Zbl. Bakt. 277, 179-187 (1992) © Gustav Fischer Verlag, Stungart/New York
Subtyping of Legionella pneumophila Serogroup 1 Strains Isolated in Germany Using Monoclonal Antibodies PAUL CHRISTIAN LOCK l , JORGEN HERBERT HELBIG!, WERNER EHRET 2 , REINHARD MARRE 3 , and WOLF WITZLEB l 1
2
3
Institut fiir Medizinische Mikrobiologie, Medizinische Akademie Dresden, 0-8019 Dresden Max-von-Pettenkofer-Institut fur Hygiene und Mikrobiologie. Ludwig-Maximilian-Universitat Munchen, W-8000 Munchen Institut fur Medizinische Mikrobiologie, Medizinische Universitat Lubeck, W-2400 Lubeck, Germany
With 1 Figure· Received August 16, 1991 . Revision received January 27, 1992· Accepted March 10, 1992
Summary Legionella pneumopbila serogroup 1 (Lpl ) strains isolated from patients and hot water supplies in different locations of Germany were subtyped using seven monoclonal antibodies (mabs) in the indirect immunofluorescence test (IFA) and in part, using a dot blot assay. Four of these mabs were produced in Dresden. Three mabs (mab 33G3, mab 32A12 and mab 144c2) were kindly supplied by]. ]oly, Quebec, Canada. Altogether, seven antigenic variants were found among Lpl strains isolated in Germany. Patient strains belonged to the Philadelphia, Benidorm, Knoxville, France, Olda-Heysham and Bellingham subgroups, whereas environmental isolates reacted like the Bellingham, Oxford, Philadelphia, Knoxville and France strains. The majority of patient strains (15 out of 26, 58 %) reacted with our mab 3/1 (corresponding to mab 2 of the standard panel), but only 26 out of 118 environmental strains (22% ) isolated from 4 of 15 hot water supplies did so (p < 0.05). The majority of water-borne Lp1 strains reacted with a mab specific of the Bellingham subgroup. Three water systems under study were associated with human legionellosis. Two of them contained Bellingham-like strains, one Philadelphia-like legionellae.
Zusammenfassung Legionella pneumophila Serogrouppe 1 (Lp l ) Stamrne, die von Patienten und aus Wassersystemen in verschiedenen geographischen Regionen Deutschlands isoliert worden waren, wurden mit sieben monoklonalen Antikorpern (mab) im indirekten Immunofloureszenztest (lFA) und teilweise in einem Dot-Blot-Assay subtypisiert. Vier der verwendeten mab wurden in Dresden produziert. Drei (mab 33G3, mab 32A12 und mab 144C2) wurden freundlicherweise von j. ]oly, Quebec, Kanada, uberlassen. Insgesamt wurden sieben antigene Varianten unter den Isolaten aus Deutschland gefunden. Patientenstamme gehorten zu
180
P. C. Luck,
J. H. Helb ig, W. Ehret, R. Marre, and W. Witzleb
den Subgruppen Philadelphia, Benidorm, Knoxville, France , Olda-He ysham und Bellingham, wahrend Umweltisolate wie die Bellingham, Oxford, France , Knoxville und Philadelph ia-Stamrne reagierten. Die Mehrzahl der Patienten isolate, (15 von 26, 58%) reagierte mit unserem mab 311 (er entspricht dem mab 2 des Standardpanels), aber nur 26 von 118 Urnweltstammen (22 %) aus 4 von 15 Warmwassersystemen taten dies (p < 0,05). Die meisten der Wasserisolate reagierten mit einem mab , der fur die Bellingham-Subgruppe spezifisch ist, Drei der unter suchten Wassersysteme waren mit Erkrankungen bei Menschen assoziiert. Zwei davon enth ielten den Subryp Bellingham, eins Stamrne, die wie Philadelphia reagierten.
