogenic Treponema species, B. burgdorferi, and four Leptospira serovars Staining by: Tissue infected with:
Source of tissue
Indirect IFA for T. pallidum
Direct homologous conjugate
T. pallidum subsp. pallidum T. pallidum subsp. pallidum T. pallidum subsp. pallidum T. pallidum subsp. pertenue B. burgdorferi B. burgdorferi Leptospira serovar canicola Leptospira serovar pomona Leptospira serovar icterohaemorragiae Leptospira serovar grippotyphosa
Rabbit testicle Rabbit skin Human skin Rabbit testicle Hamster kidney Hamster spleen Hamster kidney Hamster kidney Hamster kidney Hamster kidney
+ + + +
+ + + + + + + + + +
-
-
VOL. 29, 1991
IMMUNOFLUORESCENCE STAINING OF T. PALLIDUM
According to Lomholt (17) and Jeerapaet and Ackerman (13), it is not always easy to demonstrate T. pallidum from secondary macular or papular syphilitic lesions. In a previous study, it was difficult to find T. pallidum in the discharge of a patient with secondary roseola syphilitica (12). Biopsies of lesions are sometimes done, but results are histopathologically nonspecific and the demonstration of T. pallidum in tissue sections is required to confirm the diagnosis. Although most human immunodeficiency virus-infected individuals who are also infected with T. pallidum have reactive serological test results for syphilis and have typical clinical signs, several unusual presentations have been reported in the literature (6, 16, 21). Direct examination of lesions was required to establish a diagnosis of syphilis in one patient (8). Clinical manifestation of Lyme disease (caused by B. burgdorferi) usually differs from that of syphilis. However, Dekoning and Hoogkamp-Korstanje (5) reported that B. burgdorferi was located in the dermal papillae, in the edematous subepidermal zone, in lymph vessels, and among collagen fibers. This coincides with the location of T. pallidum in syphilitic skin lesions (12). Thus, a rapid and careful identification of T. pallidum in smears and tissues is sometimes crucial to the correct diagnosis. Direct IFA procedures with unabsorbed FITC-labeled anti-T. pallidum antibody stained T. pallidum in all tissue sections. Tissue slides pretreated with trypsin stained with only a slightly greater intensity than did slides pretreated with NH40H. The minor improvement in staining noted with unabsorbed conjugates on trypsin-treated tissue slides was insufficient to improve staining with specific FITClabeled anti-T. pallidum antibody absorbed with T. phagedenis. This suggested that trypsin digestion of tissues does not greatly increase antibody penetration. Findings obtained by both pretreatment methods were similar to the results reported earlier (10), which indicated that the specific anti-T. pallidum conjugate for chancre smears stains T. pallidum in chancre smears well but is unsatisfactory for examining human tissues for T. pallidum. Our data suggest that the failure of the FITC-labeled anti-T. pallidum IgG absorbed with T. phagedenis to stain T. pallidum in tissues may be related to insufficiently high antibody titer because a 1:2 dilution of the absorbed conjugate did not stain T. pallidum in human tissue, while an unabsorbed conjugate stained T. pallidum through a 1:80 dilution. Also, T. pallidum staining in rabbit testicular tissues appeared to be enhanced compared with staining in rabbit and human skin tissues. The most significant outcome of the study was our ability to obtain specific identification of T. pallidum in tissues by an indirect IFA procedure. Our fractionated absorbed serum was pathogen specific since it stained both T. pallidum subsp. pallidum and T. pallidum subsp. pertenue and not the other spirochetes that were tested. Current direct IFA methods which identify treponemes in tissues are still useful for screening tissues for the presence of spiral organisms and serve the same role in the diagnosis of syphilis as do the silver stains (27). For a more specific diagnosis, we recommend that tissue sections positive for treponemes be further confirmed by either the indirect procedure we described here or with a specific monoclonal antibody, when it is available (12a).
