Lyme Borreliosis Spirochete and its Tick Vector Relationship of protein expression, antigenicity and pathogenicity of Borrelia burgdorferi (Spirochaetales: Spirochaetaceae) sensu lato and Ixodes ricinus ticks (Acari: Ixodidae) THESE Présentée à la Faculté des Sciences de l'Université de Neuchâtel pour l'obtention du grade de Docteur es Sciences par Chang Min HU Licenciée en Médecine Neuchâtel 1996 IMPRIMATUR POUR LATHESE Relations entre les spirochetes de la borréliose de Lyme et leur vecteur Ixodes ricinus: expression protéinique, antigénique et pathogénicité de Borrelia burgdorferi sensu lato. de Mme Chang Min Hu UNIVERSITE DE NEUCHATEL FACULTÉ DES SCIENCES La Faculté des sciences de l'Université de Neuchâtel sur le rapport des membres du jury, M. B. Betschart (directeur de thèse), Mme L Gern (co-directrice de thèse), MM. M. Brossard, O. Péter (Sion) et M. Simon (Freiburg, Deutschland) autorise l'impression de la présente thèse. Neuchâtel, le 30 janvier 1997 Le doyen: R. Dändliker To my parents, To my husband, To my children. A mes parents, A mon mari, A mes enfans. This thesis is based on the following papers: (The complete thesis is deposited in the Central Library of Neuchâtel University) I. Comparison in the immunological properties of Borrelia burgdorferi isolates from Ixodes ricinus derived from three endemic areas in Switzerland. Hu C. M., Leuba - Garcia S., Kramer M. D., Aeschlimann A. and Gern L. Epidemiol. Infect. (1994) 112: 533 - 542 II. Changes in the protein profile and antigenicity of different Borrelia burgdorferi strains after reintroduction to Ixodes ricinus ticks. Hu C. M, Gern L. and Aeschlimann A. Parasite Immunology (1992) 14: 415 - 427 III. Antigenic variation in Borrelia burgdorferi after passage through Ixodes ricinus and Ixodes hexagonus. Gern L., Hu C. M., Toutoungi L. A. and Kramer M. D. In: Proc. of the 1st Intern. Conf. On Tick - Borne Pathogens at the Host - Vector Interface: An Agenda for Research. Munderloh U. G. and Kurtti T. J. (eds), Saint Paul, USA, 1992, pp 121 - 125 IV. Proteinic, antigenic and pathogenic variations of a clonal Borrelia burgdorferi isolate from Ixodes ricinus hemolymph. Hu C. M., Kramer M. D., Simon M. M. and Gern L. In: Proc of VI International Conference on Lyme borreliosis. Cevenini R., Sambri V. and La Placa M. (eds), Bologna: Italy June 19 - 22,1994, pp 19 V. Tick factors and in vitro cultivation influence the protein profile, antigenicity and pathogenicity of a clonal Borrelia garinii isolate from Ixodes ricinus hemolymph. Hu C. M., Simon M. M. Kramer M. D. and Gern L. INFECTION (1996) 24 (3): 251/41-257/47 VI. Apodemus sp rodents, reservoir hosts for Borrelia afzelii in an endemic area in Switzerland. Hu C. M., Humair P. F., Wallich R. and Gern L. ZbI. Bakt. Hyg. (1997) 285, 558-564 Epidemiol, lnjeci. (1994), 112, 533-542 Copyright © 1994 Cambridge University Press 533 Comparison in the immunological properties of Borrelia burgdorferi isolates from Ixodes ricinus derived from three endemic areas in Switzerland C. M. HU1, S. LEUBA-GARCIA1, M. D. KRAMER2, A. AESCHLIMANN1 and L. GERN1* 1 Institut de Zoologie, Neuciiâtel University, CH-2000 Neuchâtel, Switzerland, 2Institut für Immunologie, Ruprecht-Karls-Universität, Heidelberg, Germany (Accepted 17 January 1994) SUMMARY Borrelia burgdorferi isolates were obtained from Ixodes ricinus from three sites in Switzerland. They were examined by SDS-PAGE and immunoblotting. The phenotypes, in respect of three outer surface proteins (Osp), differed between the sites of collection. In site 1, most isolates had an OspA of 31 kDa and an OspB of 34 kDa; in site 2, isolates presenting an OspA of 33 kDa dominated and in site 3, the isolates with an OspA of 32 kDa and an OspB of 35 kDa were most frequent. This distribution differed significantly. About half of the isolates from sites 1 and 3 reacted with anti-OspA monoclonal antibody H5332 compared to 29% from site 2. Site 1 isolates reacted significantly more frequently (81 %) with another anti- OspA monoclonal antibody LA-31 than isolates from site 3 (P < 0-0001). These findings have implications for the epidemiology of Lyme borreliosis, for the further development of sérodiagnostic reagents and for the development of a vaccine. INTRODUCTION In humans, Lyme borreliosis is a disease caused by infection with Borrelia, burgdorferi [I]. B. burgdorferi is transmitted by infected ticks belonging primarily to the Ixodes ricinus complex [2]. In Europe, B. burgdorferi can be isolated from infected ticks, animals and patients. The outer membrane of B. burgdorferi contains at least three outer surface proteins (Osp) A (31-33 kDa). B (34-36 kDa) and C (20-24 kDa) [3-6]. These lipoproteins are embedded in the fluid outer membrane of B. burgdorferi and are encoded by linear plasmids [6, 7]. Although their exact functions have not yet been defined, the outer surface proteins of B. burgdorferi are generally thought to have an important role in the host-parasite interactions during the course of infection. The European isolates are more heterogeneous with respect to their antigenic profiles than the American isolates [8-12]. * Author for correspondence: Institut de Zoologie, Chantemerle 22, CH-2000 Neuchâtel. .Switzerland. 534 C. M. Hu AND OTHERS Table t. Characterization of B. burgdorferi isolates (n = 92) Osp Mws (kDa) MoAbs Strains A B C H5332 LA-2 LA-4 LA-31 LA-25 LA-27 LA-7 p3! Site 1 NE83 33 a 22 — — — + — — — + NE84 33 a 22 — — — + — — — — NE85 33 a 22 — — — + — — — — NE 190 31 34 23 — — — + — + — - NE 192 33 a 22 — — — — — — — — NE 193 31 34 — — — + — + — — NE308 33 34 — — — + — — — - NE317 32 34 22 — — — + — — — — NE323 32 35 — — — — — — — — NE 15 31 34 + — — + — + — - NE 12 31 34 22 + + + + + + + + NE14 31 34 + + — — — — — + NEl 96 31 34 + + — + — + — + NE303 32 35 23 — + — + — + — - NE304 32 a 22 — — — — — — — — NE305 31 34 22 — + + + + + - - NE378 33 a 22 — - — + — + — - NE20 32 a 23 + — — + — — — n NE25 31 34 23 + + — + + + + n NE26 33 a 22 + — — + — — — n NE8 32 a 23 + — — + — — — Ii NE 19 32 35 a + — — — — — — n NE550 31 34 a + + — + + + + n NE21 32 a 22 + + — + — — — n NE24 31 34 23 + + + + + + + n NE23 32 a 22 + + - + - - - n Site 2 XE2 33 a 22 — — — + — — — - XE4 33 a 22 — — — — — — — — XE58 33 a a — + — + — — — + NEoO 33 a a — + — + — — — + XE 173 33 a 22 — — — — + — - - XE352 32 34 23 + + — — — + — — XE355 33 a 22 — — — + — — — — XE363 32 35 a — — — — — — — — XE3 33 a a — — — h n n — n XE9 32 a 22 + — — + — — — M XElO 32 a 23 + — — + — — — Il XE22 33 a a — — — + — — — Il XElC 32 a 22 + — — — — — — Il XEl 32 a 22 — — — — — — — Il XE5 33 a 22 — — — — — — — Il XE6 33 a a — — — — + — — Il XE27 32 35 23 + - - - - - - Il Site 3 NE413 33-5 a 22 — — — — — — — — NE443 31 a 23 + — + — — — — — XE450 32 35 23 + — — — — — — + XE454 32 35 21 — — — — — — — — XE456 31 34 22 + + — + + + + + NE4C0 32 35 22 + + — — — — — - Geographical diversity of B. burgdorferi 535 Table 1 {coni.) Osp M ws i (kDa) MoAbs PoAb A A1 anti- Strains A B C H5332 LA-2 LA-4 LA-31 LA-25 LA-27 LA-7 p39 22 kDa NE461 32 35 22 — — — — — — — — + NE462 33 a 23 + — + — — — — — + NE472 32 35 23 + — — — — — — + + NE474 31 a 23 — — + — — - — — + NE477 32 35 22 + — — — — - — + + NE478 32 35 22 — — — — — — — — + NE485 32 35 22 + — — — — — — + + NE490 32 a 22 + — — + — — — — + NE496 32 35 23 + — — — — — — + — NE506 a a 22 — — — — — — — — + NE507 32 35 22 — — — — — — — + + NE517 32 35 22 — — — — — — — — + NE519 32 35 22 — — — — — — — — + NE558 32 a 22 + — — + — — — — + NE601 32 a 21 — — — + — — — — + NE603 32 a 22 + — + — — — — — + NE606 32 35 22 + — — — — — — — + NE607 32 a 22 + — — + — — — + + NE608 32 a 22 + — — + — — — — + NE623 32 35 22 + — — — — — — — + NE624 32 35 22 + — — — — — — + + NE629 32 35 a + — — — — — — + + NE630 32 35 22 + — — — — — — + + NE632 32 35 23 + — — — — — — + + NE418 32 35 a + + — + — — — - — NE426 33 a 22 + + — — — — — — + NE429 32 35 a — + — — — — — — — NE435 32 35 22 — — — — — — — — — NE438 33 a a — + — — — — — — — XE463 31 34 21 + + — + + + + + + NE467 32 35 21 + + — — — — — — + XE471 32 35 22 — + + — — — — — + XE470 32 35 a — + — — — — — — - NE493 31 34 21 + — + + + + + + + NE508 32 35 22 + + — + — — — — + NE537 32 35 22 — + — — — — — — — XE605 32 35 22 — + — — — — — — + XE200 32 a 22 + + — + + + — — + XE207 32 a 23 + + + — + — — — + XE201 32 a a — + — + — - — + - XE202 31 a 22 — + + + — — — - - XE203 32 35 a + + + + + + — + — XE204 32 36 a — — — — — — — — — + , positive réaction; —. negative reaction; a, absent; n. not tested. In this study, B. burgdorferi strains from B. burgdorferi ticks collected in three different endemic areas were screened for evidence of phenotypic differences using immunochemical methods. HVf: 112 536 C. M. Hu AND OTHERS MATERIAL AND METHODS Collection of ticks and isolation of B. burgdorferi J. ricinus ticks (adults and nymphs) were collected b}' flagging lower vegetation using a white cotton flannel flag ( 1 m2) which was dragged behind collectors in three endemic areas: site 1, Bois de l'Hôpital forest close to Neuchâtel; site 2, Staatswald forest and site 3, Karoline forest around Aarberg. Ticks were collected during April-June and August-November 1987 to 1992 in sites 1 and 2, and in 1989 and 1992 in site 3. For isolation of B. burgdorferi, the midgut of the tick was incubated for 10 days at 34 0C in individual culture tubes containing 4 ml BSK II medium [13] supplemented with rifampicin (50/tg/ml) and phosphomycin (50/*g/ml). SDS-PAGE and immunoblol analysis Each isolate was inoculated in 25 ml BSK II medium and after 10 days, the cultures were centrifuged and washed twice with PBS+ 5 mM MgCl2. Whole-cell lysates (equivalent to 107 cells/lane) were separated by SDS-PAGE using a 12-5% poly aery lam ide gel. The gels were stained with Coomassie brilliant blue R, 250 [14]. The separated proteins were transferred onto nitrocellulose paper using a transit cell (2117-250 Nova Blot Electrophoretic Transfer Kit, LKB AM Bromma, Sweden) [14]. The monoclonal antibodies (MoAbs): H5332, LA-2, LA-4, LA-31 (anti-OspA) [8, 15], LA-25, LA-27 (anti-OspB) [15], LA-7 (anti-20 kDa protein) [15], and polyclonal antibodies (PoAbs): anti-22 kDa/NE4 [14] and anti-B31, produced by immunizing a New Zealand white rabbit with strain B31 [9], were used for immuno-blotting. Bound antibodies were visualized by using peroxidase labelled anti-rabbit IgG or anti-mouse IgG antibodies (1:1000, Nordic Immuno- logical Laboratories, The Netherlands). Statistical analyses The Fischer's exact test was used to compare the distribution of the different characterized B. burgdorferi isolates derived from different areas. The difference was considered as significant if P value was < 0-017 [16]. RESULTS Characterization of B. burgdorferi isolates by immunological methods Twenty-six B. burgdorferi isolates were obtained from ticks of the Bois de l'Hôpital forest (site 1), 17 isolates from the Staatswald forest (site 2) and 49 isolates from the Karoline forest (site 3). OspA, OspB and OspC were expressed by 91 (99%), 50(54%) and 70(76%) of the 92 isolates (Table 1). Four different phenotypes could be distinguished on the basis of the expression pattern of the Osps, namely, A (11% of isolates), AB (16%), AC (41%) and ABC (32%). Péter and colleagues [11] distinguished four typing groups of B. burgdorferi (I, II, III and IV) according to the molecular weight of OspA and OspB (Fig. Ia). Most of our isolates belong to groups I, II and III and only one, from site 3, is in group IV. Seven isolates did not fit into this classification and comprised three additional groups: V, VI and VII (Fig. Ia). Geographical diversity of B. burgdorferi 537 («) HkDi 3SkDa 1A5.1 '32-JIkDa MkDa 31 kDi Ì2 kD» ' 33 5 kDa 32 kDa 31 kDa gr. I gr. II gr. Ill gr. IV gr. V gr. VI gr. VII (b) Site 1 isolates (N = 26) gr. IV-VII 8% Site 2 isolates (JV = 17) gr. Ill I 42% gr. IV-VII 6% gr- II 12% gr. I gr.III 82% Site 3 isolates (N = 49) gr. IV-VII .0% Sr%I gr- HI I J1[K . 24% I l|\0 s I JlJ1I)W--j gr. II 64% Fig. 1 a and b. Typing groups of B. burgdorferi isolates and their geographical distribution, (a) Schematic description of the typing groups according to the molecular weights of OspA and OspB. The first four groups (gr. I-IV) were suggested by Péter and colleagues [11] and the groups V-VII were proposed by us. Group VII comprised one isolate (NE506) which did not express OspA and OspB. (b) Relative distribution of the different typing groups in the three sites of isolation : groups I and 111 prevailed in site 1, group III or II prevailed in site 2 and site 3, respectively. Immunological characterization of these isolates revealed a heterogeneous reactivity with the different antibodies used (Table 1). The different phenotypes obtained from April to November did not show any differences in their seasonal distribution (data not shown). Distribution of the cliaracterized isolates in the three studied sites The isolates from the three different geographic sites of isolation were compared for: (1) the apparent molecular weights (Mw) of OspA, OspB and OspC; (2) their classification into the respective typing groups; (3) differences in the Osp phenotypes; (4) their reactivity with mono- and polyclonal antibodies specific for B. burgdorferi antigens. The apparent molecular weights of OspA, OspB and OspC The results are summarized in Table 2. OspA was present^ in nearly all isolates; 538 C. M. Hu AND OTHERS Table 2. The presence of the major proteins in isolates derived from, three different. sites OspA OspB OspC 31 kDa 32 kDa 33 kDa 34 UDa 35 kDa 36 kDa 21 kDa 22 kDa 23 kDa Site 1 10 9 7 12 3 0 0 12 6 (JV =26) (38%) (35%) (27%) (46%) (12%) (0%) (0%) (46%) (23%) Site 2 07 10 1200 8 3 (JV= 17) (0%) (41%) (59%) (6%) (12%) (0%) (0%) (47%) (18%) Site 3 6 38 3 3 29 1 5 28 8 (JV = 49) (12%) (78%) (6%) (6%) (59%) (2%) (10%) (57%) (16%) Statistical results Pl* 00072 0-7521 00565 00062 10000 10000 10000 1-0000 0-7223 P2* 00158 00004 00170 000008 000007 10000 01567 0-4669 05398 P3* 01889 00081 000002 10000 00014 10000 03165 05753 10000 * Pl : was as compared between site 1 isolates to site 2 isolates. P2 : was between the site 1 isolates to the site 3 isolates. P3 : was between the site 2 isolates to site 3 isolates. Bold type: significant differences between both. however, they varied in molecular weight with the 31, 32 and 33 kDa protein predominating in isolates from sites 1, 3 and 2 respectively (Table 2). OspB was more prevalent among isolates from sites 1 and 3 than site 2 ; the 34 IcDa protein predominated in site 1 and the 35 kDa protein in site 3. OspC was detected in 69, 65 and 83% of isolates from sites 1, 2 and 3 respectively with the 22 kDa protein being the most prevalent in all three sites. The distribution of typing groups in different areas Between the sites of isolation, the distribution of the main typing groups differed significantly (Fig. 16). In site 1, group I and III were most frequent (38 and 42%) and group I was more prevalent than in site 2 (0%, P = 00072) and in site 3 (3%, /' = 0-0008). Group III was the main group in site 2 (87%). Its presence differed significantly from site 1 (42%, /3 = 0-0058) and site 3 (24%, P = 0-00004). Group II was more frequent in site 3 (60%) than in site 1 (12%, P = 0-00007) and site 2 (12%, P = 00014). About 8% of isolates from each site belonged to groups IV, V, VI or VII. 71Ae distribution of the different Osp phenotypes Differences in the distribution of the Osp phenotypes according to the expression pattern of the different Osps were observed between the three areas studied (Fig. 2). Phenotype AC (OspA+ OspC) was the most frequent in sites 1 and 2, whereas phenotype ABC (OspA+ OspB+ OspC) predominated in site 3 (54%) and was less frequent in site 2 (12%) than in site 3 (P = 00038). Phenotype A (OspA only) was more prevalent in site 2 (29%) than in the other sites. The immuno-reactivity with different MoAbs and PoAbs About half of isolates from sites 1 and 3 reacted with MoAb H5332. The isolates Geographical diversity of B. burgdorferi 539 Site 1 isolates (JV = 26) Site 2 isolates (N = 17) Site 3 isolates (N = 49) ABC 31% *% mììSl\ Fig. 2. Relative distribution of the Osp phenotypes (in percentage) in B. burgdorferi isolates derived from the three sites of isolation. A, expression of OspA only; AB, expression of OspA and OspB; AC, expression of OspA and OspC; ABC, expression of OspA, OspB and OspC. Table 3. The reaction of B. burgdorferi isolates derived from three different sites with specific mono- and polyclonal antibodies MoAbs PoAb H5332 LA-2 LA-4 LA-31 LA-25 LA-27 LA-7 anti-22 kDa Site 1 13 10 3 21 5 11 4 11 (N = 26) (50%) (30%) (12%) (81%) (19%) (42%) (15%) (42%) Site 2 5 3 0 7 2 1 0 8 (,V= 17) (29%) (18%) (0%) (41%) (13%) (6%) (0%) (50%) Site 3 29 18 9 14 6 5 3 37 (N = 49) (59%) (37%) (18%) (29%) (12%) (10%) (6%) (76%) Statistical results Pl* 0-2194 01874 0-2658 00200 0-6897 00149 01498 10000 /'2* 0-4733 10000 0-5256 000002 0-4984 00023 0-2270 00059 /»3* 00490 0-2273 00977 0-3566 10000 10000 0-5692 00386 Pl : was as compared between the site 1 isolates to site 2 isolates. P2: was between the site 1 isolates to site 3 isolates. F3 : was between the site 2 isolates to site 3 isolates. Bold type: significant differences between both. MoAbs: monoclonal antibodies. PoAbs: polyclonal antibodies. from site 1 reacted more frequently (81 %) with MoAb LA-31 (anti-OspA of