422 Article Vol. 10. No. 5

Eur. J. Clin. Microbiol. Infect. Dis., May 1991, p. 422-427 0934-9723191/05 0422-06 $ 3.00/0

Diagnosis of Lyme Borreliosis by an Enzyme Immunoassay Detecting Immunoglobulin G Reactive to Purified Borrelia burgdorferi Cell Components

S. Bergs t r6m 1., A. Sjtistedt 2, L. Doteval l 3'4, B. Kaijser 3, B. Ekstrand-Hammarst r /Sm 5, C. Wallberg 5, G. Skogman 5, A.G. Barbour 6

An enzyme immunoassay (EIA) developed for the diagnosis of Lyme borreliosis was tested for its specificity and sensitivity in detecting IgG antibodies in patients at various stages of the disease. The EIA is based on a detergent extract ofBorrelia burgdorferi which contains 12 proteins of defined molecular weights from Borrelia burgdooreri. The assay showed a specificity of 100 % in control sera from 64 healthy individuals, using a cut-off optical density value of 0.13 (~ +- 3 SD). The sensitivity was 100 % using sera from 22 Swedish patients with late stage Lyme borreliosis and 43 % using sera from 30 patients with the initial stage of the disease. The reactivity of the sera against whole cell preparations, the outer surface proteins OspA and OspB, and the flagella of Borrelia burgdorferi was also tested and compared with the EIA. No cross-reactivity with treponemai antigens was observed when using the EIA.

Lyme borreliosis is caused by the spirochete Bor- relia burgdorferi, this organism being transmitted to humans and animals by ixodid ticks (1-3). The first sign of the disease is often the characteristic skin manifestation erythema migrans (4, 5). Later secondary or tertiary forms of the disease may develop, indicating spread of the microorganisms. In these stages there are a variety of manifesta- tions, including neurological disorders, arthritis and carditis (6). The diagnosis of Lyme borreliosis has mainly been based on clinical observations, which hampers confirmation of the diagnosis in many cases. Lyme borreliosis, like syphilis, may also be mis-diagnosed as the clinical manifesta- tions are similar to those of other diseases, par- ticularly those characterized by neurologic or arthritic symptoms. Determination of serum anti- body titres has proven to be a useful method for the diagnosis of Lyme borreliosis (7), mainly in-

1 Department of Microbiology, and 2Department of Bac- teriology, University of Ume/t, S-901 87 Ume~, Sweden.

3Department of Bacteriology, Sahlgrenska Sjukhuset, S- 413 45 Gothenburg, Sweden.

4Department of Infectious Disease, 6stra Sjukhuset, S- 416 85, Gothenburg, Sweden.

5 Symbicom AB, S-901 24 Ume~, Sweden. 6Departments of Microbiology and Medicine, University

of Texas Health Science Center, San Antonio, Texas 78284, USA.

direct fluorescent antibody (IFA) tests or an en- zyme immunoassay (EIA) directed against whole cell antigen of Borrelia burgdorferi (7-9) being used. However, both tests have shown tow sen- sitivity during the early stages of the infection (10, 11). This is a major problem in the early stage of Lyme borreliosis where there is no evidence of a skin lesion. In the later stages of the infection, IgG antibody titers are usually elevated and the dis- ease is readily diagnosed serologically. However, cross-reactivity occurs with sera from patients with other spirochetal infections, i.e. relapsing fever or syphilis, when using whole cell or soni- cated antigen from Borrelia burgdorferi (7,12). To improve the serologic diagnosis of Lyme bor- reliosis some investigators have used fractionated whole cell preparations of Borrelia burgdorferi (13, 14). A recently developed EIA using purified flagellin as antigen showed increased specificity and sensitivity as compared to an EIA based on whole cell preparations as the test antigen (15,16). Coleman and Benach (14) showed that use of frac- tionated antigenic components from Borrelia burgdorferi other than the flagellin also improved the sensitivity of the assay.

In the present report, sera from patients with dif- ferent stages of Lyme borreliosis were tested with an EIA based on a novel collection of 12 major proteins isolated from Borrelia burgdorferi. This

Vo1.10,1991 423

f rac t ion did no t con t a in the flagell in p ro te in or the cell enve lope .