447
REFERENCES 1. Barbour, A. G., W. Burgdorfer, S. F. Hayes, and B. 0. Aeschlumin. 1983. Isolation of a cultivatable spirochete from Ixodes ricimus ticks of Switzerland. Curr. Microbiol. 8:123-126. 2. Centers for Disease Control. 1989. Summary-cases of specified notifiable diseases. United States. Morbid. Mortal. Weekly
Rep. 38:391. 3. Chandler, F. W., and L. Gorelkin. 1983. Immunofluorescence staining of adenovirus in fixed tissues pretreated with trypsin. J. Clin. Microbiol. 17:371-373. 4. Coffin, D. L., and G. Maestrone. 1962. Detection of Leptospira by fluorescent antibody. Am. J. Vet. Res. 23:159-164. 5. Dekoning, J., and J. A. A. Hoogkamp-Korstanje. 1986. Diagnosis of Lyme disease by demonstration of spirochetes in tissue biopsies. Zentralbl. Bakteriol. Mikrobiol. Hyg. 263:179-188. 6. Dowson, S., B. A. Evans, and A. G. Lawrence. 1988. Benign tertiary syphilis and HIV infection. AIDS 2:315-316. 7. Faine, S. 1988. Leptospirosis, p. 344-352. In A. Balows, W. J. Hausler, Jr., M. Ohashi, and A. Turano (ed.), Laboratory diagnosis of infectious diseases: principle and practice, 4th ed. Springer-Verlag, New York. 8. Hicks, C. B., P. M. Benson, G. P. Lupton, and E. C. Tramont. 1987. Seronegative secondary syphilis in a patient infected with the human immunodeficiency virus (HIV) with Kaposi sarcoma. Ann. Intern. Med. 107:492-495. 9. Houng, S. N., H. Minassian, and J. D. Moore. 1976. Application of immunofluorescent staining on paraffin sections improved by trypsin digestion. Lab. lnvest. 35:383-390. 10. Hunter, E. F., P. W. Greer, B. L. Swisher, A. R. Simons, C. E. Farshy, J. A. Crawford, and K. R. Sulzer. 1984. Immunofluorescent staining of Treponema in tissues fixed with formalin. Arch. Pathol. Lab. Med. 108:878-880. 11. Hunter, E. F., H. Russell, C. E. Farshy, J. S. Sampson, and S. A. Larsen. 1986. Evaluation of sera from patients with Lyme disease in the fluorescent treponemal antibody-absorption test for syphilis. Sex. Transm. Dis. 13:232-236. 12. Ito, F. 1987. A study on the demonstration of Treponema pallidum in the tissues of syphilis. J. Nippon Med. Sch. 54:5462. 12a.Ito, F., E. F. Hunter, R. W. George, V. Pope, and S. A. Larsen. Unpublished data. 13. Jeerapaet, P., and A. B. Ackerman. 1973. Histopathologic patterns of secondary syphilis. Arch. Dermatol. 107:373-377. 14. Larsen, S. A. 1989. Serologic reactions when syphilis and HIV occur as coinfections. CAP Today 3:35-38. 15. Larsen, S. A., E. F. Hunter, and B. E. McGrew. 1984. Syphilis, p. 8-11. In B. B. Wentworth and F. N. Judson (ed.), Laboratory methods for the diagnosis of sexually transmitted diseases. American Public Health Association, Washington, D.C. 16. Laska, M. J., D. J. Laska, and F. W. Schmidley. 1988. Syphilitic polyradiculopathy in an HIV positive man. Neurology 38:12971301. 17. Lomholt, G. 1979. Syphilis, yaws and pinta, p. 701-736. In A. Rook, D. S. Wilkinson, and F. J. G. Ebling (ed.), Textbook of dermatology, 3rd ed. Blackwell Scientific Publications, Oxford. 18. Maestrone, G. 1963. The use of an improved fluorescent antibody procedure in the demonstration of Leptospira in animal tissues. Can. J. Comp. Med. Vet. Sci. 27:108-112. 19. McKinney, R. M., L. Thacker, M. C. Wong, and G. A. Hebert. 1978. Improved immunoabsorption procedure with anion-exchange bacterial cell columns. J. Immunol. Methods 21:1-10. 20. Nakamura, S., F. Ito, and S. Onda. 1983. Improved methods for the demonstration of Treponema pallidum in the tissues. J. Dermatol. (Tokyo) 10:433-437. 21. Radolf, J. D., and R. R. Kaplan. 1988. Unusual manifestations of secondary syphilis and abnormal humoral immune response to Treponema pallidum antigens in a homosexual man with asymptomatic human immunodeficiency virus infection. J. Am. Acad. Dermatol. 18:423-428. 22. Rentrop, M., B. Knapp, H. Winter, and J. Schweitzer. 1986. Aminoalkylsilane-treated slides as support for in situ hybridization for keratin cDNAs to frozen tissue sections under varying
448
ITO ET AL.
fixation and pretreatment conditions. Histochemistry 18:271-276. 23. Russell, H., J. S. Sampson, G. P. Smith, H. W. Wilkinson, and B. Plikaytis. 1984. Enzyme-linked immunosorbent assay and indirect immunofluorescence assay for Lyme disease. J. Infect. Dis. 149:465-470. 24. Silber, I. J., M. Kastrinakis, and 0. Taube. 1987. Painful red leg nodules and syphilis: a consideration in patients with erythema nodosum-like illness. Sex. Transm. Dis. 14:52-53. 25. Silber, I. J., and N. F. Niland. 1984. The clinical spectrum of syphilis in adolescence. J. Adolesc. Health Care 5:112-116.