B31) than isolates from site 3 (29% ; P = 0-00002) (Table 3). Reactions of the isolates with MoAb LA-27 (anti-OspB) were more frequent in site 1 (42%) as compared to site 2 (6%) and site 3 (10%); the differences were statistically significant (P = 00149 and P = 00023). In contrast, 76% of site 3 isolates reacted with anti- 22 kDa/NE4 (PoAb) and this was significantly different from isolates from site 1 (42%, P = 00059). To reveal the number B31-like strains in each area, we also compared our isolates with the B. burgdorferi strain B31, a prototype strain isolated from I. dammini. Five of the site 1 isolates had the same Osp phenotype as B31. Five additional site 1 isolates and three site 3 isolates, all of which expressed the OspC 540 C. M. Hu AND OTHBBS protein, were similar to the B31 strain in the protein Mw of their OspA (31 kDa) and OspB (34 kDa). However, only two of the site 1 isolates displayed the same reactivity as B31 with the MoAbs described by Barbour and colleagues [8], Kramer and colleagues [15] and Wallich and colleagues [12]. DISCUSSION Our study confirms the antigenic heterogeneity of the European B. burgdorferi isolates [8, 9, 11, 12]. In this study we compared isolates from ticks from three endemic areas and showed significant differences in the distribution of the different phenotypes among the isolates of these sites. According to the molecular weight of OspA and OspB (Fig. 2), each endemic area presented a main typing group which differed significantly from the two other sites : group I prevailed in site 1, group III in site 2 and group II in site 3. A recent study demonstrated that the molecular weights of OspA and OspB and the representative phenotypes of borrelia isolates from patients with disseminated Lyme borreliosis were different from those obtained from patients with the cutaneous form of Lyme borreliosis [20]. It was reported that most skin isolates presented proteins of 32 kDa (OspA) and 35 kDa (OspB) (group II in our study), whereas the isolates from the disseminated Lyme borreliosis patients expressed an OspA of 32-5 kDa and an OspB of 33-34 kDa (groups III and V in our study). In view of our results, it may be suggested that the clinical manifestations of Lyme borreliosis may differ in different geographical areas. This hypothesis remains to be confirmed. Immunoreactivity of the isolates varied between the different sites of isolation. Site 1 isolates reacted most with MoAbs LA-31 and LA-27. Site 3 isolates reacted most frequently with the MoAb H5332 and PoAb anti-22 kDa/NE4. In addition, the frequency of isolates reacting with MoAb LA-2, a MoAb which recognizes a protective epitope against B. burgdorferi infection [21], was different in each site. On the other hand, the comparison of our isolates with the strain B31 showed that only two isolates presented exactly the same reactivity with the MoAbs described by Barbour and colleagues [8], Kramer and colleagues [15] and Wallich and colleagues [12]. In view of this, it is suggested that European isolates could elicit specific antibody responses during infection, which differ from one site to another site and which is different from that induced by the strain B31. Therefore, a correct selection of the antigen or antigens seems to be necessary for the serodiagnosis of Lyme disease. The different protein profile and immunoreactivity of the isolates from different geographical locations may account for seroconversion in people who do not develop Lyme borreliosis, as described in Aarberg (site 3, in this study) where 26 % of an asymptomatic population was seropositive by ELISA and Western blotting [22,23]. Among isolates from the Aarberg area (site 3), 83% expressed the 21-23 kDa proteins and 76% reacted with PoAb anti-22 kDa/NE4 and this was significantly different from the isolates from site t. Several studies showed that these proteins are important immunogens which elicit the antibody response early after the tick bite [5, 17-19]. It remains to be elucidated whether these proteins could be responsible for the presence of asymptomatic -seropositive people in Geographical diversity of B. burgdorferi 541 Aarberg by eliciting a protective antibody response. Another explanation could be that some of the strains present in this area were less pathogenic or non- pathogenic and that could depend on their antigenic profiles. Heterogeneity observed among B. burgdorferi isolates from different sites which are fairly close to each other may have implications for human health if some strains are capable of producing early or late manifestations of Lyme borreliosis, on serodiagnosis and also on the production of protective vaccine since OspA and a 22 kDa protein are actually vaccine candidates. In fact, several studies showed that these antigens elicit protective antibody responses in animal models [21,24-26]. The reasons leading to such a geographic diversity remain unknown but could be due to differences in the reservoir hosts in these three areas. However, our results suggest that studies on the local distribution of B. burgdorferi strains in each endemic area should represent an important step toward understanding the epidemiology of Lyme borreliosis, toward improving serological testing and toward developing an efficient vaccine for populations exposed to bites by infected ticks. ACKNOWLEDGEMENTS This work is part of the PhD thesis of one of the authors (Hu C. M.) and it was supported by the Swiss National Science Foundation. We thank Jacqueline Moret for assistance in statistical analysis, Alan Barbour (University of Texas. San Antonio. USA) for providing monoclonal antibodies, Olivier Rais for technical assistance and Larry Kendall for advice on the manuscript. REFERENCES 1. Steere AC, Bartenhagen NH, Craft JE, et al. Clinical manifestations of Lyme disease. Zentralblatt für Bakteriologie, Mikrobiol Hyg 1986; A263: 201-5. 2. Burgdorfer W. Discovery of the Lyme disease spirochete and its relation to tick vectors. Yale J Biol Med 1984; 57: 515-20Ï 3. Barbour AG, Tessier SL, Todd WJ. Lyme disease spirochetes and ixodid tick spirochetes share a common surface antigenic determinant defined by a monoclonal antibody. Infect Immun 1983; 41: 795-804. 4. Barbour AG, Tessier SL, Hayes SF. Variation in a major surface protein of Lyme disease spirochete. Infect Immun 1984; 45: 94-100. 5. Wilske B. Preac-Mursic V, von Busch K. Immunochemical and immunological analysis of European Borrelia burgdorferi strains. Zentralblatt Bakteriol Hyg 1986; 263: 92-102. 6. Fuchs R. Jauris S, Lottspeich F, Preac-Mursic V, Wilske B. Soutschek E. Molecular analysis and expression of a Borrelia burgdorferi gene encoding a 22 kDa protein (pC) in Escherichia coli. Molec Microbiol 1992; 6: 503-9. 7. Barbour AG, Garon CF. Linear plasmids of the bacterium Borrelia burgdorferi have covalently closed ends. Science 1987; 237: 409-11. 8. Barbour AG, Heiland RA, Tessin HR, Heterogeneity of major protein in Lyme disease borrelia: a molecular analvsis of American and European isolates. J Infect Dis 1985: 152: 478-84. 9. Wilske B, Preac-Mursic V, Schierz G. Kiihbeck R. Barbour A(J. Kramer MD. Antigenic variability of Borrelia burgdorferi. Annals XY Acad Sci 1988; 593: 126-43. 10. Rosa PA, Hogan DM. Colony formation by Borrelia burgdorferi in solid medium clonal: analvsis of Osp locus variants. First International Conference on Tick-Borne Pathogens at the Host-Vector Interface: An Agenda for Research. Saint Paul, Minnesota. USA. 1992; 95 103. 542 C. M. HU AND OTHERS 11. Péter O, Bretz AG. Polymorphism of outer surface proteins of Borrelia burgdorferi as a tool for classification. Zentralblatt Bakteriol 1992; 277: 28-33. 12. Wallich R1 Moter SE, Kramer MD, et al. Untersuchungen zur genotypischen und phänotypischen Heterogenität von Borrelia burgdorferi, dem Erreger der Lyme-Borreliose Infection. In : Hassler D, Kramer M, Maiwald M, Marget W, Zöller L, eds. Forschritte der Infektiologie. München: MMV Verlag 1992: 176-91. 13. Barbour AG. Isolation and cultivation of Lyme disease spirochete. Yale J Biol Med 1984; 57:521-5. 14. Hu CM, Gern L, Aesehlimann A. Changes in the protein profile and antigenicity of different Borrelia burgdorferi strains after reintroduction to Ixodes ricinus. Parasite Immunol 1992; 14: 415-27. 15. Kramer MD, Schaible UE, Wallich R, Moter SE, Petzoldt D, Simon MM. Characterization of Borrelia burgdorferi associated antigens by monoclonal antibodies. Immunobiol 1990; 181: 357-66. 16. Kirk, R. Experimental design. Procedure for behavioral sciences. Belmont, California: Brooks, 1968. 17. Wilske B, Preac-Mursic V, Schierz G, Liegi G, Gueye W. Detection of IgM and IgG antibodies to Borrelia burgdorferi using different strains as antigen. Zentralblatt Bakteriol, Mikrobiol Hyg 1989; Suppl. 18: 299-309. 18. Ma B, Christen B, Leung D, Vigo-Pelfrey C. Serodiagnosis of Lyme Borreliosis by Western iminunoblot : reactivity of various significant antibodies against Borrelia burgdorferi. J Clin Microbiol 1992; 30: 370-6. 19. Gern L, Schaible UE, Simon MM. Mode of inoculation of the Lyme disease agent Borrelia burgdorferi influences infection and immune responses in inbred strains of mice. J Infect Dis 1993; 167: 971-5. 20. van Dam AP, Kuiper H, Vos K, et al. Different genospecies of Borrelia, burgdorferi are associated with clinical manifestations of Lvme borreliosis. Clin Infect Dis 1993; 17: 708-17. 21. Schaible UE, Kramer MD, Eichmann K, Modolell M, Museteanii C, Simon MM. Monoclonal antibodies specific for the outer surface protein A (OspA) of Borrelia burgdorferi prevent Lyme borreliosis in severe combined immunodeficiency (seid) mice. Proc Nat Acad Sci USA 1990;87:3768-72. 22. Gern L, Frossard E, Walter A, Aesehlimann A. Presence of antibodies against Borrelia burgdorferi in a population of the Swiss Plateau. Zentralblatt Bakteriol, Mikrobiol Hyg 1989; Suppl. 18: 321-8. 23. Gern L, Garcia S, Frossard E. Characterization and follow-up the IgG antibody response against Borrelia burgdorferi using Western blot in a seropositive (ELISA) population from an endemic area. Bull Soc Neuchâteloise Sciences Naturelles 1993; 116: 5-14. 24. Fikrig R, Barthold SW, Kantor FS, Flavell RA. Protection of mice against the Lyme disease agent by immunizing with recombinant OspA. Science 1990; 250: 553-6. 25. Fikrig R, Barthold SW, Kantor FS. Flavell RA. Long-term protection of mice from Lyme disease by vaccination with OspA. Infect Immun 1992; 60: 773-7. 26. Simon MM, Schaible UE, Kramer MD, et al. Recombinant outer surface protein A from Borrelia: burgdorferi induces antibodies protective against spirochetal infection in mice. J Infect Dis 1991; 164: 123-32. Parasite Immunology 1992. 14, 415-427 Changes in the protein profile and antigenicity of different Borrelia burgdorferi strains after reintroduction to Ixodes ricinus ticks CHANG MlN HU, LISE GERN & ANDRÉ AESCHLIMANN Institute of Zoology, Chan temerle 22. 2000 Neuchâtel. Switzerland Accepted for publication 3 March 1992 Summary Eight Swiss strains of Borrelia burgdorferi, with various protein profiles and the North-American strain B31 were artificially introduced into Ixodes ricinus ticks and reisolated IO days later. All isolates were subsequently examined by SDS-PAGE analysis. Comparing initial isolates with the rcisolates. we observed that 7 out of 9 strains changed their protein pattern with respect to the major proteins OspA. OspB and the 22 kDa protein after passage in the tick. The strains NE2. NE4 and NE83 with the initial phenotypc of OspA and 22 kDa proteins changed to the phenotypc of OspA and OspB. the strains B2 and NE202 with the initial phenotypc of OspA acquired an additional protein of 22 kDa and the strain NE58 with the initial phenotypc of OspA also acquired a protein of 22 kDa. Examination of these isolates by Western blot analysis demonstrated that the reaction with the monoclonal antibody H5332 and a monospecific polyclonal antibody PoAb/anti-22 kDa differed between the initial isolates and the rcisolates. Keywords: Ixodes ricinus. Borrelia burgdorferi, antigen, electrophoresis Introduction Borrelia burgdorferi, the causative agent of Lyme Borreliosis. is directly inoculated into the vertebrate by a tickbitc. This means that the spirochaetes spend part of their life cycle within the arthropod, an environment strikingly different from that found within the vertebrate hosts. The necessity for micro-organisms to adapt to different environmental conditions has often been underestimated in laboratory research. In Europe, as well as in the USA, many B. burgdorferi isolates have been established from ticks and vertebrates. The two outer surface proteins OspA (3 I kDa) and OspB (34 kDa). of the North American isolates are more homogeneous than the European ones which has been demonstrated with various monoclonal antibodies (MoAbs) specific to these proteins by Western blot analysis (Barbour & Schrumpf 1986. Barbour, Heiland & Tcssin 1985, Barbour, Tessici" & Hayes 1984. Kramer et al. 1990. Wilskc el al. 1986). In some of the European isolates. OspA and OspB could not be detected. Moreover, a Correspondence: Use Gern. 415 416 CM.Hu et al. number of them showed another major protein band of approximate molecular weight of 20-23 kDa, designated 'pC (Barbour et al. 1985, Wilske et al. 1986, 1988). Most of the studies on B. burgdorferi have been done after serial in vitro passages of the spirochaetes. Under these conditions, changes in protein profile and antigenicity as well as the loss of infectivity to animals have been reported in a tick isolate (Schwan and Burgdorfer 1987, Schwan, Burgdorfer & Garon 1988). To determine if the passage into in vivo models of in vitro cultivated strains, influences (a) the protein profile and antigenicity of the strains, and (b) the pathogenicity and infectivity of the strains, we reintroduced various isolates of B. burgdorferi to Ixodes ricinus, the most important vector in Europe (Barbour et al. 1983c). If such modifications occurred within ticks, then this passage into the arthropod, which represents a prerequisite condition for the transmission to human and animal host, may be of great importance in the epidemiology of Lyme borreliosis. We present data which suggest changes in the protein profiles and in the antigenicity of the strains during tick passages. The influence of the tick on the pathogenicity of B. burgdorferi will be the subject of another paper. Materials and methods STRAINS B. burgdorferi strains NE2, NE4, NE56, NE83, NE202. NE203 were isolated from I. ricinus tick midguts incubated individually in BSK II medium (Barbour et'al. 1985) Ticks were collected from the Neuchâtel area of Switzerland. The isolale B2. derived from synovial fluid of a Lyme disease patient (kindly provided by Dr J. Schmidli. Bern Hospital, Switzerland) (Schmidli et al. 1988) and the North American strain B31 were examined in parallel. These strains were maintained in our laboratory in BSK II culture medium as described by Barbour el al. (1985). Strains B3I and NE4 are high-passage strains (more than 100 in vitro passages) and the other strains were low-passaged (less than 6 passages). None of these strains developed changes after serial //; vitro passages. Cl.ONlNCi Oh lì. IiURGPORFI-IRI To isolate clonal populations of B. burgdorferi, the two strains NFÌ2 and NE83 (107 cells/ ml; Hclbcr count cell chamber) were serially diluted 10-fold from 10 :' to 10 s in BSK 11 medium (Bundoc & Barbour 1989). Three lubes containing 5 ml BSK Il were used for each dilution. All dilutions were examined after 1. 2 and 4 weeks of incubation using dark field microscopy. The spirochaetes were tested by SDS-PAGE as described later. REINTRODUCTION Ol- B. BURGDORFERI TO I RICISUS AND RHISOI.ATION The borrelia isolates ( 10s cells/ml) were reintroduced to uninfected I. ricinus females bred in our laboratory (Graf 1978). by the capillary method (Gern. Zini & Aeschlimann 1990. Monili, Gern & Aeschlimann 1989). One weck after infection. B. burgdorferi was reisola led from the tick midgut and maintained in BSK Il medium al 34 C for 7 lo 10 days. Each reisolate was named by adding the letter R to the parental name. e.g.. R2NE58 designates the reisolatc number 2 from strain NE58. Changes in the protein profiles of B. burgdorferi 417 SDS-PAGF. All isolates and their reisolates were incubated in 25 ml BSK II medium. Ten days later, the borrelia cultures were centrifuged and washed twice with PBS +5mM MgC12 (10000 jç., for 20 min at 20°C). For Polyacrylamide gel electrophoresis (SDS-PAGE), whole cells were suspended in 15 p\ distilled water and resuspended in sample buffer to give a protein concentration of 30 //g/Iane (Wilske et al. 1988) or 107 cells/lane. The pH of the separating gel buffer was 8-8 and the acrylamide concentration was 12-5%. The gels were stained with Coomassie brilliant blue R-250. Molecular weight standards were low range specification of Bio-Rad"s SDS-PAGE standards. ANTIBODIES The monoclonal antibodies H9724 specific for the flagellili and H5332 specific for the OspA of B. burgdorferi strain B31 were obtained from Alan Barbour (Barbour, Tessier& Todd 1983a). The polyclonal monospecific antibody (PoAb) anti-22kDa/NE4 was produced by immunizing a New Zealand white rabbit with the 22 kDa protein of strain NE4 as described by Wilske et al. (1988). WKSTKRN BLOT ANALYSIS Proteins of whole cell lysates. separated by SDS-PAGE (15 //g/lane) (Wilske et al. 1988) were transferred to nitrocellulose in a transit cell (2117-250 Nova Blot Electrophoretic Transfer Kit, LKB AB Bromma, Sweden). Blots were incubated with rabbit specific antibodies(l :200), monoclonal antibodies(MoAbs) H5332(l ; 10)(Barboureta/. 