Materials and Methods

Patient Sera. Serum samples from 30 patients with erythema migrans were kindly provided by Dr. E. .~sbrink and Dr. A. Hovmark of South Hospital, Stock- holm, These patients had clinically verified disease, and their sera had border-line titres in the whole cell EIA, as defined by the State Bacteriological Laboratory, Stock- holm, Sweden (OD405 values between 0.5 and 1.0 in a serum sample diluted 1:100). Twenty-two sera, also from the same source, which were from patients with clinically and serologically verified aerodermatitis chronicum atrophicans (ACA), were tested and regarded as late stage sera. Negative control sera were obtained from 64 healthy individuals from northern Sweden with no known exposure to ticks, Additional sera were obtained from ten patients with reactive arthritis, and from 13 patients with infectious mononucleosis. Forty-nine sera with rheumatoid factor titres ranging from 1/80 to 1/1280, as determined by the method of Waaler (17), were tested, likewise 70 sera with anti-nuclear antibody titres ranging from 1/100 to 1/1600, as determined by an immuno- fluorescent assay (18) on sections from rat liver. Nine sera from syphilis patients with a positive Treponema pallidum (TPHA) reaction were also tested (19), as well as ten Wassermann-positive (false positive), TPHA-nega- tire sera.

Strains and Culture. Two different strains of Borrelia burgdolferi, the Shelter Island New York strain B31 (ATCC 35210) and the Swedish strain ACA-1 (20), were used in antigen preparation. The spirochetes were grown in BSK II medium (21) at 34 °C for five days in 5 % COz/95 % atmospheric air to a cell density of 108/ml. The BSK II medium contained bovine serum albumin Fraction V (Boehringer-Mannheim, Germany).

Borrelia burgdorferi Fractionation and Antigen Prepara- tion. The different protein fractions were prepared es- sentially as described by Magnarelli et al. (22). In brief, 1.5 1 of Borrelia burgdolferi culture containing approx- imately 1012 cells in the late log phase of growth were harvested and washed once with I0 mM Tris (pH 7.4), 150 mM NaCI and 5 mM MgCI2 (TSM-buffer). Octyl-13- D-glucopyranoside (OGP; Calbiochem, USA) was added to a final concentration of 2 % in 10 ml TSM buffer and the suspension incubated at 37 °C for 60 rain. The cell lysate was centrifuged and the clear OGP-supernatant incubated at 56 °C for 30 min. The flocculent white precipitate, formed after heating, was removed by centrifugation at 48,000 x g for 30 min at 37 °C. The supernatant was dialyzed against glass-distilled water at 4 °C. The precipitate formed in the dialysis bag was recovered by centrifugation at 48,000 x g for 30 rain at 25 °C. This water insoluble pellet was sonicated and this material, referred to as the Borrelisa fraction (BF), was used as the specific antigen in the Lyme borreliosis EIA. An outer surface protein fraction (OF) was obtained from the OGP-supernatant; the supernatant was in- cubated at 56 °C in 30 min and then centrifuged at 48,000 x g for 30 min at 37 °C. The resulting pellet was solubil- ized with Sarcosyl (1% sodium lauryl sarcosinate in

10 mM Tris, pH 7.4, 150 mM NaC1, 10 mM EDTA, and 0.05 % sodium azide, TSEA). The supernatant obtained after centrifugation was dialyzed against methanol in glass-distilled water at 20 °C and collected. The .flagellin fraction (FF) was recovered from the pellet after the OGP-treatment. This pellet was washed, centrifuged and solubilized in sarcosyl. This solubilized fraction was then centrifuged at 48,000 x g for 30 min at 25 °C. The resultant pellet was resuspended in 2 % sodium dodecyl sulfate in TSEA and incubated at 65 °C for 30 min. This prepara- tion was then centrifuged at 48,000 x g for 30 min at 25 °C. The resulting supernatant was dialyzed against 25 % methanol before lyophilization.