J. CLIN. MICROBIOL. 26. Stoenner, H. G., T. Dodd, and C. Larson. 1982. Antigenic variation of Borrelia hermsii. J. Exp. Med. 156:1297-1311. 27. Swisher, B. L. 1984. Modified Steiner procedure for microwave staining of spirochetes and nonfilamentous bacteria. J. Histotechnol. 10:241-243. 28. U.S. Department of Health, Education and Welfare. 1969. Manual of tests for syphilis, p. 15-17. Public Health Service publication no. 411. National Communicable Disease Center Venereal Disease Program, Atlanta.
Vol. 29, No. 3
0095-1137/91/030444-05$02.00/0
Specific Immunofluorescence Staining of Treponema pallidum in Smears and Tissues F. ITO,' E. F. HUNTER,2 R. W. GEORGE,2 B. L. SWISHER,3 AND S. A. LARSEN2* Department of Dermatology, Second Hospital of Nippon Medical School, Kanagawa, Japan,' and Division of Sexually Transmitted Disease Laboratory Research2 and Division of Immunologic, Oncologic, and Hematologic Diseases,3 Center for Infectious Diseases, Centers for Disease Control, Atlanta, Georgia 30333 Received 25 June 1990/Accepted 27 November 1990
To date, tissue sections prepared from Formalin-fixed tissues have not been successfully stained with Treponema pallidum subspecies-specific antibody in a direct fluorescent-antibody assay. While current methods stain T. pallidum, they do not distinguish T. pallidum from other spirochetes such as Borrelia burgdorferi (E. F. Hunter, P. W. Greer, B. L. Swisher, A. R. Simons, C. E. Farshy, J. A. Crawford, and K. R. Sulzer, Arch. Pathol. Lab. Med. 108:878-880, 1984). Because trypsin pretreatment of tissue sections has enhanced other immunofluorescent-antibody (IFA) applications, we compared the use of the trypsin digestion method with the current 1% ammonium hydroxide (NH40H) method as a means to obtain specific staining of T. pallidum in tissues by both direct and indirect IFA techniques. Pretreated T. pallidum-infected tissues sections from rabbits, hamsters, and humans were quantitatively examined with the direct fluorescent-antibody-T. pallidum test conjugate absorbed with Treponema phagedenis, the Reiter treponeme. For indirect staining, a serum specimen from a patient with syphilis absorbed by affinity chromatography with T. phagedenis was used as the primary reagent, and a fluorescein isothiocyanate-labeled rabbit anti-human globulin was used as the secondary reagent. Serum specificity was established first by examining antigen smears of T. pallidum subsp. pallidum, T. pallidum subsp. pertenue, B. burgdorferi, T. phagedenis, and Treponema denticola MRB and then by examining tissues infected with these pathogens plus those infected with four Leptospira serovars. When we stained tissue using the direct IFA method that is currently a standard method for the examination of chancre smears, we found it to be unsuitable for use with tissue. Trypsin digestion did not offer an improvement over the NH40H pretreatment method in the specific identification of T. pallidum by direct IFA. However, specific identification of T. pallidum in tissue sections was obtained by the indirect IFA technique after either trypsin or NH40H pretreatment.
presumably because interference with the fixed tissues prevents the antibody from reaching the antigen. Enzyme digestion for staining antigen in Formalin-fixed sections has been described for other immunohistochemical applications (3, 9), including T. pallidum (12, 20). The trypsin digestion method supposedly unmasks immunoreactive sites of the antigen (9), and as a result, brilliant staining is obtained. We examined the use of trypsin digestion in the specific identification of T. pallidum in tissues by both direct and indirect staining methods and compared these results with the staining results obtained after using the NH4OH pretreatment method.