1983a), H9724 ( 1:200) or polyclonal monospecific antibodies (PoAbs) anti-22 kDa/NE4 ( 1:200). The immunocomplexes were detected using peroxidase labelled anti-rabbit IgG anti- bodies oranti-mouse IgG antibodies (I : 1000: Nordic Immunological Laboratories, The Netherlands). Results Sl)S-PAGE Eight B. burgdorferi strains from Switzerland and the North American isolate B31 were subjected to SDS-PAGE (Figure 1). The protein patterns of these isolates were heterogeneous. The strain NE58 presented only one major molecular mass of 33 kDa and the strain NE202 contained a 31 kDa protein and in addition one with a protein mass of 22 kDa. Some of the isolates showed similar major surface proteins, for example, the three strains NE56, B2 and NE203 expressed the protein pattern as the strains B31. and the strains NE2, NE4 and the NE83 expressed the 33 kDa and the 22 kDa proteins. These isolates were introduced to I. ricinus ticks and reisolated (Table 1). The reisolates were again examined by SDS-PAGE. Some showed different protein profiles after passage in I. ricinus. Therefore, we grouped the isolates according to their major surface proteins before and after passage into ticks (Table 2). Striking modifications were observed in the molecular weight ranges of the outer surface proteins A or B. The initial isolates included in group I presented two major 418 CM.Hu et al. 97-4 66-2 42- 7- 31-0 12345 6789 '1 i- i -» «4 • * Figure I. Eight Swiss strains of R. burgdorferi and North American isolale B3I were separated b\ SDS-PAGE and stained with Coomassie Blue R-250. Lanes I to 9: 1331. NE56. B2. NE203. N 1:2. NE4. NEX3, NE202. NE58. Molecular weight standards (MWS: kDa) are indicated on the left. proteins with molecular masses of 31 kDa and 22 kDa. or 33 kDa and 22 kDa. All reisolates from the tick midgut were found to have lost the 22'kDa. Some of them have shifted the 33 kDa band. In three reisolates a 34 kDa band and in two reisolates a 31 kDa band appeared. Thus, through the passage in tick, the strains NE4 and N 1:2 changed into another phenotype with respect to OspA and OspB ( B31 -like). The protein patterns of the two strains NE202 and NE83 remained unchanged as far as the proteins of 31 and 33 kDa were concerned (Figure 2). The initial isolates of group II presented proteins with molecular masses of 3 I and 34 kDa. Reisolates of these strains showed an additional protein band of 22 kDa. The major surface proteins OspA and OspB remained unchanged (Figure 2). Only one strain (NE58) was included in group III. The main protein of the initial isolate revealed by SDS-PAGE is a 33 kDa protein, perhaps OspA. which remained unchanged after tick passage. In contrast, a new protein band with a molecular mass of 22 kDa appeared in the reisolates (Figure 2). Tabic I. Borrelia isolates and their reisolates Strains No. of ticks infected IO Il IO Il IO S 7 7 19 No. of reisolates 5 5 6 9 7 4 4 .1 7 Ratio of chanaed reisolates 2 5 3 5 0 4/9 3 7 2 4 1 4 2 3 0 Changes in the protein profiles of B. burgdorferi 419 Table 2. Groups of Borrelia isolates and their major surface proteins Groups Strains Initial MSP* Reisolate's MSP Rate Group I NE2 33 kDa. 22 kDa 31 kDa, 34 kDa 2/5 NE4 33 kDa. 22 kDa 3IkDa. 34 kDa 3/5 NE83 33 kDa. 22 kDa 33 kDa, 34 kDa 3/7 NE202 3IkDa. 22 kDa 31 kDa 2/4 Group 11 B2 3IkDa. 34 kDa 31 kDa. 34 kDa. 22 kDa 2/3 NE203 3IkDa, 34 kDa 31 kDa, 34 kDa, 22 kDa 1/4 Group 111 NE58 33 kDa 33 kDa. 22 kDa 4/9 Group IV B3I 31 kDa. 34kDa, 22 kDat 31 kDa, 34 kDa. 22 kDat 0 NE56 31 kDa, 34 kDa, 23 kDat 31 kDa. 34 kDa. 23 kDat 0 * MSP: Major surface protein. t Trace on the SDS-PAGE. Bold type: lost or acquired proteins in reisolates. All reisolates of the strains NE56 and B3I gathered in group IV did not show any changes with respect to the proteins with molecular masses of 3 1, 34 and 22 kDa to 2.3 kDa (Figure 2). In addition, we repeated these experiments using two reisolates of NE58, i.e., R2NE58 (Figure 3a). a modified reisolate (appearance of a new 22 kDa protein) and R5NE58 (Figure 3b) an unmodified reisolate comparable to NE58. The second passage into 1. ricinus did not influence R2NE58: the resulting reisolates presented the same protein pattern as R2NE58. In contrast, the passage of R5NE58. ended up with 2 out of 7 reisolates showing modifications (appearance of a new 22 kDa protein) comparable to the modifications presented by R2NE58 after the first passage into ticks. (Figures 3a and 3b). \vksti-:rn hi.ot To investigate changes of the antigenic determinants, B. burgdorferi isolates and their modified reisolates were examined by Western blot. Western Blot analyses using monoclonal and polyclonal antibodies to OspA, flagellili and the 22 kDa protein demonstrated differences of the B. burgdorferi isolates before and after their passage in ticks. Loss of the 22 kDa protein of the group I reisolates resulted in the lack of staining with the PoAb specific for the 22 kDa protein (PoAb/anti 22 kDa,'NE4) (Table .3). Moreover, th c reaction of the three modified reisolates from NE4 (R5NE4; R7NE4 and R9NE4) with MoAb FI5332 specific for OspA changed from negative to positive staining (Table 3). On the other hand, the additional new protein of 22 kDa of the modified reisolates from groups Il and III reacted with the PoAb22kDa/NE4(Table 3). In addition,OspA of strain NE58 did not react any more with MoAb H5332 after reisolation from the tick midgut (Table 3). 420 CM.Hu et al. All strains examined in this study did not change their reactivity with MoAb H9724 specific for the flagellimi associated protein of B. burgdorferi, after tick passage (Table 3). To exclude the possibility that a simple cloning effect was responsible for the rcisolation of modified spirochaetes, we cloned the two strains NE2 and NE83 (107 cells/ ml (both strains had lost the 22 kDa protein after passage into ticks) by limiting dilution from 10 Mo IO "8. After cultivation for 4 weeks, NE2 spirochaetes were found inali tubes at the 10 6 dilution, in 2/3 tubes at the 10-7 dilution and in 1/3 tubes at the IO-8 dilution; NE83 spirochaetes were found in 2/3 tubes at the IO-6 dilution, in 2/3 tubes at the 10"7 dilution and in 1/3 tubes at the IO-8 dilution. All the clones obtained were examined by SDS-PAGE analysis (data not shown). Compared to the uncloned strains NE2 and Reisololes NE4 I 2 3 Reisololes NE2 I 2 97-4 66-2-1 ' ..„ 3 ^" 31-0- 21-5- 14-4- Reisolates NE83 I 2 3 =1 "i :1 ^ Cf ran-. im •-• 31-0H U ^ 21-5 97-4 66-2 42-7 14-4 (a) 37-4 66-2 Reisololes NE202 I 2 42-7J 31-0- 21- 5 14-4 Reisololes B2 I 2 l'i H3 1^» NE203 Reisolate NE58 Reisololes 12 3 4 ** ¦.» 97-4 66-2- 42-7 31-0 2I-5-- 14-4 97-4- 66-2 42-7 31-0- 21-5- 14-4 U kiitìS (D ) (C ) Changes in the protein profiles of B. burgdorferi 421 97-4. 66-2 42-7- 31-0 21-5- Reisolotes 831 I 2 3 4 5 6 7 e' LA tri Li bd tat Li nnccBP 97.4- 66'2- 42-7 31-0 21-5 Reisolotes NE56 12 3 4 5 6 S a a 3 pf H B •w _, Jt ~i •-» — *J « -¾ « Figure 2. Coomassic Blue stained Polyacrylamide gel of different B. burgdorferi strains and some rcisolates. Figures 2a. b, c (opposite) and d concern respectively group I. II. Ill and IV of borrclia isolates. The initial isolate is on the first lane and the rcisolates on the others. Molecular weights are indicated on the left. The arrows indicate the modified proteins after passage into I. ricinus. Reisolotes R2NE58 I 23456 789 IO 97-4- 66-2- 42-7- 31-0 ,---------------, _ - - - TT'» ,- ( 0 ) Reisolotes R5NE58 I 2 3 4 5 6 7 r* § irlifff ài H «f * ,^ «s «* IW WE: »t% **- r ^J .*» «• M «• Ü M * t H» , * • ( D ) Figure 3. Coomassic Blue stained Polyacrylamide gel oftwo reisolatcs R2NE58. R5NE58 and their rcisolatcs from tick midgut, (a) R2NE58 and its 10 rcisolates (lanes 1-10). (b) R5NE58 and its 7 rcisolatcs (lanes 1-7). Two rcisolatcs of R5NE58 modified their protein profiles after the second passage into I. ricinus as indicated by the arrows. 422 CM.Hu et al. Table 3. Reaction of borrclia isolates with different antibodies Strains PoAb/anti- MoAb/anti- MoAb/anti- 22 kDa/NE4 OspA/H5332 fiagcllin/H9724 Group NE2 -t- R3NE2 - R4NE2 - NE4 •f- R5NE4 - R7NE4 - R9NE4 - NE83 4- R3NE83 - R4NE83 - R5NE83 - NE202 + RINE202 - N5NE202 - B2 __ RIB2 + R3B2 + NE203 _. R3NE203 •1- NE58 _ R2NE58 -t- R3NE58 + R4NE58 ¦)¦¦ _ + I - + — + — + I + + -f + + + + * + + + + + + + + ¦ + + + + + + + * Results obtained bv B.Wilskc. N[:83. no differences were observed after cloning. The 22 kDa protein did not disappear suggesting that what happened in tick midguts is not a phenomenon of cloning. Then, we repeated our experiment and infected 20 ticks with the cloned strain NE83 (at dilution K)"8), named cNE83. Eleven licks were dissected I week after infection to reisolate the spirochactes from different organs: midgut, salivary glands and synganglion. Twenty one rcisolalcs were obtained and were examined by SDS-PAGE. Results showed that 2 midgut and 3 salivary glands reisolates (from 5 different ticks) present now only a weak band of 33 kDa protein suggesting that some modification may also occur in reisolates from cloned strain after passage into ticks (Figure 4). No change was observed with respect to the 22 kDa protein. Changes in the prolein profiles of B. burgdorferi 423 Figure 4. The cloned strain cNE83 (lane 1 ) and its reisolalcs from 5 different ticks were examined by SDS-PAGE. Two reisolalcs from tick midguts (Rm 1 -cNE83. Rm2-cNE83: lanes 2 and 3) and three rcisolatcs from tick salivary glands (RsI-cNE83. Rs2-cNE83 and Rs3-cNE83: lanes 4, 5 and 6) were modified after passage into ticks. The arrows indicate the modified proteins and the molecular weights (MW: kDa) arc indicated on the left. Discussion The heterogeneous protein patterns observed in the Swiss isolates correspond to those described for other European borrelia isolates (Barbour el al. 1985, Wilske el al. 1986, 1988). Their phenotypes differed in the amount and the molecular weight of their OspA (31 kDa) and OspB(34 kDa) proteins. In addition a protein of 22 kDa was expressed in 5/ 8 isolates. In this study, 8 Swiss strains and the North American strain B3I were artificially introduced to I. ricinus. One week later, they were reisolated from the tick midgut, examined by SDS-PAGE and Western blot and compared to the initial strains. In most of the cases, modifications of the protein profiles were observed. The major surface proteins OspA (31-32 kDa) and OspB (34-35 kDa). and a protein with a molecular mass of 22 kDa, were mainly involved in these changes. The protein band of 22 kDa was either 'lost' or 'acquired' in the reisolates. These modifications were accompanied by shifts of the molecular masses of the 33 kDa protein 424 C.M.Hu et al. and in some reisolates a 31 kDa and a 34 kDa protein were present. Thus reisolated spirochaetal populations contained new phenotypes expressing different surface proteins through the influence of residence in the tick midgut. In fact, the changes of OspA and the 22 kDa protein found by SDS-PAGE were supported by Western blot analysis which revealed loss or appearance of epitopes detected by specific monoclonal and polyclonal antibodies. Previous reports suggested that clonal polymorphisms of the major outer surface antigen B (OspB) of B. burgdorferi occurred (Bundoc & Barbour 1989). The modifica- tions obtained in the tick reisolates could be explained by such a clonal polymorphism but not by a cloning effect. This is suggested by the fact that the strains NE2 and NE83 lost their 22 kDa protein after tick passage, a phenomenon which was not observed after in vitro cloning. Reisolates obtained from an uncloned (NE83) and a cloned (cNE83) strain presented different protein patterns after tick passage: RNE83 lost the 22 kDa protein and presented with 33 kDa protein whereas the contrary was observed with cRNE83. This could be explained by a selection, through cloning, of a variant which after passage into ticks is more susceptible to express the 22 kDa protein than the 33 kDa protein. A comparable phenomenon was described by Bundoc & Barbour (1989): an increase in expression of the 21 or 18-5 protein was associated with a decrease or hailed production of OspB in different variants. Additional experiments using cloned strains are in process. The number of in vitro passages of the strains did not influence the occurrence of a change. This occurrence seems to depend on the strain (e.g. B3I and NE56 did not change) and on the ticks (all the reisolates from the same starting population did not always indicate change in the isolate). Changes are repeatable and constant for the same starting population as it was shown: reisolates of NE58 were either modified (appearance of a 22 kDa protein) or unmodified; if such an unmodified reisolalc is introduced again to ticks, this induces a modification comparable to thai presented by the reisolates which were modified after the first passage. Why changes occur in some licks and not in all, could be due to different physiological conditions or to other unknown factors present in ticks. Previous studies showed that B. burgdorferi spirochactes multiply in the midgut of/. ricinus females. Only in some infected ticks, a small number of spirochactes were found to escape from the midgut and to penetrate the midgut wall leading to systemic infections (Gern et al. 1990, Monin et cd. 1989). This phenomenon could be somehow linked to the antigenic changes observed in tick reisolates which also took place only in some ticks. A selection process may occur in the midgut. The modifications observed in the antigenicity of OspA and OspB in our reisolates may play a role in this selection: only the 'adapted' spirochaetcs could be able to adhere to midgut epithelial cells and traverse the midgut wall. The adherence of B. burgdorferi to the tick midgut cell surfaces has already been described (Benach et al. 1987). The OspA and OspB have been suggested to play a role in the adherence of spirochaetcs io endothelial cells (Benach et al. 1988, Comstock & Thomas 1989, Thomas & Comstock 1989) and adherence to cells is considered an important step for bacterial invasion and in pathogenesis (Finlay & Falkow 1989). Therefore, the antigenic shift of OspA and OspB and may be the 22 kDa protein, revealed after passage into tick midgut, might play an important role for the pathogenicity and infectivity of B. burgdorferi in the vertebrate host. In experimentally infected mice, the first antibodies to be detected arc those directed against OspA and OspB (Benach et al. 1988, Schwan et al. 1989). Antibodies specific for Changes in the protein profiles of B. burgdorferi 425 the outer surface protein A (OspA) can protect immunodeficient Seid mice (Schaible et al. 1990, Simon et al. 1991) as well as immunocompetent CiH mice (Fikrig et al. 1990) against clinical manifestations. However, in patients, antibodies against OspA and OspB are not expressed until the late stages of the disease (Barbour et al. 1983b, Craft et al. 1986, Dattwyler et al. 1988). The differences observed in the immune response in mice and patients could be related to the mode of inoculation of the spirochaete as it has already been suggested in a previous study in dogs (Greene et al. 1988). The authors found that experimentally infected dogs expressed antibodies directed to OspA, whereas naturally infected dogs did not. Our data could support the suggestion that the different immune responses in experimentally and naturally infected hosts like humans are influenced by the way of inoculation. Our observations might explain the protein patterns of European isolates showing such an heterogeneity with respect to their major surface proteins A and B (Figure 5). The fact that the phenotype of a B. burgdorferi population changed during residence in the tick suggests that ticks not only transmit the micro-organism but provide also the environment which leads to the variation of one or the other epitope of B. burgdorferi antigens. The mechanisms responsible for such changes in the arthropod environment remain unknown. Experiments concerning pathogenicity and the molecular genetics of these isolates and their modified reisolates are in process. In vertebrate host, antigenic variation is also suggested but not established at this point. Additional investigations of its occurrence in animals and human infections, as well as in ticks, should be done. OspA (NE58) OspA and 2ZkOo (NE2. NE4. NE83 NE202) OspA end OspB (NE203 B2I OspA, OspB ond 22 kOo (831. NE56I Major proteins of or ig inol isola tes. (before passage into ticks) Passage in Major proteins of their tick midgut reisolates (after passage into ticks) Figure 5. Schema of the modifications of the major proteins of B. burgdorferi after passage into I. ricinus. It is shown thai changes from a phenotype (major proteins) to another one occur in the tick midgut (indicated by arrows) and the reisolatcd phcnotypes (left) are similar to other original phenolypes of Ii. burgdorferi strains (right). 