Introduction Bacteria of the genus Legionella are ub iquitous water inhabita nts. They can be isolated from various man-made aqua tic systems such as hot water supplies in hospitals, hotels and other buildings, coo ling towers, whirlpools, show ers, dental units or from natural water environments. Infections in susceptible humans can occur following aeros olisation and transm ission of Legionella-contaminated water to the lungs. Legionella infections include pneumonia (legionnaires' disea se), influenza-like respirator y illness (Pontiac fever ) and asympto ma tic seroconversion (4) . Up to now , 37 different spe cies with at least 56 sero groups h ave been described. Legionella pneumopbila is still the mo st common spec ies. Using absorbed rabbit antisera, it can be divided int o 14 sero gro ups . Serogroup 1 (Lp 1) acco unts for th e majoriry o f clinical and envir onmental isolates (4, 17). Using monoclonal anti bo dies, th is sero gro up can be divided into 12 different antigenic subgro ups (12, 18, 20 ). Introduced by [oly et al. (12), the se subgroups we re named according to representative strains of Legionnella pneumopbila serog ro up 1. The distribution of different subgroups among clinical and environmental isolates in Euro pe w as studied in Gre at Britain (2 1), France (2) and Italy (1). In Germany, one study w as carried out in a Berlin ho spital (19). It ha s been the aim of the pr esent study to investigate for the first time the distribution of monoclonal an tibody su bgroups among clinical and environmental Lp 1 str ains iso lated in vario us geographical region s of Germany.
Materials and Methods
Legionella strains. Legionella strains were isolated on BCYE agar supplemented with alpha-ketoglutarate and antib ioti cs with or without acid pretreatment of samples (4, 8) from clinical and environmental specimens or supplied by colleagues (Tables 2 and 3). All strai ns reacted with FITC-Iabelled rabbit serum prepared against Lpl strain Philadelphia-I but not with absorbed sera to serog roups 2 to 14 (13). To determ ine th e reactiviry of our mon oclona l antibodies, 10 Lpl strains from the standard panel of Ioly et al. (12) were used (Table 1). Monoclonal antibodies . Monoclon al anti bodies from Dresden were selected from fusions of P3X630-Ag8/635 mouse myelom a cells with spleen cells from Balblc mice immun ized with form alin-killed Lpl strains : Philadelph ia-l for the production of mab 311, a mixture of Olda, Philadelphia-I , Knoxville-I , Bellingham-l for the production of mab 8/4 and mab 8/5) and Bellingham-1 for the production of mab 1011. Hybridomas were screened in an ELISA using sonicated Lp1 stra ins and in the IFA using heat (15 min; 100 °C) or formalinkilled (10% , 15 min) who le cell antigens. Briefly, ELISA microt iter plates were coated with sonicated Lp 1 strains (1 ug/rnl), blocked with Tween-Gela fusal-PBS, incubated with undiluted hybr idoma cell superna tants and incubated for 90 min at 37 °C. After carefull wash-
Legionella pneumophila Serogroup 1 Subtyping Table 1. Reactivity of monoclonal antibodies
to
181
Legionella pneumophila serogroup 1 type strains
Indirect immunofluorescence test with Mab Strain Philadelphia Allentown Benidorm Knoxville France OLDA Oxford Heysham Camperdown Bellingham
ATCCNr.
Mab 1"** Mab 2***
33152 43016 43108 33153 43112 43109 43110 43107 43113 43111
+ + + + + + + + + +
8/5*
3/1 *
+ + + + + 0 0 0 0 0
8/4*
33G3** 32A12** 144C2** 10/1* Mab 4*** Mab 5*** Mab 6*** Mab 7***
+
0 0 0 0 0 0 0 0 0
0 0 v****
0
+ + + 0 0
+
+ + + 0 0 0 0 0 0 0
+ 0 0
+ 0
+ + + 0 0
0 0
+ 0 0
+ 0
+ 0
+
- produced in Dresden -* kindly supplied by J.Joly, Quebeck, Canada according to the international standard panel variable results
---