Enzyme tmmunoassay. The three different fractions used in this study and designated OF, FF and BE correspond to fractions E, D and B, respectively, described by Mag- narelli et al. (22). Protein concentrations were adjusted to 6 lag/ml for each fraction obtained from both Borrelia burgdorferi strains B31 and ACA-1 by addition of phos- phate-buffered saline (PBS). These fractions were coated onto flat-bottom polystyrene microtitre plates (high bind- ing immunoplates; Nunc, Denmark). To each well 50 lal of each suspended fraction were added. The plates were kept in a moist environment and incubated at room temperature overnight. The plates were washed manual- ly, and i00/al of serum diluted 1:200 or 1:500 in PBS with 1% milk powder (wt/vol) was added to the wells and incubated for 2 h at 20 °C. After washing, 50 pl of alkaline phosphatase conjugated rabbit anti-human IgG (Dakopatts, Denmark) diluted 1:250 in PBS was added. The microtitre plates were incubated at 37 °C for 2 h with gentle agitation. The plates were washed and 100 pl of substrate, 1 mg/ml of p-nitrophenyl phosphate (Sigma Chemical, USA), in ethanolamine buffer (pH 9.8) with 5 mM MgCI2 were added. The enzymatic reaction was stopped by the addition of 50 pl ]M NaOH. The optical density (OD) was measured at 405 nm by a microplate photometer (Flow Laboratories, USA).

Polyacrylamide Gel Electrophoresis (PAGE). The protein profiles of whole celt preparations and the three fractions from Borretia burgdolferi strains B31 and ACA- 1 were analyzed by PAGE as described earlier (23). The acrylamide in the separating gel was 15 %. The gels were stained with Coomassie brilliant blue R-250 (Sigma Chemical). Molecular weight standards were obtained from Pharmacia (Sweden) and included proteins ranging in size from 14.4 to 94 kDa. Murine monoclonal anti- bodies (H5332, H6831, and H9724) (23-25) were assayed against whole cell preparations and the fractions by im- munoblot analysis (24) to confirm the presence of the outer surface proteins and the flagellin in Borrelia burgdorferi B31 and ACA-1.

Results

Protein Fractions. T h e p ro te in profi les of whole- cell p r epa ra t ions and f rac t iona ted cell- lysates of Borrelia burgdorferi s t ra ins B31 and A C A - I are shown in F igure 1. F rac t ions B F and O F both con- t a ined ou te r surface p ro t e ins (Osp) O s p A and OspB with mo l e c u l a r sizes of 31 and 34 kD a for s t ra in B31 and 32 and 36 k D a for s t ra in A C A - 1 ,

424 Eur. J. Clin. Microbiol. Infect. Dis.

MW

i ....

% kO(~

67 kDa

4.3 kDa

30 kDo

2 3 /* 5 6 7

m

U

20.1 kOo

Figure 1: Coomassie blue stained SDS-10 % polyacryl- amide gel of whole-cell and fractionated Borrelia burgdor- feri strains B31 and ACA-1. The sizes of the molecular weight standards are shown on the left and right. Lane 1: whole cell ACA-1; Lane 2: whole-cell B31; Lane 3: BF- fraction ACA-1; Lane 4: BF-fraction B31; Lane 5: FF-fr~,c- tion ACA-1; Lane 6: FF-fraction B31; Lane 7: OF-fraction ACA-1; Lane 8: OF-fraction B31. The amount of proteins in lane 1, whole cell extracts of ACA-1, is ten times lower than the amount of proteins in the whole cell extract of B31 in lane 2.

respectively. The Osp-prote ins were highly en- riched in the OF fractions and consti tuted the major component . The presence of OspA and OspB in these antigen fractions was verified by immunoblot t ing with the monoclonal antibodies (data not shown). Strain ACA-1 did not react with the OspB monoclonal ant ibody H6831 as shown earlier by Barbour et al. (21).

The BF fraction contained 12 major proteins. They had approx imate molecular sizes of 11 kDa (12 k D a in ACA-1) , 16 kDa (17 kDa in ACA-1) , 19, 21,22, 27 (only in ACA-1) , 31 (32 in ACA-1) 34 (36 in ACA-1) , 38 (only in B31), 39, 59, 60 and 68 kDa. The FF fraction revealed a major band with an apparen t molecular size of 41 kDa. The flagel- lin protein was identified by immunoblot t ing using the monoclonal ant ibody H9724 (data not shown). Addit ional minor bands with approx-

imate molecular sizes of 60-100 kDa were also seen. The pat tern was similar for both strains. In fraction FF no Osp were present.