426 C. M. Hu étal. Acknowledgements This work is part of the PhD thesis of one of the authors (C.M.H.) and it was supported by the Swiss National Research Foundation. We thank Bettina Wilske for having tested some strains with monoclonal antibodies. We also thank Olivier Rais for technical assistance, Allan Barbour for providing monoclonal antibodies and J. Schmidli for providing the strain B2. We gratefully acknowledge Ulrich Schaible and Reinhard Wallich for critical review of the manuscript. References Barbour A.G. & Schrumpi- M.E. (1986) Polymorphisms of major surface protein of Borrelia burgdorferi. Zentralblatt für Bakteriologie und Hygiene (A) 263, 83 Barbour A.C. Heiland R.A. & Tessin H.R. (1985) Heterogeneity of major proteins in Lyme disease borrcliac: A molecular analysis of American and European isolates. Journal of Infectious Diseases 152, 478 Barhour A.C. Tessier S.L. & Hayes S.F. (1984) Variation in a major surface protein of Lyme disease spirochete. Injections and Immunity 45, 94 Barbour A.C. Tessier S.L. & Todd W.J. (1983a) Lyme disease spirochetes and ixodes tick spirochetes share a common surface antigenic determinant defined by a monoclonal antibody. Infection and Immunity 41, 795 Barbour A.C., Burgdoreer W., Grunwaldt E. & Sterri- A.C. (1983b) Antibodies of patients with Lyme disease to components of the Ixodes dammini spirochete. Journal of Clinical Investigations 82, 504 Barbour AC, Burgdori-er W., Hayes S.F., Péter O. & Aeschlimann A. (1983c) Isolation ofa cultivable spirochete from Ixodes ricinus tick of Switzerland. Current Microbiology 8, 123 Bi-nach J.L., Coleman J.L., Garcia-Monco J.C.C & Deponie P.C. (1988) Biological activity of Borrelia burgdorferi antigens. Annals oj New York Academy of Sciences 539, 115 BenACii J. L, Coleman J. L., Skinner R. A. & Bosler E.M. (1987) Adult Ixodes dammini on rabbits: a hypothesis for the development and transmission of Borrelia burgdorferi. Journal of Infectious Diseases 155, 1300 Bundoc V.C & Barbour AC. (1989) Clonal polymorphisms of outer membrane protein OspA of Borrelia burgdorferi. Injection and Immunity 57, 2733 Crai-t .I.E.. Fischer D.K.. Shimamoto CT. & Stei-RE A.C. ( 1986) Antigens of Borrelia burgdorferi recognized during Lyme disease: appcarcncc of a new immunoglobulin M response and expansion of the immunoglobulin G response late in the illness. Journal of Clinical Investigations 78, 934 Q)MSTOCK L.E. & Thomas D.D. (1989) Penetration of endothelial cell monolayers by Borrelia burgdorferi. Infection and Immunity 57, 1626 Daitwyler R. .1.. Voi.kman D.J.. Halperin JJ., Luit B..1.. Thomas J. & Golighti.y MC. (1988) Specific immune response in Lyme borrcliosis: characterization of Tcell and B cell response to Borrelia burgdorferi. Annals of New York Academy of Sciences 539, 93 KiKRiCi E., Barthold S.W., Kantor F.S. & Fla veli. R.A. (1990) Protection of mice against the Lyme disease agents by immunizing with recombinant OspA. Science 250, 533 Finlay B.B. & Falkow F.S. (1989) Common themes in microbial pathogenicity. Microbiology Review S3, 210 Gern L., Zhu 7.. & Aeschi.imann A. (1990) Development of Borrelia burgdorferi in Ixodes ricinus females during feeding. Annales de Parasitologic Humaine et Comparée 65, 89 Changes in the protein profiles of B. burgdorferi 427 Graf .I.Fr. (1978) Copulation, nutrition et ponte chez Ixodes ricinus L. (Ixodoidca: lxodidac)-lère partie. Bulletin de la Société Entomologique Suisse 51, 89 Greene R.T.. Walker R.L.. Nicholson W. et al. (1988) Immunoblot analysis of immunoglobulin G response to the Lyme disease agent (Borrelia burgdorferi) in experimentally and naturally exposed dogs. Journal of Clinical Microbiology 26, 648 Krami;r MD.. ScHAiBLi: U.E.. Wallich R.. Moter S.E., Petzoldt D. & Simon M.M. (1990) Characterization of Borrelia burgdorferi associated antigens by monoclonal antibodies. linmunobiologv 181, 357 Monin R.. Gi-rn L. & Ar-SCHLiMANN A. (1989) A study of the different modes of transmission of Borrelia burgdorferi by Ixodes ricinus. Zentralblatt für Bakteriologie Suppl. 18, 14 SrHAiItU-: U.E., Kramer M.D., Eichmann K., Modoli-ll M., Museteanu C. & Simon M. (1990) Monoclonal antibodies specific for the outer surface protein A (OspA) of Borrelia burgdorferi prevent Lyme borrcliosis in severe combined immunodeficiency (seid) mice. Proceedings of the National Academy of Sciences of the United States of America 87, 3768 Sciimiih.i J.. HiJNZiKER T., MOESU P. & ScHAAD W.B. (1988) Cultivation of Borrelia burgdorferi from joint fluid three months after treatment of facial palsy due to Lyme borrcliosis. Journal of Infectious Diseases 158, 905 Schwan T. & Burgdorier W. (1987) Antigenic changes of Borrelia burgdorferi as a result of in riiro cultivation. Journal of infectious Diseases 156, 852 Schwan T.G.. Burgdoreer W. &GaronC.F. (1988) Changes in infectivity and plasmid profile of the L, y m c disease spirochete Borrelia burgdorferi, as a result of in ritro cultivation. Infection and Immunity 56, 1831 Schwan T.G.. Kime K.K.. Schrumpf M.E.. Coi: .I.E. & Simpson W.J. ( 1989) Antibody response in white-footed mice (Peromyscus lcucopus) experimentally infected with the Lyme disease spirochete (Borrelia burgdorferi). Infection and Immunity 57, 3445 Simon M.M.. Schaihu; U.E.. Wallich R. & Kramer M.D. (1991) A mouse model for Borrelia burgdorferi infection: approach to a vaccine against Lyme disease. Immunology Today 12, I 1 Thomas D.D. & Comstock L.E. (1989) Interaction of Lyme disease spirochetes with cultured eucaryotic cells. Injection and Immunity 57, 1324 Wii.ske Ii.. Preac-Mursic V.. Schierz G. & Busch K.V. (1986) Immunochemical and immunological analysis of European Borrelia burgdorferi strains. Zemralblatt für Bakteriolo- gie und Hygiene 263, 92 WiLSKi B.. Preac-Mursic V.. Schierz G.. Kuhkeck R., Barbour AG. & Kramer M.D. (1988) Antigenic variability of Borrelia burgdorferi. Annals of New York Academy ofSciences 593, 123 Proceedings of 1st International Conference on Tick-Borne Pathogens at the Host-Vector Interface: Agenda for Research. Eds U. G. Munderloh and T. J. Kurtti. Saint Paul, Minnesota, 1992:121-126 Antigenic variation in Borrelia burgdorferi after passage through Ixodes ricinus and Ixodes hexagonus ticks L. Gern, C. M. Hu, L. A. Toutoungi and M. Kramer* Institut de Zoologie, Université de Neuchâtel, Switzerland and * Institut für Immunologie, Universität Heidelberg, FRG Introduction Borrelia burgdorferi, the causative agent of Lyme borreliosis is transmitted to humans and animals by infected ticks. In Europe, Ixodes ricinus is the main vector of this spirochete (Burgdorfer et al, 1983), and recently the hedgehog tick, I. hexagonus has proved to transmit B. burgdorferi (Gern et ai, 1991). The spirochetes spend part of their life cycle within the arthropod. In laboratory research, this aspect has often been underestimated. Most of the studies on B. burgdorferi have been performed with bacteria passaged in vitro. Under these conditions, changes in the antigen profile as well as loss of infectivity for animals have been reported (Schwan et al., 1987, 1988). In this study, various in vitro grown isolates of B. burgdorferi were reintroduced into I. ricinus and I. hexagonus and afterwards recultivated to determine whether the passage through ticks influences the protein and antigen profile of the spirochetes. 121 Materials and Methods Strains B. burgdorferi strains NE2, NE4, NE56, NE58, NE83, NE202, NE203, NE196 were isolated from I. ricinus midguts in BSKII medium (Barbour, 1984). Ticks were collected from the Neuchâtel area of Switzerland. The isolate B2 derived from synovial fluid of a patient (Schmidli et al, 1988) and the North American strain B31 were also examined. To isolate clonal populations of B. burgdorferi, strains NE2 and NE83 were cloned by limiting dilution (Bundoc and Barbour, 1989). Re introduction of B. burgdorferi into I. ricinus and I. hexagonus and reisolation The isolates NE2, NE4, NE56, NE58, NE83, NE202, NE203, B2, B31 and the cloned strain NE83 (cNE83) (108 cells I ml) were reintroduced into uninfected I. ricinus females, by the capillary method (Gern et al., 1990). One week thereafter, spirochetes were reisolated from the tick midgut. All isolates and reisolates were submitted to SDS-PAGE and Immuno Blot as described elsewhere (Hu et al, 1992). The monoclonal antibodies (MoAb) H9724 and H5332 specific for the flagellin and OspA, respectively, of B. burgdorferi strain B31 were obtained from Alan Barbour (Barbour et al., 1983). Other monoclonal antibodies (LA- 31, LA-4, LA-28; LA-25, LA-32, LA-27; LA-7) recognizing epitopes of OspA, OspB and 20 kDa antigens (Kramer et al, 1990) were used for Immuno Blot as well as polyclonal monospecific antibodies against a 22 kDa protein of strain NE4 (PoAb) (anti-22 kDa/NE4; Ha et al, 1992). The isolate NE 196 (B31 type; 107cells I ml) was introduced into 10 flat I. hexagonus females using the capillary method and was reisolated one week later. The cloned strain cNE83 (107cells I ml) was used to infect 30 engorged I. hexagonus nymphs by the same method and was reisolated after tick moulting. Results I. ricinus, I. hexagonus and uncloned strains Some reisolates showed altered protein profiles after passage in I. ricinus. Striking modifications were observed in the molecular weight ranges of OspA and OspB. We grouped the isolates according to their major surface proteins before and after passage into ticks. In group I, four isolates (NE2, NE4, NE83 and NE202) presented two major proteins with molecular masses of 33 kDa and 22 kDa, or 31 kDa and 22 kDa. All modified reisolates from the tick midgut were found to have lost the 22 kDa band. Some of them have shifted 122 and altered the 33 kDa band: In three reisolates a 34 kDa band and in two a 31 kDa band appeared. The strains NE4 and NE2 changed into a B31-like phenotype with respect to OspA and OspB. Loss of the 22 kDa protein of the group I reisolates resulted in lack of staining with PoAb specific for the 22 kDa protein of NE4. Moreover, the reaction of the modified reisolates from NE4 with moAb H5332 changed from negative to positive. The initial isolates of group II (B2 and NE203) presented proteins with molecular masses of 31 and 34 kDa and reisolates showed an additional protein band of 22 kDa. Only one strain, NE58, was included in group III. The main protein of this isolate is a 33 kDa protein, perhaps OspA, which remained unchanged after tick passage. In contrast, a new protein band of 22 kDa appeared in the reisolates. The additional new protein of 22 kDa of the modified reisolates from groups II and III reacted with PoAb 22 kDa/NE4. In addition, OspA of strain NE58 did not react any more with moAb H5332 after reisolation from the tick. All reisolates of group IV, NE56 and B31, did not show any changes. After passage through I. hexagonus, reisolates from NE 196 (type B31 ) did not show any changes. I. ricinus, I. hexagonus and the cloned strain cNE83 To exclude the possibility that a simple cloning effect was responsible for the reisolation of modified spirochetes, we cloned the strains NE83 and NE2 (both strains had lost the 22 kDa protein after passage through I. ricinus). When compared to the uncloned strains, no differences were observed after cloning: the 22 kDa did not disappear suggesting that loss of the 22 kDa protein (as observed in tick midguts) is merely not a phenomenon of cloning. The cloned strain cNE83 was reintroduced into I. ricinus and I. hexagonus. Five reisolates from I. ricinus presented only a weak band of OspA with 33 kDa and the 22 kDa protein was unchanged (Fig.l). The 33 kDa was not detectable in 3/6 reisolates from I. hexagonus (Fig.l). Immuno Blot results of cNE83 and its reisolates from I. ricinus and I. hexagonus are summarized in Table 1. 123 Fig.l. Protein profiles of the B. burgdorferi cloned strain cNE83 before passage through ticks and its reisolates from I. ricinus (left) and I. hexagonus (right). R: reisolates. Table 1 Immuno-reactions of cNE83 before and after passage through I. ricinus and I. hexagonus Specific cNE83 cNE83 (Ir) cNE83 (Ih) antibodies_____________5 reisolates_________6 reisolates anti-OspA H5332 - ND -/-/-/-/-/- LA-2 - ND -/-/-/-/-/- LA-4 + ND -l-l-l-l-l- LA-31 + -/+/-/+/- +1-1-1-1+1+ anti-OspB LA-25 LA-27 LA-32 + .1-1-1-1-ND ND -N-I-I-I--N-I-I-I--l-l-l-l-l- anti-OspC 22kDa/NE4 + +/+/+/+/+ +1+1+1+1+1+ anti-20kDa LA-7 _ ND -l-l-l-l-l- ami-flagellin H9724 + +/+/+/+/+ +1+1+1+1+1+ ND: not done Ir: Ixodes ricinus Ih: Ixodes hexagonus Discussion In this study, 9 Swiss strains and, the North American strain B31 were artificially introduced into I. ricinus or I. hexagonus and reisolated from their midguts. In most cases, modifications of the protein and antigen profiles were observed. OspA, OspB and a protein with a molecular mass of 22 kDa were mainly changed. The 22 kDa band was either "lost" or "acquired" in the reisolates and these modifications were accompanied by shifts of the 33 kDa protein: i.e. in some reisolates a 31 and a 34 kDa protein was present. Changes were also found by Immuno Blot analysis which revealed loss or appearance of specific epitopes. Reisolates obtained from an uncloned (NE83) and cloned strain (cNE83) presented different pattern after tick passage: reisolates from NE83 lost the 22 kDa protein and presented the 33 kDa whereas the contrary was observed with the reisolates of cNE83. Selection of a B. burgdorferi variant that is more prone to express the 22 kDa protein than the 33 kDa protein, possibly induced by the cloning process, could account for this phenomenon. A comparable observation was described by Bundoc and Barbour (1989): an increase in expression of the 21 or 18.5 protein was associated with a partial or complete loss of OspB in different variants. Moreover Immuno Blot analysis of the reisolates of cNE83 from both tick species revealed a loss of epitopes detected by OspA- and OspB-specific antibodies (Table 1). Previous studies showed that B. burgdorferi multiplies in the midgut of I. ricinus females. Only in some infected ticks, few spirochetes escape from the midgut and induce a systemic infection (Gern et al, 1990). It is tempting to speculate that this is somehow linked to the antigenic changes observed in tick reisolates which also appeared only in some ticks. Moreover, recent studies revealed that spirochetemia in mice infected via tick bites is higher than in experimentally infected animals suggesting that the passage of the spirochetes in the tick environment enhances the infectiviry of B. burgdorferi (Gern et al. submitted). In vertebrate hosts, antigenic variation has also been suggested (Schwan et al. 1991). Additional investigations in animals, ticks and humans should be addressed to this matter. Aknowledgements We thank Alan Barbour for providing monoclonal antibodies and Jurg Schmidli for providing the B. burgdorferi strain B2. This work is part of the PhD thesis of one of the authors (C.M.H.) and it was supported by the Swiss National Research Foundation. 125 References Barbour A.G. 1984. Yale J. Biol. Med. 57:521-525. Barbour, A.G., S.L. Tessier and WJ. Todd. 1983. Inf. Imm. 41: 795-804. Bundoc V.G. and A.G. Barbour. 1989. Inf. Immun. 57: 2733-2741. Burgdorfer W., A.G. Barbour, S.F. Hayes, O. Péter and A. Aeschlimann. 1983. Acta Trop. 40: 79-83 Gern L., Z. Zhu and A. Aeschlimann. 1990. Ann. Paras. Hum. Comp. 65: 89-93. Gern L., L.N. Toutoungi, CM. Hu and A. Aeschlimann. 1991. J. Med.Vet. Entom. 5:431-435. Gern L., U.E. Schaible and M.M.Simon. Submitted. Hu CM., L. Gern and A. Aeschlimann. 1992. parasite Immunology. 14: 415- 427 Kramer M.D., U. Schaible, R. Wallich, S.E. Moter, D. Petzoldt and M.M. Simon. 1990. Immunbiol. 181:357-366. Schmidli J., T. Hunziker, P. Moesli and W.B. Schaad. 1988. J. Inf. Dis. 158: 905-906. Schwan T.G. and W. Burgdorfer. 1987. J. Infect. Dis. 156: 852-853. Schwan T.G., W. Burgdorferand Garon. 1988. Inf. Imm. 56:1831-1836. Schwan T.G., R.H. Karstens, M.E. Schrumpfand W.J. Simpson. 1991. Can. J. Microbiol. 37: 450-454. 126 Proceedings of the VI International Conference on Lyme Borreliosis. Eds. R. Cevenini, V. Sambri and M. La Placa. Bologna, Italy, June 19-22, 1994. 