Treponema pallidum subsp. pallidum infections have increased in number. In 1989 there were nearly 43,000 confirmed cases reported in the United States; this exceeded the number reported for any year since World War 11 (2). Recent reports have alerted us that syphilis may be easily overlooked in teenagers, in whom the disease often presents with atypical characteristics (24, 25). Generally, the manifestations of syphilis are recognizable, but in some cases the clinical disease mimics other infections, and serologic tests may not be able to resolve the coinfection. For example, up to 22.5% of serum specimens from patients with Lyme disease have shown cross-reactivity by the fluorescent treponemal antibody-absorption test (11, 23). In addition, the standard tests for syphilis may fail to react in early primary syphilis in humans with or without human immunodeficiency virus infection. Although these tests have rarely, if ever, been reported to be falsely nonreactive in human immunodeficiency virus-seronegative patients with secondary syphilis (14), delayed development of an expected serologic reaction for syphilis in an human immunodeficiency virus-seroreactive person was reported by Hicks et al. (8). For these reasons, a rapid and specific test for identifying T. pallidum in Formalin-fixed tissues is needed. T. pallidum in chancre smears (15) can easily be stained with specific fluorescein-labeled anti- T.pallidum conjugate absorbed with Treponema phagedenis. However, this conjugate has not been used successfully with Formalin-fixed tissues (10), *
MATERIALS AND METHODS Antigen and tissue preparation. Adult male New Zealand White rabbits were inoculated intratesticularly with 0.5 ml of either T. pallidum subsp. pallidum (Nichols strain) or T. pallidum subsp. pertenue (Gauthier strain) suspended in passage medium (1 part 0.85% NaCl and 1 part heatinactivated normal rabbit serum). Orchitis developed in 10 to 11 days in rabbits that were inoculated with T. pallidum subsp. pallidum and in 13 to 18 days in rabbits that were inoculated with T. pallidum subsp. pertenue. Infected testes were fixed in 10% neutral-buffered Formalin for at least 24 h at room temperature, dehydrated, cleared, and infiltrated with paraffin on an automatic tissue processor. To obtain skin lesion material, adult male New Zealand White rabbits were inoculated intradermally on the back with 0.1 ml of T.
pallidum subsp. pallidum suspended in 50% normal rabbit
Corresponding author. 444
VOL. 29,
1991
IMMUNOFLUORESCENCE STAINING OF T. PALLIDUM
serum. Twenty-six days later, infected skin lesions were excised and placed in 10% neutral-buffered Formalin and processed as described above. Borrelia burgdorferi-infected hamster kidney, bladder, and spleen were received from R. C. Johnson, University of Minnesota (Minneapolis). Leptospira-infected hamster kidneys (four serovars: canicola, grippotyphosa, icterohaemorragiae, and pomona) were received from the National Veterinary Service Laboratories, Animal and Plant Health Inspection Service, U.S. Department of Agriculture. Paraffin-embedded sections of Formalin-fixed human skin lesions from patients with diagnosed primary and secondary syphilis were examined; the patients were from First Hospital of Nippon Medical School, Tokyo, Japan, and the University of Virginia Medical Center, Charlottesville. For smears, T. phagedenis Reiter and Treponema denticola MRB were grown in National Institutes of Health thioglycolate broth containing 10% normal rabbit serum that was heated to 60°C for 2 h. Cultures were harvested after 4 days of incubation at 37°C. The spirochetes were sedimented by centrifugation at 755 x g and washed three times in phosphate-buffered saline (PBS). B. burgdorferi organisms were grown at 33°C for 5 to 7 days in a modified BarbourStoener-Kelly (BSK) medium (1, 26). Leptospira interrogans serovar pomona was grown at 28 to 30°C for 7 days (7) in polysorbate liquid medium-5 liquid (Armour Pharmaceuticals, Kankakee, Ill.). These spirochetes were sedimented by centrifugation at 12,000 x g and then washed three times in PBS. The fluorescent treponemal antibody-absorption test (28) antigen was used for the T. pallidum subsp. pallidum smears, whereas T. pallidum subsp. pertenue antigen was extracted from infected rabbit testicular tissue. Slide preparation. Antigens for direct and indirect staining were placed on 12-well microscopic slides, air dried, and then fixed with reagent-grade acetone for 10 min. Formalinfixed paraffin blocks were cut approximately 3-km thick, floated on a 38°C water bath, and collected on aminoalkylsilane-coated microscope slides (22). Sections were dried in a 60°C oven for 30 min, deparaffinized through two changes of Shandon Xylene Substitute (Shandon, Inc., Pittsburgh, Pa.) combined with xylene, and rehydrated in absolute ethanol-95% ethanol-80% ethanol and placed in a phosphate-buffered solution (pH 7.60) until they were ready to be