Enzyme Immunoassays. Borrelia burgdorferi whole cell prepara t ions and the three f ract ionated prepara t ions (OF, BF, and FF) were used as anti- gens in the EIA. In Table 1 the result obta ined with the three different Borrelia burgdorferi frac- tions is shown. Both Borrelia burgdorferi strains, B31 and ACA-1, were used in equal amounts in the whole cell and fraction EIAs. A set of different positive Lyme borreliosis sera were assayed. The cut-off value in these tests was de termined using sera from 64 persons f rom Nor thern Sweden with no known exposure to Lyme borreliosis as the mean optical density plus three s tandard devia- tions (~ _+ 3 SD). F rom the results presented in Table 1 it can be seen that the BF fraction ex- hibited the highest sensitivity and specificity in this EIA. As shown earlier by Magnarell i et al. (22) an E I A using a whole cell p repara t ion of Bor- relia burgdorferi was less specific than an E I A based on fract ionated antigen. We also obtained similar data when compar ing the BF fraction and whole cell E I A (data not shown). The BF fraction was further assayed using sera from persons who had other diseases (Table 2). Ten sera from patients with reactive arthritis and serum specimens which had shown reactivity in tests for rheumatoid factor (49 sera) and in the Wasser- mann test (10 sera) did not give any significant net absorpt ion in the E I A based on the BF antigen. Two of the 70 (3 %) anti-nuclear ant ibody-posi- tive sera had a net absorpt ion above the cut-off value. One of 13 (8 %) sera f rom mononucleosis patients exceeded the cut-off value in the EIA.

In another exper iment testing 12 low titre sera in addition to 18 sera f rom Table 1, it was shown that 13 of these 30 sera f rom patients with early bor- reliosis (43 %) had an elevated absorpt ion value above the cut-off value de te rmined in the BF- EIA. All high titre sera tested had an absorpt ion greater than the cut-off value (Figure 2).

Table 1: Reactivity of early and late stage borreliosis sera in the EIA using fractionated Borrelia burgdo~feri.

Serum Dilution BF fraction FF fraction OF fraction

A405 NO. (%) A405 No. (%) A405 No. (%) (x -+ SD) positive ~ -+ SD) positive ~ ± SD) positive

Early stage 1/200 0.28 -+ 0,18 10/18 (55) 0.22 -+ 0.16 3/15 (20) 0.067 +_ 0.031 2/11 (18)

Late stage 1/200 1,289 + 0,38 10/10 (100) 0.28 _+ 0.19 7/11 (63) 0.22 -+ 0.093 1/9 (11)

Vol. 10, 1991 425

Table 2: Specificity of Borrelia burgdorferi fraction BF measured as reactivity of various patient serum samples in the EIA.

Serum No. of serum No. showing samples tested cross-reactivity

Rheumatoid factor 49 0

Anti-nuclear antibody 70 2

Wassermann positive 10 0

Syphilis 9 0

Reactive arthritis 10 0

Mononucleosis 13 1

'1. /+ i

1.2-

1.0~-

0.8-

0.6-

0.4-

0,2-

Early borretiosis ko,e borretiosis Heolfhy con,ro[s

Cu~" off

Figure 2: The IgG antibody responses to the Borrelia burgdopferi BF fraction EIA in sera from 52 patients with early and late stage Lyme borreliosis. Control sera from 64 healthy individuals were also measured. The different sera were diluted 1:500. The cut-off value, calculated from the 64 healthy control sera, is marked by a dotted line in the figure.

Discussion

EIAs developed for the diagnosis of Lyme bor- reliosis have been based on whole cell prepara- tions or supernatants of crude fractions. Such EIA methods have shown good sensitivity, but have lacked specificity (7, 9, 10). Other antigenic

preparations have been used, such as the flagellin (15, 16) and fractionated antigens (14, 15). These tests have shown greater specificity and a sen- sitivity almost as good as or even greater than that of the purified flagellin EIA described by Hansen and .~sbrink (16), in comparison to the EIA test based on whole cell antigens. Thus, these methods have been improvements of the EIA based on whole cell antigen. The EIA in the present study uses a modification of the antigen preparation developed by Magnarelli et al. (22). Of the dif- ferent antigenic preparations, the B fraction, which is the equivalent of BF in our study, was found in that study to exhibit greatest sensitivity and specificity (22). The B fraction EIA was then shown to have 90 % sensitivity compared to the EIA based on whole cell antigen, and greater specificity. Whereas Magnarelli et al (22) inves- tigated sera from individuals infected in the USA, we examined sera from individuals infected in Europe.