19-22 19 PROTEINIC, ANTIGENIC AND PATHOGENIC VARIATIONS OF A CLONAL BORRELIA BURGDORFERI ISOLATE FROM IXODES RICINUS HEMOLYMPH CM. Hu*, MD. Kramer**, MM. Simon °, L. Gern* ?Institut de Zoologie, Université de Neuchâtel, Switzerland ??Immunologie, Ruprecht-Karls-Universität, Heidelberg, Germany 'Max-Planck Institut für Immunbiologie, Freiburg, Germany In this study, we obtained a B. burgdorferi isolate from tick hemolymph which was highly pathogenic for SCID mice and expressed two main proteins in the molecular weight range -20 kDa, i.e. 22 kDa and 23 kDa, detected by a specific immune serum. Continuous cultivation in vitro resulted in a successive loss of the 22 kDa and 23 kDa molecules, loss of reactivity with the immune serum and a loss of pathogenicity in the SCID mice. After cloning, this non pathogenic strain was reintroduced into ticks by capillary feeding. Most of the reisolates from various tick organs again reacted with the specific immune serum. Two out of four isolates which re- expressed the 22/23 kDa proteins induced clinical arthritis in SCID mice. These results suggest that tick factors influence the protein profiles, antigenicity and pathogenicity of B. burgdorferi spirochetes. In Europe, Borrelia burgdorferi is transmitted by Ixodes ricinus ticks. In most host-seeking ticks, spirochetes are limited to 20 the midgut. During the blood meal, they penetrate the midgut wall and induce a systemic infection via hemolymph (Gem et al., 1990). Immunochemical analysis of the European isolates derived either from tick midgut or from different animal hosts showed that their outer surface proteins (OspA, OspB and OspC) are variable in their molecular weights and reactivity with different antibodies (Kramer et al., 1992). The great majority of tick isolates investigated so far were obtained from either tick midgut or tick homogenates. Previous studies have shown that the reintroduction of B . burgdorferi into ticks resulted in changes of protein profile and antigenicity (Hu et al., 1992). In an attempt to further study the influence of the tick's physiology on the phenotypic variability of B. burgdorferi, spirochetes were isolated from I. ricinus hemolymph. Here we describe one isolate, and its clonal derivative and report on its variations in protein profile, antigenicity and pathogenicity. Ixodes ricinus adults were collected by flagging vegetation in a forest near Neuchâtel (Bois de l'Hôpital, Switzerland). The tick hemolymph was collected using capillaries after cutting the coxa I from the first leg and was inoculated in a tube containing 4 ml of BSK II medium. The culture tubes were examined after 10 days by dark-field microscopy. One isolate was obtained and was subcultured in BSK II medium for 15 passages. In addition, the isolate was cloned at passage 3 and reintroduced into uninfected J. ricinus adults using the capillary feeding technique. One week later, the spirochetes were reisolated from various tick organs including midgut, salivary glands and genital tissues. The protein profiles and antigenicity of the isolate, its clonal derivative as well as the reisolates were examined by SDS-PAGE and by immunoblotting using 21 a panel of monoclonal antibodies and an immune serum (IS) specific for 22/23 kDa proteins. In addition, the pathogenicity of all spirochetal preparations were examined in adult SCID mice by monitoring for clinical signs of arthritis in the tibiotarsal joints for 70 days post inoculation. Results showed that the original isolate expressed four major protein bands corresponding to 33 kDa, 32 kDa and two proteins in the 20 kDa range, 22/23 kDa. Such a protein pattern has never been observed before in our laboratory among -200 tick isolates from the midgut which is the primary organ site for spirochetes in unfed ticks. This suggests that the particular phenotype of the present isolate is related to the particular hemolymph environment. During serial subcultivation in vitro, the original isolate lost the 22 kDa protein after 5th passage and the 23 kDa protein after 15th passage as well as its reactivity with the IS. After cloning of the original isolate at passage 3, the cloned strain lacked the 22 kDa protein and presented a phenotype similar to that of the 5th in vitro passage of the original isolate. When the cloned B. burgdorferi strain was reintroduced into ticks and reisolates from various tick organs were analysed, variations in the protein profile were observed. Two of the reisolates reexpressed the 22 kDa protein. These results indicate that the expression of the 22/23 kDa proteins is influenced by tick factors. Moreover, we demonstrate that proteins and epitopes of B. burgdorferi which are lost during in vitro cultivation might be re-acquired after re-exposure to its former environment. The original isolate was highly pathogenic for SCID mice (Schaible et al., 1993) but lost this activity after continuous in vitro passages. Moreover, the clonal strain was also unable to induce clinical arthritis in SCID mice. Whereas, after passage through ticks, two out of four 22 reisolates again induced disease in SCID mice. Together the results suggest that tick factors may influence both the phenotype and the pathogenicity of B. burgdorferi. This work is from a part of PhD thesis of Hu, CM. and was partially supported by the Swiss National Science Foundation (32-29964.90), by the Bundesministerium für Forschung und Technologie (01 KI 9104) and by the Commission of the European Communities (P92321). We thank S. di Nuncio, O. Rais for their help in collecting ticks and for their technical assistance, and Alan Barbour for providing monoclonal antibody H5332. Gern L., Zhu Z. and Aeschlimann A. 19 9 0 Development of Borrelia burgdorferi in Ixodes ricinus females during blood feeding. Annales de Parasitologie Humaine et Comparée 65, 8 Hu C. M., Gern L., Aeschlimann A. 1992 Changes in the protein profile and antigenicity of different Borrelia burgdorferi strains after reintroduction to Ixodes ricinus. Parasite Immunology 14, 415 Kramer M. D.# Schaible Ü. E., Wallich R., Moter S. E., Petzoldt D. and Simon M. M. 1990 Characterization of Borrelia burgdorferi associated antigens by monoclonal antibodies. Immunobiology 181, 357 Schaible U.E., Wallich R., Kramer M.D., Gern L., Anderson J. F., Museteanu C. and Simon M.M. 1993 Immune sera to individual Borrelia burgdorferi isolates or recombinant OspA thereof protect SCID mice against infection with homologous strains but only partially or not at all against those of different OspA/OspB genotype. Vaccine 11, 1049 Originalia______________________________________________________________ C. M. Hu, M. Simon, M. D. Kramer, L. Gern Tick Factors and In Vitro Cultivation Influence the Protein Profile, Antigenicity and Pathogenicity of a Cloned Borrelia gannii Isolate from Ixodes ricinus Hemolymph Summary: A Borrelia garinii isolate (NEIlH) was obtained from the hemolymph of infed Ixodes ricinus. NEIlH expressed four major proteins of 33 kDa, 32 kDa, 23 kDa and 22 kDa. During in vitro culture, NEIlH successively lost the expression of the 22 kDa and 23 kDa proteins and the NEIlH variant (NEllHpl5) was not recognized by an immune se- rum specific for the OspC protein (anti-OspC IS). However, when reintroduced into tick midguts, NEllHpl5 spirochetes present in the midgut again reacted with anti-OspC IS. A clone derived from the wild type line, cNEHH, lacked the 22 kDa but not the 23 kDa pro- tein. The 23 kDa protein of cNEHH was recognized by anti-OspC IS. In addition, the two descendant lines (NEllHpl5 and cNEllH) lost their capacity to induce clinical arthritis in SCID mice. When cNEHH was reintroduced into ticks and reisolated from various tick or- gans, most reisolates presented the same reaction with anti-OspC IS as cNEHH. Interest- ingly, two reisolates obtained from the tick midgut reexpressed large amounts of the 22 kDa protein which was recognized by anti-OspC IS and these two reisolates induced clinical ar- thritis in SCID mice. The results confirm that proteins of 22/23 kDa are differentially ex- pressed during in vitro subcultures and in ticks, and show that proteins which are not de- tectable after in vitro culture may be reexpressed after reexposure of B. burgdorferi to its former environment in the tick. The data suggest that the pathogenicity of B. burgdorferi for mice might be influenced by environmental factors via differential expression of 22/23 kDa proteins. Introduction The Lyme disease spirochete, Borrelia burgdorferi [1] is transmitted in Europe by Ixodes ricinus [2]. In most host- seeking infected ticks, spirochetes are localized in the mid- gut. The salivary route of B. burgdorferi transmission is now generally accepted for I. ricinus [3] and for the Amer- ican tick vector I. scapularis [4,5]. This implicates that spi- rochetes penetrate the midgut wall and via hemolymph in- fect other tick organs during the blood meal. After molt- ing, some ticks may have a systemic infection [6,7]. To our knowledge, only a few isolates of B. burgdorferi have been obtained from tissues other than tick midgut and most tick isolates previously described have been derived from tick midgut or tick homogenates [6, 8, 9]. Phenotypic and genotypic analysis of European B. burg- dorferi isolates from I. ricinus ticks showed that there are at least three species: B. burgdorferi sensu stricto (ss), B. garinii and B. afzelii [8,10, H]. B. burgdorferi ss express- es outer surface protein (Osp) A of 31 kDa and OspB of 34 kDa, B. garinii has an OspA of 32-33 kDa and B. afze- lii expresses an OspA of 32 kDa and an OspB of 35 kDa. In addition to OspA and OspB, many European B. burg- dorferi sensu lato (si) isolates express a major protein with a molecular weight of about 22 kDa described as OspC [8, 12]. Most American isolates belong to B. burgdorferi ss and lack the expression of OspC protein [12,13]. Previous studies have shown that cultivation of tick-de- rived B. burgdorferi si isolates may result in differential expression of cell surface proteins and/or antigen variation and that the altered phenotype is associated with a loss of virulence in animal hosts [14,15]. In addition, it was shown that both expression of OspA, OspB and/or OspC as well as immunogenicity of B. burgdorferi si isolates were influ- enced by passage through ticks [16,17]. These results indi- cate an environmental influence on the phenotype and/or infection potential of B. burgdorferi. In the present study, we employed a B. garinii isolate (NEIlH) obtained from the hemolymph of a systematical- ly infected I. ricinus to examine the protein profile, antige- nicity and pathogenicity of its uncloned and cloned de- scendants after in vitro passages and reintroduction into ticks. The data demonstrate that spirochetes express pro- teins differently under in vivo from in vitro conditions and suggests that environmental factors are critical in the con- trol of their pathogenic potential. Received: 25 January 1996/Revision accepted: 27 February 1996 Dr. med. Chang Min Hu, Dr. rer. Lise Gern, Institut de Zoologie, Uni- versité de Neuchâtel, Neuchâtel, Switzerland; Dr. rer. PD. M. Simon, Max-Planck-Institut für Immunbiologie, D-79108 Freiburg; Dr. rer. med. PD. M. D. Kramer, Institut für Immunologie, Ruprecht-Karls-Uni- versität, D-69120 Heidelberg, Germany. Correspondence to: Dr. L. Gern, Institut de Zoologie, Emile-Argand 11, CH-2000 Neuchâtel, Switzerland. Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 251/42 C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity Mw (kDa) 97.4 66.2 NE11H o m T- co m T- i- Q. Q. Q. Q. Q. 45.0 31.0 21.0 14.4 a) Mw (kDa) p1 p3 p5 p1Q p15 31.0 21.0 b) Figure 1. a) SDS-PAGE stained with Coomassie blue of NE11H and its in vitro passages (p1, p3, p5, p10 and p15); b) reaction of NE11H h vitro passages with IS anti-OspC. The protein mo- lecular weight standards are shown on the left. Materials and Methods Ticks and B. burgdorferi isolates: Twenty /. ricinus adults were collected by flagging vegetation in a forest near Neuchâtel (Bois de l'Hôpital, Switzerland) in June 1990. Ticks were washed in 70% ethanol and rinsed in sterilized PBS pH 7.4. Tick nemo- lymph was collected using capillaries after cutting the first leg and was immediately inoculated in tubes containing 4 ml of BSK II medium [18] and maintained at 340C. After 10 days of incuba- tion the culture tubes were examined by dark-field microscopy over a period of 6 weeks. NEIlH was cultivated in BSK II until the fifteenth serial in vitro passage (p) (NEl 1 HpI-NEl lHpl5). The third passage (p3) containing 107 cells/ml (Helber counting cell chamber) was cloned by limiting dilution [16,19]. Serial ten- fold dilutions from 10"3 to 10~9 were carried out in BSK II medi- um and three tubes containing 5 ml BSK II were used for each dilution. After 3 weeks of cultivation, none of the tubes at the 10"9 dilution contained spirochetes and only one of three tubes was positive at the 10"8 dilution for B. burgdorferi This cloned strain, termed cNEl 1H, was reintroduced into 20 un- infected I. ricinus adults from our laboratory colony, using the capillary feeding technique as described previously [16,20]. Dur- ing capillary feeding all ticks were maintained at 34°C for 2 hours. Infected ticks were maintained at room temperature and dissected 10 days later. The midgut (m), salivary glands (s), gen- ital tissues (g) as well as hemolymph (h), removed as described above, were incubated individually in tubes containing 4 ml of BSK II medium for regrowth of spirochetes. Each reisolate was designated with a R (reisolate) followed by the tick number and the organ of origin (m, s, g or h). In addition, the fifteenth in vitro passage (p) of NEIlH (NEIlHpI5) was reintroduced into five uninfected /. ricinus adults. Ten days later, these ticks were dissected and each lick midgut was examined for the presence of spirochetes by direct immunofluorescence analysis (DIFA). Moreover, the expression of OspC on spirochetes in the tick midgut was tested by indirect immunofluorescence analysis (IDIFA). immunofluorescence analysis: The midguts of ticks infected with NEllHpl5 were smeared on two glass microscope slides, dried and fixed with acetone for 10 min. One slide was used to detect the presence of B. burgdorferi si spirochetes by direct immuno- fluorescence using a polyclonal antibody serum against B. burg- dorferi si conjugated with fluorescein isothiocyanate [21]. The second slide was used to detect the expression of OspC using a rabbit anti-OspC antiserum [22] kindly obtained from T. Schwan (Rocky Mountain Laboratories. Montana, USA) and visualized by a goat anti-rabbit fluorescein isothiocyanate conjugate. Each incubation lasted 1 h at 34°C. The slides were then examined at x 400 magnification by fluorescence microscopy. SDS-PAGE and Western blotting analysis: Serial in vitro subcul- tures of NEl 1H (pl-pl5), cNEl IH and its reisolates from differ- ent tick organs were examined by SDS-Polyacrylamide Gel Elec- trophoresis (SDS-PAGE). Whole cells underwent electrophore- sis on a 12.5% acrylamide gel with 107 cells/lane and stained with Coomassie brilliant blue R-250 [16]. Western blotting was car- ried out as described previously [16] using the following mono- clonal antibodies (MoAbs): H5332. LA-31, LA-4. LA-28 reac- tive with OspA [23, 24]; LA-25, LA-27, LA-32 reactive with OspB [24] and LA-7 reactive with 20 kDa protein [24]. More- over, one rabbit polyclonal immune serum (IS): anti-OspC obtained from T. Schwan [22] was used. Pathogenicity: Adult SCID mice were bred under pathogen-free conditions at the Max-Planck Institut für Immunbiologie in Frei- 42/252 Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München. München 19% C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity burg. Germany. Mice were inoculated subcutaneously into the base of the tail (3 mice per isolate) with IxIO8 viable spirochetes of NEllHp3. NEItHpl5, its clonal derivative (cNEl IH) and of the reisolates of cNEllH from tick organs: R5g, R8m, RlOm. RlIm. Mice were inspected under blinded conditions every 3 days over 70 days after inoculation for clinical signs of arthritis in the tibiotarsal joints as described in detail [25]. Results SDS-PAGE and lmmunoblot analysis of in Vitro Passage of NEUH and its Pathogenicity In the molecular weight range below 35 kDa, NEIlHpI (first in vitro passage of NEIlH) expressed four major protein bands corresponding to 33 kDa. 32 kDa, 23 kDa and 22 kDa (Figure 1 ). After the fifth in vitro passage (p5). the 22 kDa protein was no more detectable, followed by the subsequent decline of expression of the 23 kDa band during the further 10 passages (shown after pi5, Figure 1a). All isolates from the in vitro passages of NEl IH reacted with mAbs H5332 and LA-31 directed against OspA. None of them reacted with mAbs LA-4, LA-28 (specific for OspA), LA-25, LA-27, LA-32 (specific for OspB) and LA-7 (specific for 20 kDa protein) generated against B. Table 1 : Pathogenicity of NE 11 Hp3 and its variants in SCID mice. Arthritis in mice* Isolates 8 12 15 19 21 26 29 33 36 50 69 days** NEllHp3*** NEllHpl5 cNEllH R5g R8m RlOm RlIm __________ 1/3 - - - 1/3 2/3 3/3 3/3 3/3 3/3 3/3 3/3 •Each isolate was inoculated into three mice; **Mice were inspected about every 3 days after inoculation: ***Shown previously {Schaible et il IW3):-: No arthritis. burgdorferi ss [15]. During serial in vitro subcultivations, spirochetes of strain NEl IH lost the ability to bind anti- bodies from anti-OspC (Figure 1b): the 22 kDa and 23 kDa proteins of NEIlH (pi) and p3 spirochetes reacted with anti-OspC IS whereas only the 23 kDa protein of p5 and plO reacted. Spirochetes of NEl 1H (pi 5) did not react at all with anti-OspC IS. In contrast to the potential of the low passage strain NEllHp3 to induce arthritis in SCID mice [26]. NEllHpl5 (after 12 additional passages) lost this ability (Table 1). Reisolates from midgut Reisolates from genital tissue Reisolates from salivary gland MW (kDa) _____ 66.0 45.0 _____ _____ _____ EEE CM CO EEEEEeIIco m (p n o a cccrcrcrûctrcrircrtrcccc -" #¦_____ .-. ««4P - 00)0)0)0)0)00)0)0 t-wP3tm_s(oo)T- crcccccccccccccccca: ¦ V ¦ __-_¦ _¦ Figure 2. SDS-PAGE (Coomassie stain) of cNE11H reisolates from different organs of I. ricinus, a) reisolates from tick midgut (Rm); b) reisolates from genital tissue (Rg); c) reisolates from salivary glands (Rs). The protein molecular weight standards are shown on the left. Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München. München 19% 253/43 C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity Table 2: Immuno-reactions of cNE11H and its reisolates from ticks. cNEllH Reisolates from ticks R4m R4g R4s R5m R5g R5s R6m R6g R6s R8m R8g R8s RlOm RlOg RlIm Antibodies* H5332 + + + + + + + + + + + + + + + + LA-31 + + + + -- + + + + - + + + + + anti-OspC 23 23 - 23 23 23 23 23 23 23 - 23 23 22 23 22 *+: Positive reaction; -: Negative reaction; 22: Reaction with the 22 kDa protein; 23: Reaction with the 23 kDa protein; m: midgut, g: genital tissues, s: salivary glands. Analysis of the 22-23 kDa Protein Expression in the Tick Midgut As shown above, in v/Yro-cultured NEllHpl5 spirochetes lack 22 kDa and 23 kDa proteins as revealed by SDS- PAGE (Figure 1a) and they do not react with anti-OspC IS (Figure 1b). To see whether cultured NEllHpl5 spirochetes reexpress the 22 kDa protein when present in the tick midgut the in vitro variant NEllHpl5 was reintro- duced into ticks by capillary feeding. Analysis by DIFA re- vealed that 4/5 tick midguts were infected with spiro- chetes. The spirochetes in 3/4 tick midguts also reacted with the anti-OspC IS by IDIFA. However, the number of OspC-expressing spirochetes in tick midgut was less (one to two per spot of 4 mm diameter) than the number of spi- rochetes detected by antiserum against B. burgdorferi si (> 10 per spot). In vivo Passage ofcNEJIH and its Pathogenicity cNEllH, a clone derived from NEllHp3, was shown to still express, besides the 33 kDa and 32 kDa proteins, the 23 kDa protein but not the 22 kDa protein (Figure 2). As expected, only one band of 23 kDa was detectable in the Western blot with anti-OspC IS (Table 2). cNEllH was reintroduced into 20 ticks and spirochetes were reisolated from various organs 10 days later. Most ticks (10/12; tick number 1-10) had a systemic infection with spirochetes isolated from midgut and one or two ad- ditional organs (Figure 2). No reisolate was obtained from the hemolymph whereas 11 reisolates (R) were derived from the midgut (m), 10 from the genital tissues (g) and 8 from the salivary glands (s) of a total of 12 infected ticks (tick number 1-11 and 13) (Figure 2). The protein profiles of reisolates were not significantly different from that of cNEllH with respect to the 23 kDa protein (Figure 2). However, RmIO and RmIl, two midgut reisolates, re-ex- pressed a protein of 22 kDa (Figure 2a). Western blot analysis of the tick reisolates also showed a different reaction pattern particularly with the anti-OspC IS and MoAbs LA-31 (anti-OspA). All reisolates showed the same reactions of the 23 kDa protein with anti-OspC IS as cNEllH except reisolates R4g and R8m which did not react at all and RlOm and RlIm whose 22 kDa protein re- acted with anti-OspC IS (Table 2). On the other hand, R5m, R8m and R5g lost their ability to react with MoAbs LA-31. Experimental infection of SCID mice with cNEllH showed that cNEllH was unable to induce disease in these mice (Table 1 ). Four reisolates R5g, R8m, RlOm and RlIm, which were phenotypically different from cNEllH either by expression of a very low level of 23 kDa protein and by a lack of reaction with mAb LA-31 directed against OspA (R5g and R8m) (Table 2 and Figure 2) or by expres- sion of a very high amount of the 22 kDa protein (RlOm and RlIm) (Figure 2), were tested for their ability to in- duce arthritis in SCID mice. Whereas R5g and R8m were unable to induce clinical symptoms in SCID mice, RlOm and RlIm were shown to induce either moderate (RlOm, 1/3 SCID mice) or severe (RlIm, 3/3 SCID mice) clinical arthritis (Table 1). Discussion The main finding of this study is that a clonal isolate of B. garinii derived from tick hemolymph which had lost its pathogenic potential during in vitro culture regained this capacity upon reintroduction into ticks. The alteration of pathogenic potential is associated with a differential ex- pression of proteins in the range of 22/23 kDa. The B. burgdorferi isolate NEIlH was obtained from the hemolymph of a systematically infected unfed I. ricinus fe- male and belongs to OspA genotype II (B. garinii) based on Southern blot analysis [6]. This isolate expresses two protein bands in the molecular weight range of 22 kDa and 23 kDa which reacted with an immune serum directed against OspC (anti-OspC) [22]. This phenotype has been observed only twice so far among about 200 tick isolates obtained in our laboratory [6, 8,16]. These three isolates, which express two anti-OspC reactive proteins, were ob- tained from unfed ticks which were systematically infect- ed by B. garinii [6]. The relation between expression of the 22 kDa and 23 kDa proteins of B. garinii and the presence of a systemic infection in ticks remain to be studied. During serial in vitro subcultivation, NEIlH successively lost the expression of the 22 kDa and 23 kDa proteins af- ter the fifth and fifteenth passages respectively, together with the respective epitopes recognized by a monospecif- ic IS directed against OspC [22]. Phenotypic changes in spirochetal populations during their in vitro propagation have been described before by others [13,27] and are most probably induced by culture conditions. It is therefore 44/254 Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity possible that the in vitro conditions lead to the generation of genotypic/phenotypic variants and that this process re- sults in differential expression of 22 kDa and 23 kDa mole- cules [12, 19, 28]. In fact, this could be proven for the 22 kDa molecule of NEIlH by both in vitro passage and clon- ing procedures: p5 and cNEllH did not express the 22 kDa protein. The finding that cNEHH, after re-introduction into ticks, could be reisolated at a very high rate from organs of in- fected ticks other than midgut demonstrates the capacity of this variant to induce systemic infections in this vector. However, the possibility that this was due to the high con- centration of spirochetes introduced into the midgut by capillary or to a contamination between organs during tick dissection cannot be excluded. The re-expression of the 22 kDa protein in two reisolates of cNEllH from ticks and of the epitope (22 kDa protein) recognized by anti-OspC IS indicates an influence of the tick environment on the cell surface phenotype of spiro- chetes. Indeed, when NEllHpl5 was reintroduced into ticks, the 22 and 23 kDa proteins, which were not expressed in cultured NEllHpl5 were expressed in the tick midgut as detected by IDIFA using anti-OspC IS [22]. Variation of OspC expression was described to be associated with the number of in vitro passages [29], with passage into ticks [16] as well as with temperature variations [30]. Schwan et al. [30] observed that OspC was produced by B. burgdorferi ss spirochetes when grown in vitro at 32°C-37°C or in en- gorged tick midguts. When B. burgdorferi ss spirochetes were grown at 24°C very little OspC was expressed, as well as when spirochetes were present in unfed ticks. In fact, the contrary was observed in our study. When spirochetes were cultivated at 34°C they lost the expression of OspC, where- as they expressed OspC while present in the tick midgut. An explanation might be found in the Borrelia species: OspC of B. burgdorferi ss is not expressed in tick midgut [30] whereas OspC of B. garinii spirochetes is. In order to see whether OspC is expressed in unfed I. ricinus ticks, nymphal and adult ticks were collected from vegetation in the Neuchâtel area and examined by IDIFA using anti- OspC IS. When examined for the presence of spirochetes, only spirochetes found in adult ticks reacted with anti- OspC IS (unpublished data). Data from our laboratory show that B. garinii has been more frequently isolated from adult ticks (30/51) than from nymphs (5/21) (p=0.009). The significance of this relationship between expression of OspC and the presence of B. garinii in unfed field collect- ed adult I. ricinus remains to be explored. The present study is in line with our previous results, which demonstrated that the variations of OspC of B. ga- rinii were related to tick passage [16,17], and suggests that expression of the 22/23 kDa proteins and/or immunologi- cal determinants of B. garinii might be influenced by envi- ronmental conditions. In the tick, the 22 kDa protein of B. burgdorferi may play a role in penetrating different tick tissues and furthermore, in the transmission of spirochetes to the animal hosts. In fact, antibodies directed against the 22/23 kDa protein are detected early after infection in pa- tients [31, 32] and in animals infected via tick bites [33]. The mechanism responsible for changes in the protein ex- pression of B. burgdorferi si after passage into ticks is not known. There are at least two hypotheses to explain this phenomenon: variants may reflect the presence of geneti- cally independent clones as shown for the Osp locus [19, 28], on the other hand, they may be generated by regula- tory processes at the genetic level. A recent study [34] identified plasmid sequences homologous to the genes for two purine biosynthesis enzymes (guaA and guaB). These genes are adjacent to the outer surface protein OspC (22 kDa) gene of B. burgdorferi. Margolis et al. [34] suggest that this would allow the spirochetes to readily adapt to the different purine levels in vertebrates (low level) and ticks (high level). A more detailed genetic analysis of cNEllH and its reisolates is required to elucidate the mechanisms responsible for our observation. The changes observed in the expression of proteins and their antigenic determinants of NEIlH and cNEllH after serial in vitro subcultures and re-introduction in ticks seem to influence their pathogenicity in SCID mice. Similar to the strain B31 [14], the low passage strain NEIlH was highly pathogenic for SCID mice but lost its pathogenici- ty after prolonged in vitro subcultivation. Most notably, after re-exposure of cNEHH to tick environment two re- isolates from their midgut (RlOm and RHm) acquired path- ogenicity in SCID mice, though to various degrees. Both reisolates expressed the 22 kDa protein after passage through ticks whereas the two nonpathogenic re-isolates only expressed a very low level of this protein. These re- sults indicate that factors of the tick which allow the ex- pression of the 22 kDa protein also might influence the pathogenicity of B. burgdorferi. A relationship between infectivity and OspC expression was described previously [29]. Here we show that the 22 kDa protein correlates with the pathogenicity of B. garinii and that the expression of this protein is influenced by tick factors. This association is in line with finding that the infectivity of spirochetes for hamsters is maintained for a longer period upon their co- cultivation with tick cells as compared to BSK II medium alone [35]. These findings warrant further studies on the association between the 22 kDa protein and the pathoge- nicity of B. burgdorferi. Acknowledgements We thank Sofia di Nuncio and O. Rais for their help in collecting ticks and for their technical assistance. A. Barbour for providing monoclonal antibodies H5332, and T. G. Schwan for providing anti-OspC serum. We also thank B. Betschart for helpful com- ments on the manuscript. This study was partially supported by the Swiss National Science Foundation (32-29964.90) and by the Bundesministerium für Forschung und Technologie (O1K19104) and by the Commission of the European Communities (P920321). This study forms a part of PhD thesis by C. M. Hu. This study was presented in part at the VI International Confer- ence on Lyme Borreliosis, Bologna, Italy, 19-22 June, 1994. Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 255/45 C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity Zusammenfassung: Faktoren der Zecke und in v/fro-Kultivie- rung beeinflussen das Proteinmuster, die Antigenität und die Pathogenität eines aus Hämolymphe gewonnenen klonierten Borrelia garinii Isolats. Ein Borrelia garinii Isolât (NEIlH) wurde aus der Hämolymphe nicht gefütterter Ixodes ricinus Zecken gewonnen. NEIlH exprimierte vier Hauptproteine mit einem Molekulargewicht von jeweils 33,32,23 und 22 kDa. Im Laufe der in vitro Kultur entstand eine Variante (NEIlHpIS), die durch den Verlust der 22 und 23 kDa Pro- teine gekennzeichnet war. NEllHpl5 zeigte keine Reaktion mit einem Immunserum gegen OspC. Wurde NEllHpl5 in Zecken passagiert, fand sich wieder eine Reaktion mit dem an- ti-OspC-Immunserum. Ein Klon der Ausgangslinie NEIlH (cNEHH) war durch das Fehlen des 22 kDa, nicht jedoch des 23 kDa Proteins, gekennzeichnet. Das 23 kDa Protein von cNEllH wurde durch anti-OspC-Immunserum erkannt. So- wohl NEIlHpIS als auch cNEllH waren nicht in der Lage, ei- ne klinisch nachweisbare Arthritis in SCID-Mäuse zu induzie- References 1. Burgdorfer, W., Barbour, A. G., Hayes, S. F., Benach, J. L., Griin- waldt, E., Davis, J. P.: Lyme disease - a tick-borne spirochetosis? Sci- ence 216 (1982)1317-1319. 2. Burgdorfer, W., Barbour, A. G., Hayes, S. F., Péter, O., Aeschli- mann, A.: Erythema chronicum migrans - a tick borne spirochetosis. Acta Trop. 40 (1983) 79-83. 3. Gern, L., Zhu, Z., Aeschlimann, A.: Development of Borrelia burg- dorferi in Ixodes ricinus female during blood feeding. Ann. Parasit- ol. Hum. Comp. 65 (2) (1990) 89-93. 4. Ribeiro, J. M. C, Mather, T. N., Piesman, J., Spielman, A.: Dissem- ination and salivary delivery of Lyme disease spirochetes in vector ticks (Acari: Ixodidae). J. Med. Entomol. 24 (2) (1987) 201-205. 5. Zung, J. L., Lewengrub, S., Rudzinska, M. A., Spielman, A., Telford III, S. R., Piesman, J.: Fine structural evidence for the penetration of the Lyme disease spirochete Borrelia burgdorferi through the gut and salivary tissues of Ixodes dammini. Can. J. Zool. 67 (1989) 1737-1990. 6. Leuba-Garcia, S., Kramer, M. D., Wallich, R., Gern, L.: Characteri- zation of Borrelia burgdorferi isolated from different organs of Ix- odes ricinus ticks collected in nature. Zentralbl. Bakt. Hyg. 280 (1994) 468^75. 7. Lebet, N., Gern, L.: Histological examination of Borrelia burgdorf- eri infections in unfed Ixodes ricinus nymphs. Exp. Appi. Acarol. 18 (3) (1994) 177-183. 8. Hu, C. M., Leuba-Garcia, S., Aeschlimann, A., Kramer, M. D., Gern, L.: Comparison in the immunological properties of Borrelia burg- dorferi isolates from Ixodes ricinus derived from three endemic areas in Switzerland. Epidemiol. Infect. 112 (1994) 533-542. 9. Péter, O., Bretz, A. G.: Polymorphism of outer surface proteins of Borrelia burgdorferi as a tool for classification. Zentralbl. Bakt. Hyg. 277 (1992) 28-33. 10. Baranton, G., Postic, D., Saint Girons, I., Boerlin, P., Piffaretti, J. C, Assous, M., Grimont, P. A. D.: Delineation of Borrelia burgdorferi sensu stricto, Borrelia garinii sp. nov., and group VS461 associated with Lyme borreliosis. Int. J. Syst. Bacteriol. 42 (3) (1992) 378-383. 11. Canica, M. M., Nato, F., Du Merle, L., Mazie, J. C, Baranton, G., Postic, D.: Monoclonal antibodies for identification of Borrelia afze- la sp. nov., associated with late cutaneous manifestations of Lyme borreliosis. Scand. J. Infect. Dis. 25 (4) (1993) 441-448. 12. Wilske, B., Preac Mursic, V., Jauris, S., Hofmann, A., Pradel, I., Soutschek, E., Schwab, E., Will, G., Wanner, G.: Immunological and molecular polymorphisms of OspC, an immunodominant major out- er surface protein of Borrelia burgdorferi. Infect. Immun. 61 (1993) 2182-2191. ren. Nach Reinokulation von cNEllH in Zecken und an- schließender Reisolation aus verschiedenen Zeckenorganen, zeigten die meisten Reisolate eine Reaktion mit anti-OspC- Immunserum wie der Ausgangsklon cNEllH. Interessanter- weise fand sich bei zwei Reisolaten aus dem Zeckenmit- teldarm eine starke Expression des 22 kDa Proteins, welches durch das anti-OspC-Immunserum erkannt wurde. Darüber- hinaus induzierten diese beiden Reisolate eine klinisch nach- weisbare Arthritis in SCID-Mäusen. Zusammengefaßt zeigen unsere Resultate, daß die spirochätalen 22/23 kDa Proteine während der Zeckenpassage differentiell exprimiert werden. Darüberhinaus zeigen sie auch, daß Spirochätale Proteine, die nach Kultivierung in vitro nicht mehr nachweisbar sind, dann reexprimiert werden können, wenn B. burgdorferi der Mikro- umgebung des Zeckenorganismus ausgesetzt wird. Wir postu- lieren, daß die Pathogenität von B. burgdorferi in Mäusen durch umgebungsbedingte différentielle Expression der 22 und 23 kDa Proteine beeinflußt werden könnte. 13. Wilske, B., Preac Mursic, V., Schierz, G., Kühbeck, R., Barbour, A. G., Kramer, M. D.: Antigenic variability of Borrelia burgdorferi. Ann. N. Y., Acad. Sei. 539 (1988) 126-143. 14. Schwan, T. G., Burgdorfer, W., Garon, C. F.: Changes in infectivity and plasmid profile of the Lyme disease spirochetes Borrelia burg- dorferi, as a result of in vitro cultivation. Infect. Immun. 56 (1988) 1831-1836. 15. Moody, K. D., Barthold, S. W., Terwilliger, G. A.: Lyme borreliosis in laboratory animals: effect of host species and in vitro passage of Borrelia burgdorferi. Am. J. Trop. Med. Hyg. 43 (1) (1990) 87-92. 16. Hu, C. M., Gern, L., Aeschlimann, A.: Changes in the protein profile and antigenicity of different Borrelia burgdorferi strains after réin- troduction to Ixodes ricinus. Parasite Immunol. 14 (1992) 415-427. 17. Gem, L., Hu, C. M., Toutoungi, L. N., Kramer, M.: Antigenic varia- tion in Borrelia burgdorferi after passage through Ixodes ricinus and Ixodes hexagonus. In: Proceedings of the First International Confer- ence on Tick-Borne Pathology Host-Vector Interface. St. Paul, USA 1992, pp. 121-125. 