stained. Ammonium hydroxide pretreatment of tissue sections. Deparaffinized tissue sections on microscopic slides were flooded with 1% NH40H for 3 min and then rinsed in 3% polysorbate 80 (Tween 80) in PBS, followed by rinsing for 10 min in PBS (4, 10, 18). Trypsin pretreatment of tissue sections. Deparaffinized tissue sections on aminoalkylsilane-coated slides were covered with PBS (pH 7.2) containing 0.25% trypsin (1:250 trypsin; ICN Biomedicals, Cleveland, Ohio) in a moist chamber and incubated at 37°C for 20 to 30 min. Slides were rinsed for 5 min in each of three changes of PBS (12). Conjugates and serum for immunofluorescent-antibody (IFA) staining. Fluorescein isothiocyanate (FITC)-labeled human anti-T. pallidum immunoglobulin G (IgG), which was unabsorbed and absorbed with T. phagedenis, and rabbit FITC-labeled anti-human IgG were reference reagents obtained from the Inventory Control Activity, Technical Service Branch, Scientific Resources Program, Center for Infectious Disease, Centers for Disease Control, Atlanta, Ga. FITC-labeled rabbit anti-L. interrogans and anti-B. burgdorferi were obtained from Katherine Sulzer, Leptospirosis Reference Laboratory, Division of Bacterial Diseases, Cen-
445
TABLE 1. Direct IFA staining of T. pallidum-infected tissues stained with FITC-labeled anti-T. pallidum globulin absorbed with T. phagedenis following trypsin and NH40H pretreatment IFA staining intensity after pretreatment with the following at the indicated conjugate dilution: Trypsin NH40H
Type of tissue
1:2
Rabbit testicular Rabbit skin Human skin Antigen smear
2+ 1+
1:4
1:8
2+ -
1:16
1:2
1:4
1:8
1+
-
1+
2+
2+
-
-
-
-
-
-
-
+
+
-
-
-
-
3-4+
3+
3+
2+
1:16
ter for Infectious Disease, Centers for Disease Control. Human serum specimens from patients with secondary syphilis were obtained from the Serum Bank maintained in
the Treponemal Pathogenesis and Immunobiology Branch, Division of Sexually Transmitted Diseases Laboratory Research, Center for Infectious Disease, Centers for Disease Control. Cross-reacting treponemal antibody was removed by passing the serum through a triethylaminoethyl cellulose column (19) containing T. phagedenis and was concentrated with an Amicon PM-10 filter to 11 mg of protein per ml. Direct IFA procedure. For antigen slides, labeled antibody was diluted serially in PBS containing 2% Tween 80. For tissue slides, conjugates were diluted in PBS containing 2% Tween 80 and 1:20,000 Evans blue dye. After conjugate was added, the slides were placed in a moist chamber and incubated for 30 min at 37°C. Slides were rinsed with a stream of PBS, soaked in PBS for two 5-min periods, rinsed in distilled water, and then blotted to remove excess water. Slides were mounted in glycerol-saline (9 parts glycerol plus 1 part PBS [pH 7.2]) and covered with a cover slip. Indirect IFA procedure. For antigen and tissue slides, the T. phagedenis-absorbed human serum was serially diluted in PBS, beginning at a 1:2 dilution for antigen smears and a 1:5 dilution for tissues. Slides were covered with the appropriate serum dilution and placed in a moist chamber. Slides were incubated at 37°C and rinsed as described above for the direct IFA procedure. FITC-labeled rabbit anti-human IgG diluted 1:5 in PBS containing 2% Tween 80 and 1:20,000 Evans blue dye for tissue sections was added to the slides. Slides were placed in a moist chamber and incubated as described above. Washing steps were repeated, and the slides were mounted with the same medium described above and covered with a cover slip. Observation of antigen smears and tissue sections. After staining, antigen smears and tissue sections were examined
TABLE 2. Indirect IFA staining of T. pallidum-infected human tissues stained with anti-T. pallidum serum absorbed with T. phagedenis following trypsin or NH40H pretreatment
Type of tissue
Rabbit testicular Rabbit skin Human skin
IFA staining intensity after pretreatment with the following at the indicated serum dilution: Trypsin NH40H 1:5
1:10
1:20
1:40
1:5
1:10
1:20
1:40
4+ 3+ 3+
3+ 2+ 4+
3+ 2+ 3+
2+ 2+ 1+
3+ 2+ 3+
4+ 1+ 2+
4+ 2+ 3+
2+ 1+
2+
446
J. CLIN. MICROBIOL.
ITO ET AL.