As shown in the present paper, and also by other investigators (23, 26), Borrelia burgdorferi strains of different geographical origin differ in their protein profiles as shown on Coomassie-staining of PAGE gels. Thus, the pattern of antibody responses, besides being dependent on the stage of infection, may vary between individuals from different parts of the world. Despite these geo- graphical differences, in this study the BF fraction was also found to be the most suitable of the frac- tions tested as antigen in the EIA. The EIA was designed using two strains, one American and one European, in order to detect Borrelia burgdorferi- specific antibodies regardless of the geographical origin of the infecting bacteria. The practical benefits of this design remain to be determined.

The sera included as negative controls showed

426 Eur. J. Clin. Microbiol. Infect. Dis.

little reactivity in the EIA. The cut-off O D value was 0.11 using a serum-dilution of 1:500. This may partly be due to the antigen used, but may also depend on limited exposure of individuals in Northern Sweden to Borrelia burgdorferi. The use of sera from these individuals shows that cross- reactivity to the antigens in the BF fraction is neg- ligible. Despite this, the IgG antibody responses of individuals previously infected with Borrelia burgdorferi were similar to those recorded in tests with whole cell Borrelia burgdorferi. The BF frac- tion is therefore best suited for serologic confir- mation of the late stages of Lyme borreliosis.

Cross-reactivity with treponemal antibodies in tests for Borrelia burgdorferi has been reported (7,12). Likewise, sera from patients with relapsing fever have been shown to cross-react (12). In Europe, cross-reactivity to Treponemapallidum is the major problem. Little is known regarding cross-reactivity to other treponemas, such as Treponema dentieola.

A number of different sera were also investigated for false positive reactions or the presence of anti- bodies cross-reactive to the BF fraction. These in- cluded sera known to be rheumatoid factor posi- tive, which may be problematic in different im- munosorbent assays due to binding of the Fc frag- ment of the immunoglobulin. Also included were sera from patients with acute mononucleosis, in which polyclonal activation is known to occur, and sera displaying anti-nuclear antibodies. One of 13 sera from mononucleosis patients and 2 of 70 sera with anti-nuclear antibodies exceeded the cut-off value. None of these three sera had a net absorp- tion above 0.2. We consider this rather low cross- reactivity to be acceptable since such sera usually show cross-reactivity in most immunosorbent as- says.

A c k n o w l e d g e m e n t s

We express our gratitude to Drs. Eva/~sbrink and Anders Hovmark who kindly provided us with sera from patients with erythema migrans and acrodermatitis ehronieum atrophicans. We thank Torsten Svedberg for photo- graphic assistance and Britt-Inger Strfmberg for help in preparing the manuscript. Dr. Christopher Korch is great- ly acknowledged for critically reading and improving the manuscript. We also thank Dr. Theodor Etmros for valu- able discussions throughout this study. This work was supported by grants from the Swedish Medical Research Council (Dnr 07922), and by Symbicom AB (Ume~, Sweden).

R e f e r e n c e s

1. Burgdorfer W, Barbour AG, Hayes SF, Benaeb JL, Grunwaldt E, Davis, JP: Lyme disease - a tick borne spirochetosis? Science 1983, 216: 1317-1319.

2. Steere AC, Grodzicki RL, Kornblatt AN, Craft JE, Barbour AG, Burgdorfer W, Schmid GP, Johnson E, Malawista SE: The spirochetal etiology of Lyme dis- ease. New England Journal of Medicine 1983, 308: 733-740.

3. Johnson RC, Schmid GP, Hyde FW, Steigerwalt AG, Brenner D J: Borrelia burgdorferi sp. nov. Etiologic agent of Lyme disease. International Journal of Sys- tematic Bacteriology 1984, 4: 496-497.

4. Steere AC, Bartenhagen NH, Craft JE, Hutchinson G J, Newman JH, Rahn DW, Sigal LH, Spieler PN, Stenn KS, Malawista SE: The early clinical manifes- tations of Lyme disease. Annals of Internal Medicine 1983, 99: 76-82.

5. ~,sbrink E, Hovmark A: Successful cultivation of spirochetes from skin lesions of patients with erythema chronicum migrans Afzelius and acrodermatitis chro- nica atrophicans. Acta Pathologica, Microbiotogica et Immunologiea Scandinavica (B) 1985, 93: 161-163.

6. Stiernstedt G, Granstriim M, Hederstedt B, Skiilden- berg B: Diagnosis of spirochetal meningitis by enzyme- linked immunosorbent assay and indirect immunosor- bent assay in serum and cerebrospinal fluid. Journal of Clinical Microbiology 1985, 21: 819-825.