18. Barbour, A. G.: Isolation and cultivation of Lyme spirochete. Yale J. Biol. Med. 57 (1984) 521-525. 19. Bundoc, V. G., Barbour, A. G.: Clonal polymorphisms of outer membrane protein OspB of Borrelia burgdorferi. Infect. Immun. 57 (1989) 2733-2741. 20. Monin, R., Gem, L., Aeschlimann, A.: A study of the modes of trans- mission of Borrelia burgdorferi by Ixodes ricinus. Zentralbl. Bakt. Hyg. (Suppl. 18) (1989) 14-20. 21. Gern, L., Toutoungi, L. N., Hu, C. M., Aeschlimann, A.: Ixodes (Pholeoixodes) hexagonus, an efficient vector of Borrelia burgdorf- eri in the laboratory. Med. Vet. Entomol. 5 (1991) 431-435. 22. Schwan, T. G., Schrumpf, M. E., Karstens, R. H., Clover, J. R., Wong, J., Daugherty, M., Struthers, M., Rosa, P. A.: Distribution and molecular analysis of Lyme disease spirochetes, Borrelia burg- dorferi, isolated from ticks throughout California. J. Clin. Microbiol. 31 (1993) 3096-3108. 23. Barbour, A. G., Schrumpf, M. E.: Polymorphisms of major surface protein of Borrelia burgdorferi. Zentralbl. Bakt. Hyg. (A) 263 (1986) 83-91. 24. Kramer, M. D., Schaible, U. E., Wallich, R., Moter, S. E., Petzoldt, D., Simon, M. M.: Characterization of Borrelia burgdorferi associat- ed antigens by monoclonal antibodies. Immunobiol. 181 (1990) 357-366. 25. Schaible, U. E., Gay, S., Museteanu, C, Kramer, M. D., Zimmer, G., Eichmann, K., Museteanu, U., Simon, M. M.: Lyme borreliosis in the severe combined immunodeficiency (seid) mouse manifests predom- 46/256 Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 C. M. Hu et al.: Phenotypic Variations and B. burgdorferi Pathogenicity inantly in the joints, heart, and liver. Am. J. Pathol. 137 (1990) 811-820. 26. Schaible, U. E., Wallich, R., Kramer, M. D., Gern, L., Anderson, J. F., Museteanu, C, Simon, M. M.: Immune sera to individual Borre- lia burgdorferi isolates or recombinant OspA thereof protect SCID mice against infection with homologous strains but only partially or not at all against those of different OspA/OspB genotype. Vaccine 11 (1993) 1049-1054. 27. Schwan, T. G., Burgdorfer, W.: Antigenic changes of Borrelia burg- dorferi as a result of in vitro cultivation. J. Infect. Dis. 156 (5) (1987) 852-853. 28. Rosa, P. A., Hogan, D. M.: Colony formation by Borrelia burgdorf- eri in solid medium clonal analysis of Osp locus variants. In: Proceed- ings of the First International Conference Tick-Borne Pathology Host-Vector Interface. St. Paul, USA 1992, pp. 95-103. 29. Masuzawa, T., Kurita, T., Kavabata, H., Yanagihara, Y.: Relation- ship between infectivity and OspC expression in Lyme disease Bor- relia. FEMS Microbiol. Lett. 123 (1994) 319-324. 30. Schwan, T. G., Piesman, J., Golde, W. T., Dolan, M. C, Rosa, P. A.: Induction of an outer surface protein on Borrelia burgdorferi during tick feeding. Proc. Natl. Acad. Sci. USA 92 (1995) Microbiology pp. 2909-2913. Book Review___________________________ A. Bauernfeind, M. I. Marks, B. Strandvik (eds.) Cystic Fibrosis Pulmonary Infections: Lessons from Around the World 352 pages Birkhäuser, Basel 1995 Price: DM 268. This book is very interesting and unique since it gathers for the first time clinically very important information about the preva- lence of the different microbial pulmonary infections in cystic fi- brosis patients, treatment regimes, survival curves, lung-function studies etc. from all over the world. There is no other source of this wealth of information currently available, and it is therefore highly recommended to clinicians and scientists interested in cys- tic fibrosis. The book was initiated after the International Con- gress of Chemotherapy in Berlin 1991 by the three editors from Germany, USA and Sweden. They have recruited over 90 au- thorities from all over the world to contribute, including authors from 21 different countries: Canada, USA, Mexico, Argentina, Venezuela, Uruguay, Australia, Japan, Russia, Israel, Turkey, Greece, Italy, The Czech Republic, Spain, Hungary, Germany, United Kingdom, The Netherlands, Sweden and Denmark. All authors work in well-known centers, and therefore all clinical as- pects of diagnosis and treatment of CF pulmonary infection are covered. 31. Ma, B., Christen, B., Leung, D., Vigo-Pelfrey, C: Serodiagnosis of Lyme borreliosis by Western immunoblot: reactivity of various sig- nificant antibodies against Borrelia burgdorferi. J. Clin. Microbiol. 30 (1992) 370-376. 32. Wilske, B., Preac Mursic, V., Schierz, G., Liegi, G., Gueye, W.: De- tection of IgM and IgG antibodies to Borrelia burgdorferi using dif- ferent strains as antigen. Lyme borreliosis II, Zentralbl. Bakt. Hyg. (Suppl. 77) (1989) 299-309. 33. Gern, L., Schaible, U. E., Simon, M. M.: Mode of inoculation of the Lyme disease agent Borrelia burgdorferi influences infection and im- mune responses in inbred strains of mice. J. Infect. Dis. 167 (1993) 971-975. 34. Margolis, N., Hogan, D., Tilly, K., Rosa, P. A.: Plasmid location of Borrelia purine biosynthesis gene homologues. J. Bacteriol. 176 (1994) 6427-6432. 35. Kurtti, T. J., Munderloh, U. G., Krueger, D. E., Johnson, R. C, Schwan, T. G.: Adhesion to and invasion of cultured tick (Acarina: Ixodidae) cells by Borrelia burgdorferi (Spirochaetales: Spirochaeta- ceae) and maintenance of infectivity. J. Med. Entomol. 30 (3) (1993) 586-596. The first section contains six chapters on viral, fungal, and atypi- cal bacterial infection in cystic fibrosis, pharmacokinetics of an- tibacterials in cystic fibrosis, antimicrobial pharmacotoxicity, microbial resistance, microbial virulence and pathogenesis in cystic fibrosis, and lung transplantation. The second section contains 22 chapters describing the general approach to cystic fibrosis and pulmonary infection in 22 diffe- rent countries (West and East Germany are represented by two different chapters). These chapters are highly interesting and in- structive. They show the results of many years of intensive efforts in CF centers in Europe and North America, where the median survival of patients is now over 30 years, in contrast to the pro- blems which still exist in other countries, for instance in South America where intensive effort in CF centers only started some 10 years ago, and where the survival is much shorter. It is there- fore possible, for the first time in one publication, to compare the present situation in countries where centers have just started with the results in countries where centers were established over 30 years ago, and realize how much the care of CF patients can be improved by intensive team-work over many years. The book is filled with useful clinical details with regard to regimes, sur- veillance etc. and it can be recommended as a standard book for any CF center in the world. N. H0iby Copenhagen Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 257/47 Case Report P. Meuleman, K. Erard, M. C. Herregods, W. E. Peetermans, J. Verhaegen Bioprosthetic Valve Endocarditis Caused by Neisseria elongata Subspecies nitroreducens Summary: A new case of Neisseria elongata ssp. nitroreducens bacteremia and endocardi- tis in a 74-year-old woman who had undergone aortic valve replacement in 1992 is report- ed in detail. N. elongata ssp. nitroreducens differs from the other subspecies of N. elonga- ta in the additional reduction of nitrate without gas formation. Like most Neisseria spp. ex- cept Neisseria meningitidis and Neisseria gonorrhoeae, this N. elongata ssp. nitroreducens is usually classified in the group of "non-pathogenic" Neisseria spp. This case report indi- cates that the presence of subspecies of this group is significant when isolated from normal- ly sterile sites and can cause severe disease in susceptible individuals. Introduction Neisseria elongata subspecies nitroreducens is a non-mo- tile, short, gram-negative coccobacillus. It is considered to be part of the normal human oral flora. Despite the fact that only Neisseria gonorrhoeae and Neisseria meningitidis are universally accepted as pathogens in the genus Neisse- ria, some of the so-called "non-pathogenic" Neisseria spe- cies, in particular N. elongata ssp. nitroreducens, can cause serious infections. The most common illnesses noted with this subspecies were endocarditis, bacteremia and osteo- myelitis [I]. We report in detail on the isolation of a N. elongata ssp. ni- troreducens from blood cultures of a patient with a previ- ous history of valve damage and replacement, hospitalized with the diagnosis of possible infective endocarditis. Case Report A 74-year-old woman was admitted to hospital because of re- peated episodes of fever. In 1988 a pacemaker was implanted be- cause of a high-degree atrioventricular heart blockage, and in 1992 she underwent a biological prosthetic aortic valve implan- tation because of severe aortic stenosis. The medication for ven- tricular arrhythmia consisted of flecainide. The patient had been well until 1 month before admission, when she presented with fe- ver and vague discomfort. There was no anamnestic or clinical clue to point to the cause of the fever, so the general practition- er started a treatment course of amoxicillin (500 mg t.i.d. for 4 days). Initially the fever disappeared, but 4 days after withdraw- al of the antibiotic, temperature rose again and a quinolone (ofloxacin 200 mg b.i.d. for 10 days) was administered. Despite an initial benefit, the fever soon returned after withdrawal of ofloxacin. The patient was then admitted to the hospital. On physical examination she appeared well. Her temperature was 36.4°C. There were no signs of heart failure and the skin was normal (no signs of vasculitis, no splinter hemorrhages). Labora- tory data included a leucocyte count of 11,900/mm3 (with 86.3% neutrophils), a platelet count of 159,000/mm3, a CRP of 15.5 g/dl and an ERS of 54 mm/h and no microscopic hematuria. Pro- thrombin time, immune complexes and complement were nor- mal. An electrocardiogram presented normal pacemaker activity on demand. A radiograph of the chest showed cardiac enlargement, a slight left pleural effusion and signs of chronic bronchitis. Fun- doscopy showed neither Roth spots nor a clue for the cause of fe- ver. Transthoracic cardiac ultrasound examination did not reveal any vegetation. The prosthetic biological aortic valve appeared normal. The left and right ventricular functions were normal. The atria were slightly dilated. There was a peak instantaneous pressure gradient over the aortic valve of 55 mmHg and a slight central aortic regurgitation, a moderate mitral regurgitation and a tricuspid regurgitation without pulmonary hypertension. Transoesophageal cardiac ultrasound examination revealed de- generative changes of the mitral valve, a slight thickening of the aortic wall at the prosthetic valve implantation, but no signs of abscess formation and no vegetation. Three days after admission the patient developed several epi- sodes of high fever (39.3°C). During each episode, blood cultures were taken and in 19 vials of 24 blood culture sets N. elongata ssp. nitroreducens was found. Stomatological examination revealed poor dental hygiene, and on the orthopantomogram a small ra- diolucent spot could be detected at the apex of tooth 36. Following Durack's new criteria for diagnosis of infective endo- carditis [2], since one major and two minor criteria were met, the diagnosis of possible infective endocarditis could be made. The patient received gentamicin (160 mg - 240 mg p.d.) for 14 days and ampicillin (6 x 2 g p.d.) for 6 weeks. The fever disap- peared completely within 12 h. The inflammatory parameters normalised progressively. The pleural effusion also cleared after 1 week. A weekly control of the transthoracic cardiac ultrasound did not reveal any change. Meanwhile, tooth 36 was extracted. Microbiological Examination For 7 successive days we received a total of 24 blood culture sets from this patient at our laboratory. They were examined daily with the Bactec System (Bactec 730; high volume resine, Bactec plus 26 (aerobic) and 27 (anaerobic), Becton-Dickinson). Out of Received: 9 August 1995/Revision accepted: 27 November 1995 Dr. P. Meuleman, Prof. Dr. J. Verhaegen, Dept. of Microbiology, Bacte- riology; Dr. K. Erard, Prof. Dr. M. C. Herregods, Dept. of Internal Med- icine, Cardiology; Prof. Dr. W. E. Peetermans, Dept. of Internal Medi- cine, General Internal Medicine, University Hospitals, B-3000 Leuven, Belgium. Correspondence to: Prof. Dr. J. Verhaegen, Dept. of Microbiology, Uni- versity Hospital St. Raphael, Capucijnenvoer, 35, B-3000 Leuven, Bel- gium. 48 / 258 Infection 24 (1996) No. 3 © MMV Medizin Verlag GmbH München, München 1996 ZbI. Bakt. 285, 558-564 (1997) © Gustav Fischer Verlag, Jena Apodemus sp. Rodents, Reservoir Hosts for Borrelia afzelii in an Endemie Area in Switzerland CHANG MIN HU1, PIERRE-FRANCOIS HUMAIR1, REINHARD WALLICH2 and LISE GERN1 1 Institut de Zoologie, Université de Neuchâtel, Switzerland 2 Angewandte Immunologie, FS 0440, Deutsches Krebsforschungszentrum, Heidelberg, Germany Received March 29, 1996 • Revision received May 22, 1996 • Accepted July 9, 1996 Summary Borrelia burgdorferi is maintained in nature in transmission cycles alternatively involv- ing ticks and reservoir hosts. Small rodents like Apodemus mice and Clethrionomys voles are the primary reservoir of Lyme disease in Europe. In this study, we analyzed by SDS- PAGE and Western blot 20 borrelial isolates from xenodiagnostic ticks fed on four Apode- mus sp. mice captured in the Staatswald forest (Switzerland). All isolates but one showed a homogeneous protein pattern expressing an outer surface protein, (Osp) A of 32 kDa and an OspB of 35 kDa and reacted with monoclonal antibody (mAb) I 17.3 specific for B. af- zelii. One isolate expressed an OspA of 32.5 kDa and an OspB of 35 kDa and did not react with species-specific mAbs I 17.3, D6 and H3TS, but was shown to belong to B. afzelii by Southern blot analysis. The possibility exists that non-cultivatable borreliae are present in xenodiagnostic ticks. However, our results clearly show that Apodemus sp. are reservoir hosts for B. afzelii, since this genospecies is transmitted from Apodemus sp. to feeding lar- val ticks. Introduction The enologie agent of Lyme disease, Borrelia (B.) burgdorferi, is maintained in en- zootic cycles involving wild vertebrate hosts and different species of ticks belonging primarily to the Ixodes (I.) ricinus complex (12). Ticks acquire the infection during a blood meal on infected animals, maintain transstadially spirochetes, and transmit these bacteria to other animals during the next feeding (3). In Europe, Apodemus mice have been shown to be important reservoirs for 5. burgdorferi si being infective for I. rici- nus (5, 10, 14). Various studies have shown the phenotypic and genotypic heterogeneity of B. burg- dorferi isolates with respect of their abundant outer surface proteins (Ops) A, B and C (19, 20, 21). Now, three genospecies are known to cause Lyme borreliosis in Europe: B. burgdorferi sensu stricto, B. garinii and B. afzelii (2, 4). B. burgdorferi ss expresses Burkholderia pickettii in Acanthamoeba 557 References l.Burchard, CD. and M.Bierther: Study on the micromorphology of Mycobacterium leprae. Arch. Dermatol. Res. 277 (1985) 220-224 2. De Jonckheere, J. F. and H. van de Voorde: Differences in destruction of cysts of patho- genic and non pathogenic Naegleria and Acanthamoeba by chlorine. Appi. Environ. Mi- crobiol. 31 (1976) 294-297 3. De Jonckheere, J. R: Use of an axenic medium for differentiation between pathogenic an non-pathogenic Naegleria fowleri isolates. Appi. Environ. Microbiol. 33 (1977) 751- 757 4. Gilardi, G. L.: Pseudomonas and related genera. In: Manual of clinical microbiology, Eds. Balows, W.J. Hausier Jr., K.L. Herrmann, H.D. Isenberg and H.J. Shadomy, p. 429-441 5th Ed. American Society for Microbiology, Washington, D.C. (1991) 5. Hahn, H., D. Falke und P. Klein: Legionellen. In: Medizinische Mikrobiologie, S. 367- 371, Springer Verlag, Berlin - Heidelberg - New York (1991) 6. Hauröder-Philippczyk, B. and R. Michel: Pseudomonas pickettii as intracellular para- site in Acanthamoebae. Bioforum 14 (1991) 34 7. Kilvington, S. and J. Price: Survival of Legionella pneumophila within cysts of Acantha- moeba polyphaga following chlorine exposure. J. Appi. Bacteriol. 68 (1990) 519-525 8. Ly, T. M. C. and H. E. Müller: Interactions of Listeria monocytogenes, Listeria seeligeri, and Listeria innocua with protozoans. J. Gen. Appi. Microbiol. 36 (1990) 143-150 9. Menn, Th. und R. Michel: Untersuchung gebräuchlicher Flächendesinfektionsmittel auf ihre nematozide und amoebizide Wirkung. Krankenhaushygiene und Infektionsverhü- tung 6 (1988) 158-164 10. Michel, R. and B. Hauröder-Philippczyk: Observation of a natural infection of Acanth- amoeba sp. by Pseudomonas pickettii. J. Protozool. 39 (1992a) 15A 11. Michel, R. and B. Hauröder-Philippczyk: Cocultivation of Acanthamoeba castellami and Pseudomonas aeruginosa leads to infection of the amoebae. Publication of the Japanese-German-Center Berlin, Series B 5 (1992b) 174-178 12. Michel, R. und Th. Menn: Acanthamoeben, Naeglerien und Invertebraten in Feuchtbe- reichen von Physiotherapieeinrichtungen in Krankenhäusern. ZbI. Hyg. B 191 (1991) 423-473 13. Müller, H. E. und T. M. C. Ly: Über die Phagocytose von Listerien durch Protozoen und einige sich daraus ergebende Schlußfolgerungen zur Pathogenität und Virulenz. 