TABLE 3. Specificity of the indirect staining procedure with anti-T. pallidum human serum absorbed with T. phagedenis in antigen smears of homologous and heterologous spirochetes Staining intensity at the following
Antigen
T. pallidum subsp. pallidum T. pallidum subsp. pertenue T. phagedenis T. denticola B. burgdorferi
serum dilution:
1:2
1:4
1:8
1:16
1:32
1:64
1:128
1:256
1:512
1:1,024
4+ 4+ 1+ + ±
4+ 4+
4+ 3+
± +
± -
3+ 3+ -
3+ 2+ -
3+ 2+ -
2+ 2+ -
2+ 1+ -
1+ 1+ -
1+ ± -
on a Leitz Ortholux II microscope equipped with incident illumination with a mercury lamp (HBO-50). Observations were made by using x40/1.30 oil or x 100/1.32-0.60 oil objectives and x10 oculars. A K filter cube containing BP (band pass) 470 to 490, RKP (Reflexions-Kurz pass) 510, and LP (long pass) 515 filters was used for reading FITC fluorescence. The reader was blinded to the results for the tissue sections.
-
treated slide, but the intensity
RESULTS When rabbit and human tissues were pretreated with trypsin or NH40H and stained with unabsorbed FITClabeled anti-T. pallidum globulin, T. pallidum stained brilliantly (3+ to 4+) from the lowest dilution (1:5) through a 1:80 dilution. Slight variations in staining intensities were observed when the antibody was examined quantitatively and the reader was blinded to the results. These variations were considered minor differences in visual interpretation since all treponemes stained well. In most tissue sections, there was a slight but unquantifiable increase in the intensity of T. pallidum staining when slides were pretreated with trypsin. When tissues were stained with specific absorbed human anti-T. pallidum conjugate, the T. pallidum staining intensity was reduced from that obtained with the unabsorbed conjugate: staining could be observed only with rabbit tissues, and staining could be observed only at a low dilution (Table 1). There was essentially no improvement of staining of rabbit tissue sections when slides were pretreated with trypsin and compared with NH40H-treated slides. Furthermore, the use of trypsin did not increase the permeability of human tissue to staining by the direct method. T. pallidum in rabbit skin tissue was observed at a 1:2 conjugate dilution on a trypsin-
was
too weak for diagnostic
use.
In contrast, specific staining of T. pallidum was demonstrated in rabbit and human tissue sections by an indirect IFA procedure when the reactive human serum absorbed with T. phagedenis by affinity chromatography and an FITC-labeled anti-human antibody were used. Staining intensity and antibody titer were only slightly improved when slides were pretreated with trypsin (Table 2) and, again, may have been only minor differences in visual interpretation. The specificity of the indirect staining procedure was demonstrated quantitatively on antigen smears of homologous and heterologous spirochetes. As seen in Table 3, only minor cross-reactive antibody was observed with T. phagedenis, T. denticola, and B. burgdorferi. Based on the results given in Tables 2 and 3, a 1:8 or a 1:16 dilution of absorbed human serum was selected for further tissue evaluation. The specificity of the selected dilution (1:16) in the indirect IFA procedure for tissue is shown in Table 4. No difference in staining intensity between tissues infected with T. pallidum subsp. pallidum and T. pallidum subsp. pertenue was observed. B. burgdorferi and four Leptospira serovars in infected hamster tissues were not stained with indirect anti-T. pallidum serum but were detected with control homologous anti-B. burgdorferi or Leptospira serovar globulin. DISCUSSION With the reported increase in the incidence of syphilis, clinicians are more likely to encounter patients with syphilis. Usually, these patients are screened by a nontreponemal blood test for syphilis (28) plus a dark-field examination (28) or the Parker ink method (12) when facilities are available.