7. Craft JE, Grodzicki RL, Steere AC: Antibody response in Lyme disease: evaluation of diagnostic tests. Journal of Infectious Diseases 1984, 149: 789-795.

8. Russell H, Sampson JS, Schmid GP, Wilkinson HW, Plikaytis B: Enzyme-linked immunosorbent assay and indirect immunofluorescens assay for Lyme disease. Journal of Infectious Diseases 1984, 149: 465--470.

9. Magnarelli LA, Meegan JM, Anderson JF, Chappell WA: Comparison of an indirect fluorescent-antibody test with an enzyme-linked immunosorbent assay for serological studies of Lyme disease. Journal of Clinical Microbiology 1984, 20: 181-184.

10. Shresta M, Grodzicki RL, Steere AC: Diagnosing early Lyme disease. American Journal of Medicine 1985, 78: 235-240.

11. Magnarelli LA, Anderson JF: Early detection and per- sistence of antibodies to Borrelia burgdorferi in per- sons with Lyme disease. Zentralblatt ftir Bakteriologie, Mikrobiologie und Hygiene (A) 1986, 263: 392--400.

12. Magnareili LA, Anderson JF, Johnson RC: Cross- reactivity in serologic tests for Lyme disease and other spirochetal infections. Journal of Infectious Diseases 1987, 156: 183-188.

13. Grodzicki RL, Steere AC: Comparison of immuno- blotting and indirect enzyme-linked immunosorbent assay using different antigen preparations for diag- nosing early Lyme disease. Journal of Infectious Dis- eases 1988, 157: 790-797.

14. Coleman JL, Benach JL: Isolation of antigenic com- ponents from the Lyme disease spirochete: their role in early diagnosis. Journal of Infectious Diseases 1987, 155: 756-765.

15. Hansen K, Hindersson P, Pedersen NS: Measurements of antibodies to Borrefia burgdorferi flagellum im- proves serodiagnosis in Lyme disease. Journal of Clini- cal Microbiology 1988, 26: 338-346.

16. Hansen K, Asbrink E: Serodiagnosis of erythema migrans and acrodermatitis chronica atrophicans by

Vol . 10, 1991 427

the Borretia burgdotferi flagellum enzyme-linked im- munosorbent assay. Journal of Clinical Microbiology 1989, 27: 545-551.

17. Waaler E: On the occurrence of a factor in human serum activating the specific agglutination of sheep blood corpuscles. Acta Pathologica et Microbiologica 1940, 17: 172-177.

18. Burnham TK, Bank PW: Antinuclear antibodies. I, Pattern of nuclear immunofluorescens. Journal of In- vestigative Dermatology 1974, 62: 526-534.

19. Wassermann A, Neisser A, Bruch C: Eine serodia- gnostische Reaktion bei Syphilis. Deutsche Medizini- sche Wochenschrift 1906, 32: 745-746.

20. /~,sbrink E, Hovmark A, Hederstedt B: The spirochctal etiology of aerodermatitis chronicum atrophicans Herxheimer. Acta Dermatologica et Venereologica 1984, 64: 506-512.

21. Barbonr AG: Isolation and cultivation of Lyme disease spirochete. Yale Journal of Biological Medicine 1984, 57: 521-525.

22. Magnarelli LA, Anderson JF~ Barbour AG: Enzyme- linked immunosorbent assays for Lyme disease: reac- tivity of subunits of Borrelia burgdorferi. Journal of Infectious Diseases 1989, 159: 43-49.

23. Barbour AG, Heiland RA, Howe TR: Heterogeneity of major proteins in Lyme disease borreliae: a mole- cular analysis of North American and European iso- lates. Journal of Infectious Diseases 1985,152: 478--484.

24. Barbour AG: Immunochemical analysis of Lyme dis- ease spirochetes. Yale Journal of Biological Medicine 1984, 57: 581-586.

25. Barbour AG, Hayes SF, Heiland RA, Schrumpf ME, Tessier SL: A Borrelia specific monoclonal antibody binds to a flagellar epitope. Infection and Immunity 1986, 52: 549-554.

26. Wilske B, Preac-Mursic V, Schierz G, von Busch K: Immunochemical and immunological analysis of Euro- pean Borrelia burgdorferi strains. Zentralblatt ffir Bak- teriologie, Mikrobiologie und Hygiene (A) 1986, 263: 92-102.