42. Ta- gung DGHM Hannover, FRG (1989) 14. Page, F. C: An illustrated key to freshwater and soil amoebae. Freshwater Biological As- sociation, The Ferry House, Ambleside, Cumbria (1976) 15. Ralston, E., N.J. Palleroni, and M. Doudoroff: Pseudomonas pickettii, a new species of clinical origin related to Pseudomonas solanacearum. Int. J. System. Bacteriol. 23 (1973) 15-19 16. Riley, P. S. and R. E. Weaver: Recognition of Pseudomonas pickettii in the clinical la- boratory: Biochemical characterization of 62 strains. J. Clin. Microbiol. 1 (1975) 61-64 17. Rowbotham, TJ.: Preliminary report on the pathogenicity of Legionella pneumophila for freshwater and soil amoebae. J. Clin. Pathol. 33 (1980) 1179-1183 Rolf Michel und Bärbel Hauröder, Ernst-Rodenwaldt-Institut, Postfach 7340, D-56065 Koblenz, Germany Apodemus rodents as reservoirs for B. afzelii 559 an outer surface protein (Osp) A of 31 IcDa and an OspB of 34 IcDa, B.garinii has an OspA of 32-33 kDa and B. afzelii expresses an OspA of 32 IcDa and an OspB of 35 IcDa (2, 4). The expression of OspC with an approximate molecular weight of 22 kDa is more variable than that of the OspA and OspB among B. burgdorferi si strains (21). The transmission cycle of the three genospecies in nature needs to be clarified. Re- cently, Humair et al. (11) demonstrated that ears of wild rodents were exclusively in- fected by B. afzelii. A preferential association seems to exist between rodents and B. af- zelii and rodents may be important reservoir hosts for this Borrelia species in nature. In order to evaluate the ability of Apodemus sp. to transmit Lyme borreliosis genospe- cies to feeding ticks, we used tick xenodiagnosis on naturally infected Apodemus ro- dents and analyzed isolates from xenodiagnostic ticks. Materials and Methods Mice and ticks Four naturally infected rodents, three Apodemus (A.) sylvaticus (G149, Gl85 and G210) and one A. flavicollis (G220) captured in the Staatswald forest (Berne, Switzerland) (10), were used as sources of infection for I. ricinus larvae. In a previous study (5), we demon- strated that G149 transmitted B. burgdorferi si. to 62.8% of I. ricinus larvae, G210 to 31.3% and G220 to 81.4%. I. ricinus larvae derived from our laboratory colony (7), being free of borrelial infection, were used for xenodiagnosis. Approximatively 50 I. ricinus larvae were placed on the head of each Apodemus rodent. Engorged larvae were maintained in glass vials at room temper- ature and at 100% relative humidity until moulting had become completed. B. burgdorferi isolations Unfed nymphs resulting from the engorged larvae were used for isolation of B. burgdor- feri si.: 10 nymphs from G149, 14 nymphs from Gl85, 12 nymphs from G210 and 12 nymphs from G220. Nymphal ticks were washed in 70% ethanol for 1 min and rinsed in sterile PBS pH 7.4. Each nymph was cut into two pieces using sterilized scissors and placed into tubes containing 4 ml BSK II medium (1) supplemented with phosphomycin (50 ug/ml, Boehringer, Mannheim, Germany) and rifampin (50 ul/ml, Ciba-Geigy, Basel, Switzerland) (8). Culture tubes were incubated at 34 0C and examined by dark-field micros- copy every 5-7 days for two months. All positive cultures (1 ml) were inoculated into 25 ml BSK II medium for further growth and then prepared for sodium dodecyl sulfate- polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblot analysis. SDS-PAGE and immunoblot analysis Isolates were subcultured 2-3 times before analysis by SDS-PAGE and Western blot. SDS- PAGE and immunoblot analysis were carried out according to the methods described pre- viously (8). All isolates were centrifuged at 10000 g for 20 min and washed twice with PBS supplemented with MgCI2 (5 mM). Whole-cell lysates of each isolate were electrophorized on a Polyacrylamide gel at 12.5% for the separating gel and 6% for the stacking gel with a borrelial concentration of 107 cells/lane (Helber cell counting chamber). The gels were stained with Coomassie brilliant blue R 250. The proteins from whole cell lysates separated by SDS-PAGE were transferred to nitro- cellulose paper using a transit cell (2117-250 Nova Blot Electrophoretic Transfer Kit. LKB AB Bromma, Sweden). The following monoclonal antibodies (mAbs) were used to identify the three genospecies as previously described (11): H3TS against OspA protein of B. burg- dorferi ss (2); I 17.3 against OspB of B. afzelii (4) and D6 against 12 kDa protein of B.ga- 560 Chang Min Hu et al. rinii (2). An immune serum (IS anti-22kDa/NE4) was used to examine the expression and the antigenicity of the 22 kDa protein (8). IS anti-22 kDa/NE4 was produced by immuniz- ing a New Zealand white rabbit with the 22 kDa protein of NE4 strain (8) identified as OspC by Wilske (personal communication). Immunocomplexes were detected using perox- idase-labelled anti-mouse IgG antibody (for H3TS and I 17.3), IgM antibody (for D6) and anti-rabbit IgG antibody (for IS anti-22KDa/NE4) diluted 1:1000 (Nordic Immunological Laboratories, Netherlands). Southern blot hybridization Total genomic DNA was extracted from NE388, an isolate which did not react with H3TS, I 17.3 and D6, and digested with 100 units of restriction endonuclease Hind III. Southern blot was performed as described previously (19). Statistical test Fischer's exact test was used to compare the tick infection rates determined by immuno- fluorescence of ticks fed on G149, G210 and G220 (5) with the tick infection rate deter- mined by Borrelia isolation. The difference was considered as significant when p < 0.05. Results A total of 20 isolates was obtained from 48 nymphs fed as larvae on four infected Apodemus rodents. Three isolates were obtained from the cultivation of 10 nymphs (30%) fed on G149, 5 from 14 nymphs (35%) fed on G185, 3 from 12 nymphs (25%) fed on G210 and 9 from 12 (75%) nymphs fed on G220 (Fig. 1). In order to elucidate whether non-cultivatable Borrelia could be present in the xenodiagnostic ticks, we compared the success of borrelial isolation obtained in this study with the infection rate of xenodiagnostic ticks fed on G149 (155/247, 62.8%), G210 (88/281, 31.3%) and G220 (219/269, 81.4%) described previously (5) and determined by immunoflu- orescence. No significant differences were observed for G210 and G220 (p = 0.759 and p = 0.704, respectively) and a weakly significant difference was observed for G149 (p = 0.048). The infection rate of ticks fed on G185 is unknown. All tick isolates but one (NE388) show similar protein profiles expressing an OspA of 32 kDa and an OspB of 35 kDa. Western blot analysis showed that all isolates ex- cept NE388 reacted only with mAb I 17.3 that identifies B.afzelii (Fig. 1). Isolate NE388 which had an OspA of 32.5 kDa and an OspB of 35 kDa (Fig. 1) did not react with any of the 3 mAbs but was identified as B. afzelii by Southern blotting (Fig. 2). The expression of the 22 kDa (OspC) protein was variable among the 20 tick iso- lates and it presented heterogeneous reactions with Is anti-22 kDa/NE4 (Fig. 1). Discussion Lyme borreliosis spirochetes isolated from ticks and humans show a great pheno- typic and genotypic heterogeneity (2,19,20). At present, eight Borrelia genotypes have been described (16). In Europe, at least three B. burgdorferi genospecies can coexist in the tick population and circulate between tick vectors and reservoir hosts (9). At present, the transmission of these Borrelia genospecies from different reservoir hosts to feeding ticks is still unclear. G149 "GÏ55 Apodemus rodents as reservoirs for B. afzelii |C±iô IC22Ô 561 Mw (kDa) 97.5 66.2 45.0 -- ta»» 31.0 -- W W W W J*) u w w a a ZZZZZ M » O 00 OC O rt f> f> Cd W W ZZZ « m fj ^ «n >o r~ » a ^ » a. a a «»> w «*> »*> «O W « U H |d U U Id Id ZZZZZZZ (d (d Z Z 21.0 .. 14.5 -- Western blot I 17.3 + + + + + + + + -++ + + + + + + + + + H3TS .................... D6 .................... anti-22 +-- + + + + - +-+ + +.--+ + .+ kDa/NE4 Fig. 1. Protein patterns of B. burgdorferi isolates from ticks fed on the four infected Apode- mus rodents, G149, Gl 85, G210 and G220. The gels were stained by Coomassie brilliant blue R 250 and the protein molecular weight standards are shown on the left. In Europe, Apodemus mice and Clethrionomys voles are important reservoirs for B. burgdorferi si, transmitting the infection to a great number of larvae (5, 10, 14). These rodents are infested with I. ricinus larvae and nymphs harbouring one or more Borrelia genospecies (11, 15, 17). Recent studies have shown that different B. burgdor- feri genospecies can infect the internal organs of small mammals (6,13). In contrast, Humair et al. ( 11 ), described only B. afzelii to be prominent in the ears of rodents cap- tured in two endemic areas and these authors suggested that rodents could be reser- voirs for ß. afzelii. In the present study, we showed that all isolates obtained from xenodiagnostic I. rici- nus ticks, fed on naturally infected rodents, were identified as ß. afzelii. The possibil- 562 Chang Min Hu et al. Fig. 2. Southern blot hybridization of NE388 and other B. burgdorferi species. The sources of DNA are indicated above the lanes. ACA-I and MMS represent B. afzelii, ZQl repre- sents B.garinii and ZS7 represents B. burgdorferi sensu stricto. NE388 showed a single Hind III DNA fragment of 1.7kb for the ospA gene as B. afzelii strain ACA-I. ity exists that borreliae non-cultivatable are present in xenodiagnostic ticks. However, the success of isolation was not significantly different from the infection rates of ticks fed on G210 and G220 and determined by immunofluorescence, and only weakly sig- nificantly different for ticks fed on G149. All rodents were captured in an area where isolates from unfed field-collected ticks belonged to B. burgdorferi ss, B.garinii, B.afzelii and group VS116 (9, Hutnair, per- sonal communication). This assumes that various Borrelia species circulate between reservoirs and ticks in this area. Since the main role of a reservoir host is to transmit infection to feeding ticks, we have shown in the present study that rodents are reser- voirs for B. afzelii. Rodents are important hosts for I. ricinus larvae (10) which impli- cates that B. afzelii might mainly be found in unfed nymphs. In fact, studies have dem- onstrated that most of B. burgdorferi isolates from field-collected nymphs belonged to B. afzelii (11, Humair, personal communication). This seems to support our viewpoint. Protein and antigenic variations among the B. afzelii isolates were observed: an iso- late (NE388) presented an OspA of 32.5 kDa and did not react with mAb 117.3 and moreover, a great heterogeneity in the expression of OspC was present among isolates from I. ricinus ticks fed on the same rodent. In contrast, B. afzelii isolates obtained from Apodemus and CIethrionomys ear biopsies by Humair et al. (11) were much more ho- mogeneous. Successive isolations from the same host did not show variation in the ex- pression of Osps. The reason of the difference observed in Osp expression between ro- dents isolates and rodent-feeding tick isolates remains unknown but could be related to the tick environment as suggested in previous studies (8). These results also show that variation in the Osps expression exists during the natural transmission cycle. Apodemus rodents as reservoirs for B. afzelii 563 In conclusion, our study shows that small rodents are reservoirs of B. afzelii since they transmit this genospecies to feeding ticks. The exact reason for such an associa- tion remains unknown. B. afzelii is observed in human skin lesions (4, 18) and in ro- dent skin biopsies (11) and since ticks attach to animal skin, I. ricinus larvae infesting rodents pick up B. afzelii. Acknowledgements. This work forms a part of the PhD thesis of Hu, C. M. We thank Olivier Rais for technical assistance, Alan Barbour, O. Péter and G. Baranton for provid- ing monoclonal antibodies. This work was partially supported by the Swiss National Sci- ence Foundation (32-29964.90) and the Federal Office for Education and Science (No. 93.0363) as part of the European Union Biomedical and Health Research (Biomed I) Programme (No.BMH-CT93-1183). References 1. Barbour A. G.: Isolation and cultivation of Lyme disease spirochetes. Yale J. Biol. Med. 57 (1984) 521-525 2. Baranton, G., D. Postic, I. Saint Girons, P. Boerlin, J. C. Piff aretti, M. Assous, and P. A. D. Grimont: Delineation of Borrelia burgdorferi sensu stricto, Borrelia garinii sp. nov., and group VS461 associated with Lyme borreliosis. Int. J. Syst. Bacteriol. 42 (1992) 378-383 7>.Burgdorf er, W., A. G. Barbour, S. F. Hayes, O. Péter, and A. Aeschlimann: Erythema chronicum migrans - a tick borne spirochetosis. Acta Tropica, 40 (1983) 79-83 4. Conica, M. M., F. Nato, L. Du Merle, ]. C. Mazie, G. Baranton, and C. Postic: Monoclo- nal antibodies for identification of Borrelia afzelii sp. nov. associated with late cutane- ous manifestations of Lyme borreliosis. Scand. J. Infect, dis. 25 (1993) 441-448 5. Gern, L., M. Siegenthaler, C. M. Hu, S. Leuba-Garcia, P. E Humair, and J. Moret: Bor- relia burgdorferi in rodents (Apodemus flavicollis and A. sylvaticus): duration and en- hancement of infectivity for Ixodes ricinus ticks. Eur. J. Epidemiol. 10 (1994) 75-80 6. Gorelova, N. B., E. I. Korenberg, Y V. Kovalevskii, and S. V. Shcherbakov: Small mam- mals as reservoir hosts for Borrelia in Russia. ZbI. Bakt. 282 (1995) 315-322 7. Graf, J. E: Copulation, nutrition et ponte chez Ixodes ricinus L. (Ixodoidea: Ixodidae)- lère partie. Bulletin de la Société Entomologique Suisse, 51 (1978) 89-97 8. Hu, CM., L. Gern, and A. Aeschlimann: Changes in the protein profile and antigenic- ity of different Borrelia burgdorferi strains after reintroduction to Ixodes ricinus ticks. Parasite Immunol. 14 (1992) 415-427 9. Hu, C. M., S. Leuba-Garcia, A. Aeschlimann, M. D. Kramer, and L. Gern: Comparison in the immunological properties of Borrelia burgdorferi isolates from Ixodes ricinus de- rived from three endemic areas in Switzerland. Epidemiol. Infect. 112 (1994) 533-542 10. Humair, P. F., N. Turrian, A. Aeschlimann, and L. Gern: Borrelia burgdorferi in a focus of Lyme borreliosis: the epizootiologic contribution of small mammals. Folia Parasito- logica, 40 (1993) 65-70 11. Humair, P. E, O. Péter, R. Wallich, and L. Gern: Strain variation of Lyme spirochetes iso- lated from Ixodes ricinus ticks and rodents collected in two endemic areas in Switzer- land. J. Med. Entomol. 32 (1995) 433-438 12. Jaenson, T. G. T.: The epidemiology of Lyme borreliosis. Parasitology Today 7 (1991) 39-45 13. Khanakah, G., E. Kmety, A. Radda, and G. Stanek: Micromammals as reservoir of Bor- relia burgdorferi in Austria. Abstract. In: VI Intern. Conf. on Lyme Borreliosis. R. Ce- venini, V. Sambri, and M. la Placa (eds.), Bologna, Italy, Juin 19-20, 1994, P077W 14. Kurtenbach, K., H. Kampen, A. Dizij, S. Arndt, H. M. Steitz, U. E. Schaible, and M. M. Simon: Infestation of rodents with larval Ixodes ricinus (Acari: Ixodidae) is an impor- 37 ZbI. Bakt. 285/4 564 Chang Min Hu et al. tant factor in the transmission cycle of Borrelia burgdorferi s.l. in German woodlands. J. Med. Entomol. 32 (1995) 200-210 15. Leuba-Garcia, S., M. D. Kramer, R. Wallicb, and L. Gern: Characterization of Borrelia burgdorferi isolated from different organs of Ixodes ricinus ticks collected in nature. ZbI. Bakt. 280 (1994) 468-475 16. Postic, D., M. V. Assous, P. A.D. Grimont, and G. Baranton: Diversity of Borrelia burg- dorferi sensu lato evidenced by restriction fragment length polymorphism of rrf (5S) - rrl (23S) intergenic spacer amplicons. Int. J. Syst. Bacteriol. 44 (1994) 743-752 17. Rijpkema, S. G. T., M.G.C. H. Molkenboer, L. M. Schouls, F.Jongejan, and J. R P. Schellekens: Simultaneous detection and genotyping of three genomic groups of Borre- lia burgdorferi sensu lato in Dutch Ixodes ricinus ticks by characterization of the am- plified intergenic spacer region between 5S and 23S rRNA genes. J. Clin. Microbiol. 33 (1995)3091-3095 18. van Dam, A.P., H.Kuiper, K.Vos, A. Widjojokusumo, B. M. Dejongh, L.Spanjaard, A. C. P. Ramselaar, M. D. Kramer, and J. Dankert: Different genospecies of Borrelia burgdorferi are associated with distinct clinical manifestations of Lyme disease. Clin. In- fect. Dis. 17 (1993) 708-717 19. Wallich, R., S. E. Moter, M. D. Kramer, L. Gern, H. Hofmann, U. E. Schaible und M. M. Simon: Untersuchungen zur genotypischen und phänotypischen Heterogenität von Bor- relia burgdorferi dem Erreger der Lyme-Borreliose, pp. 176-190 (1992). In: D. Hassler, M. Kramer, M. Maiwald, W. Mardet, and L. Zöller (eds.), Infection, Fortschritte der In- fektiologie. MMV Medizin Verlag, Munich 20. Wilske, B., V.Preac-Mursic, U.B. Göbel, B.Graf, S.Jauris, E.Soutschek, E.Schwab, and G. Zumstein: An OspA serotyping system for Borrelia burgdorferi based on reac- tivity with monoclonal antibodies and OspA sequence analysis. J. Clin. Microbiol. 31 (1993) 340-350 21. Wilske, B., V.Preac-Mursic, S.Jauris, A.Hofmann, I.Pradel, E.Soutschek, E.Schwab, G. Will, and G. Wanner: Immunological and molecular polymorphisms of OspC, an im- munodominant major outer surface protein of Borrelia burgdorferi. Infect. Immun. 61 (1993)2182-2191 Lise Gern, Institut de Zoologie, Université de Neuchâtel, Emile-Argand 9, CH-2000 Neuchâtel, Switzerland, Tel.: 4138233052, Fax: 4138233001