TABLE 4. Specificity of indirect staining procedure with anti-T. pallidum serum absorbed with T. phagedenis in tissues infected with pathogenic Treponema species, B. burgdorferi, and four Leptospira serovars Staining by: Tissue infected with:
Source of tissue
Indirect IFA for T. pallidum
Direct homologous conjugate
T. pallidum subsp. pallidum T. pallidum subsp. pallidum T. pallidum subsp. pallidum T. pallidum subsp. pertenue B. burgdorferi B. burgdorferi Leptospira serovar canicola Leptospira serovar pomona Leptospira serovar icterohaemorragiae Leptospira serovar grippotyphosa
Rabbit testicle Rabbit skin Human skin Rabbit testicle Hamster kidney Hamster spleen Hamster kidney Hamster kidney Hamster kidney Hamster kidney
+ + + +
+ + + + + + + + + +
-
-
VOL. 29, 1991
IMMUNOFLUORESCENCE STAINING OF T. PALLIDUM
According to Lomholt (17) and Jeerapaet and Ackerman (13), it is not always easy to demonstrate T. pallidum from secondary macular or papular syphilitic lesions. In a previous study, it was difficult to find T. pallidum in the discharge of a patient with secondary roseola syphilitica (12). Biopsies of lesions are sometimes done, but results are histopathologically nonspecific and the demonstration of T. pallidum in tissue sections is required to confirm the diagnosis. Although most human immunodeficiency virus-infected individuals who are also infected with T. pallidum have reactive serological test results for syphilis and have typical clinical signs, several unusual presentations have been reported in the literature (6, 16, 21). Direct examination of lesions was required to establish a diagnosis of syphilis in one patient (8). Clinical manifestation of Lyme disease (caused by B. burgdorferi) usually differs from that of syphilis. However, Dekoning and Hoogkamp-Korstanje (5) reported that B. burgdorferi was located in the dermal papillae, in the edematous subepidermal zone, in lymph vessels, and among collagen fibers. This coincides with the location of T. pallidum in syphilitic skin lesions (12). Thus, a rapid and careful identification of T. pallidum in smears and tissues is sometimes crucial to the correct diagnosis. Direct IFA procedures with unabsorbed FITC-labeled anti-T. pallidum antibody stained T. pallidum in all tissue sections. Tissue slides pretreated with trypsin stained with only a slightly greater intensity than did slides pretreated with NH40H. The minor improvement in staining noted with unabsorbed conjugates on trypsin-treated tissue slides was insufficient to improve staining with specific FITClabeled anti-T. pallidum antibody absorbed with T. phagedenis. This suggested that trypsin digestion of tissues does not greatly increase antibody penetration. Findings obtained by both pretreatment methods were similar to the results reported earlier (10), which indicated that the specific anti-T. pallidum conjugate for chancre smears stains T. pallidum in chancre smears well but is unsatisfactory for examining human tissues for T. pallidum. Our data suggest that the failure of the FITC-labeled anti-T. pallidum IgG absorbed with T. phagedenis to stain T. pallidum in tissues may be related to insufficiently high antibody titer because a 1:2 dilution of the absorbed conjugate did not stain T. pallidum in human tissue, while an unabsorbed conjugate stained T. pallidum through a 1:80 dilution. Also, T. pallidum staining in rabbit testicular tissues appeared to be enhanced compared with staining in rabbit and human skin tissues. The most significant outcome of the study was our ability to obtain specific identification of T. pallidum in tissues by an indirect IFA procedure. Our fractionated absorbed serum was pathogen specific since it stained both T. pallidum subsp. pallidum and T. pallidum subsp. pertenue and not the other spirochetes that were tested. Current direct IFA methods which identify treponemes in tissues are still useful for screening tissues for the presence of spiral organisms and serve the same role in the diagnosis of syphilis as do the silver stains (27). For a more specific diagnosis, we recommend that tissue sections positive for treponemes be further confirmed by either the indirect procedure we described here or with a specific monoclonal antibody, when it is available (12a).
447
REFERENCES 1. Barbour, A. G., W. Burgdorfer, S. F. Hayes, and B. 0. Aeschlumin. 1983. Isolation of a cultivatable spirochete from Ixodes ricimus ticks of Switzerland. Curr. Microbiol. 8:123-126. 2. Centers for Disease Control. 1989. Summary-cases of specified notifiable diseases. United States. Morbid. Mortal. Weekly
Rep. 38:391. 3. Chandler, F. W., and L. Gorelkin. 1983. Immunofluorescence staining of adenovirus in fixed tissues pretreated with trypsin. J. Clin. Microbiol. 17:371-373. 4. Coffin, D. L., and G. Maestrone. 1962. Detection of Leptospira by fluorescent antibody. Am. J. Vet. Res. 23:159-164. 5. Dekoning, J., and J. A. A. Hoogkamp-Korstanje. 1986. Diagnosis of Lyme disease by demonstration of spirochetes in tissue biopsies. Zentralbl. Bakteriol. Mikrobiol. Hyg. 263:179-188. 6. Dowson, S., B. A. Evans, and A. G. Lawrence. 1988. Benign tertiary syphilis and HIV infection. AIDS 2:315-316. 7. Faine, S. 1988. Leptospirosis, p. 344-352. In A. Balows, W. J. Hausler, Jr., M. Ohashi, and A. Turano (ed.), Laboratory diagnosis of infectious diseases: principle and practice, 4th ed. Springer-Verlag, New York. 8. Hicks, C. B., P. M. Benson, G. P. Lupton, and E. C. Tramont. 1987. Seronegative secondary syphilis in a patient infected with the human immunodeficiency virus (HIV) with Kaposi sarcoma. Ann. Intern. Med. 107:492-495. 9. Houng, S. N., H. Minassian, and J. D. Moore. 1976. Application of immunofluorescent staining on paraffin sections improved by trypsin digestion. Lab. lnvest. 35:383-390. 10. Hunter, E. F., P. W. Greer, B. L. Swisher, A. R. Simons, C. E. Farshy, J. A. Crawford, and K. R. Sulzer. 1984. Immunofluorescent staining of Treponema in tissues fixed with formalin. Arch. Pathol. Lab. Med. 108:878-880. 11. Hunter, E. F., H. Russell, C. E. Farshy, J. S. Sampson, and S. A. Larsen. 1986. Evaluation of sera from patients with Lyme disease in the fluorescent treponemal antibody-absorption test for syphilis. Sex. Transm. Dis. 13:232-236. 12. Ito, F. 1987. A study on the demonstration of Treponema pallidum in the tissues of syphilis. J. Nippon Med. Sch. 54:5462. 12a.Ito, F., E. F. Hunter, R. W. George, V. Pope, and S. A. Larsen. Unpublished data. 13. Jeerapaet, P., and A. B. Ackerman. 1973. Histopathologic patterns of secondary syphilis. Arch. Dermatol. 107:373-377. 14. Larsen, S. A. 1989. Serologic reactions when syphilis and HIV occur as coinfections. CAP Today 3:35-38. 15. Larsen, S. A., E. F. Hunter, and B. E. McGrew. 1984. Syphilis, p. 8-11. In B. B. Wentworth and F. N. Judson (ed.), Laboratory methods for the diagnosis of sexually transmitted diseases. American Public Health Association, Washington, D.C. 16. Laska, M. J., D. J. Laska, and F. W. Schmidley. 1988. Syphilitic polyradiculopathy in an HIV positive man. Neurology 38:12971301. 17. Lomholt, G. 1979. Syphilis, yaws and pinta, p. 701-736. In A. Rook, D. S. Wilkinson, and F. J. G. Ebling (ed.), Textbook of dermatology, 3rd ed. Blackwell Scientific Publications, Oxford. 18. Maestrone, G. 1963. The use of an improved fluorescent antibody procedure in the demonstration of Leptospira in animal tissues. Can. J. Comp. Med. Vet. Sci. 27:108-112. 19. McKinney, R. M., L. Thacker, M. C. Wong, and G. A. Hebert. 1978. Improved immunoabsorption procedure with anion-exchange bacterial cell columns. J. Immunol. Methods 21:1-10. 20. Nakamura, S., F. Ito, and S. Onda. 1983. Improved methods for the demonstration of Treponema pallidum in the tissues. J. Dermatol. (Tokyo) 10:433-437. 21. Radolf, J. D., and R. R. Kaplan. 1988. Unusual manifestations of secondary syphilis and abnormal humoral immune response to Treponema pallidum antigens in a homosexual man with asymptomatic human immunodeficiency virus infection. J. Am. Acad. Dermatol. 18:423-428. 22. Rentrop, M., B. Knapp, H. Winter, and J. Schweitzer. 1986. Aminoalkylsilane-treated slides as support for in situ hybridization for keratin cDNAs to frozen tissue sections under varying
448
ITO ET AL.
fixation and pretreatment conditions. Histochemistry 18:271-276. 23. Russell, H., J. S. Sampson, G. P. Smith, H. W. Wilkinson, and B. Plikaytis. 1984. Enzyme-linked immunosorbent assay and indirect immunofluorescence assay for Lyme disease. J. Infect. Dis. 149:465-470. 24. Silber, I. J., M. Kastrinakis, and 0. Taube. 1987. Painful red leg nodules and syphilis: a consideration in patients with erythema nodosum-like illness. Sex. Transm. Dis. 14:52-53. 25. Silber, I. J., and N. F. Niland. 1984. The clinical spectrum of syphilis in adolescence. J. Adolesc. Health Care 5:112-116.
J. CLIN. MICROBIOL. 26. Stoenner, H. G., T. Dodd, and C. Larson. 1982. Antigenic variation of Borrelia hermsii. J. Exp. Med. 156:1297-1311. 27. Swisher, B. L. 1984. Modified Steiner procedure for microwave staining of spirochetes and nonfilamentous bacteria. J. Histotechnol. 10:241-243. 28. U.S. Department of Health, Education and Welfare. 1969. Manual of tests for syphilis, p. 15-17. Public Health Service publication no. 411. National Communicable Disease Center Venereal Disease Program, Atlanta.