ing, bound monoclonal antibodies were detected using peroxidase-conjugated rabbit anti mouse immunoglobulin (90 min; 3rC) and a-phenylenediamine as chromogen. The IFA was carried out as described below. Clones were obtained by the dilution method. Four of them were chosen for further studies. Immunoglobulin isotypes were determined in the radial immunodiffusion test using antisera from Sigma (Miinchen, Germany). Further characterisation of our mabs was done by Western blot studies with purified outer membranes. Briefly, SDS-PAGE of sonicated Legionella cells treated with N-Iauroylsarcosine (Serva, Heidelberg, Germany), was performed using a vertical electrophoresis system (GE 2/4, LKB, Freiburg, Germany). Separated Legionella antigens were transferred to nitrocellulose (0.45 urn, Bio-Rad, Munich, Germany) using a LKB 2005 Transphor unit (Pharmacia, Freiburg, Germany) (5). Nitrocellulose sheets were blocked with 10% BSA-PBS and incubated with 10 ml of undiluted cell culture supernatant of hybridoma cells for 60 min at 37°C. Following intensive washing in PBS-Tween, bound mabs were detected by POD-conjugated anti-mouse immunoglobulin (60 min; 37°C) and diaminobenzidine hydrochloride (Chemapol, Prag, CSFR) (9). Mabs 33G3, 32A12 and 144C2 were kindly supplied by JeanJoly, University of Quebec, Canada. These antibodies were named mab 5, mab 6 and mab 7 of the standard panel (12), respectively. Mab 4/4 produced after fusion of mouse myeloma cells and spleen cells of Balbi c mice immunized with 1. pneumophila serogroup 6 strain Chicago-2 as described above and reacting with lipopolysaccharide epitopes of 1. pneumophila serogroups 2-6, 8-10, 12-14 strains (n = 208) served as a negative control. Subtyping of Legionella pneumophila serogroup 1 strains. The indirect immunofluorescence test (IFA) was performed with Legionella cells scraped from BCYE-agar and fixed in 10% formalin for 10 min or in 1% formalin overnight. Antigens were further diluted in 0.5% normal yolk-sac buffer according to Wilkinson et al. (22). The bacterial suspensions (10 ul) were ploud on Teflon-coated microscope slides and air-dried at room temperature. 10 ul of undiluted hybridoma cell culture supernatant was added to each well and slides were incubated in a moist chamber for 45 min at 3rc. Next, they were washed with PBS and FITC-conjugated goat anti-mouse immunoglobulin (Staatliches Institut fur Immunpraparate und Nahrmedien, Berlin) was added at a dilution of 1 : 30. Following additional
P. C. Liick, J. H. Helbig, W. Ehret, R. Marre, and W. Witzleb
182
incubation for 30 min at 3rC, the slides were washed, airdried and examined with an epilight fluorescence microscope at a magnification of 400. All tests were done at least twice and read by two persons independently. Only bright (3+ ) or good (2+) fluorescence of Legionella isolates were considered positive. Dot-Blot-assay. To compare the IFA with a dot blot assay, a few strains were tested in both tests. Briefly, 20 ul of Legionella IFA antigens were spotted on nitrocellulose filters (0.45 urn, Schleicher and Schuell, Dassel, Germany) in a dot blot chamber (von Keutz, Reisingen, Germany), dried overnight at room temperature and incubated with 20 111 of monoclonal antibod y. Filters were incubated for 60 min at room temperature, carefully washed and horseradish-peroxidase-conjugated goat anti-mouse immunoglobulin was added for 60 min. After additional washing steps of 3 X 20 min, a colour reaction developed by using 0.005 % HzO z in diaminobenzidine-hydrochlorid (Chemapol, Prague, CSFR) (10 mg in 20 ml 0.05 M Tris-HCI, pH = 7.6) (9).
Results
Isorype determination using the Ouchterlony technique showed that mabs 311, 8/4, 10/1 and 8/5 were IgG 3 , IgG 3 , IgG3 and IgM, respectively. Results of immunoblots with selected strains of the standard panel revealed a ladder-like reaction pattern (data not shown). So it can be concluded that our mab s recognize lipopolysaccharide structures of the cell wall of Legione/la strains. A comparison of IFA and dot-blot assa y showed complete agreement of both tests for 23 arbitarily selected Legione/la strains isolated in Germany and the 10 type strains listed in Table 1. The results of a typical dot-blot assay with 14 strains are shown in Fig. 1. When comparing the reacti vity of our mabs with the 10 type-strains of the standard panel reported by jol» et a!' (12), it become s obvious that mab s 8/5, 311 and 10/1
Lpl strains Mabs 8/5 3/1
10/1 33 G3
Dr Mil Ro Ha Bh Ca He Ox 01 Fr Kn Bd AI Ph
•
•••
•
I I I
• • ••
•
32A12 144C2
8/4
•
Fig. 1. Reactivity of monoclonal antibodies with Legionella pneumophila serogroup 1 strains in the dot blot assay (Dr = Dresden 123/89; Mii = Miinchen 1; Ro = Restock/ Dresden 1180/89; Ha = Hamburg 1156; Bh = Bellingham; Ca = Camperdown; He = Heysham; Ox = Oxford ; 01 = Olda; Fr = France; Kn = Knoxville; Bd = Benidorm; Al = Allentown; Ph = Philadelphia)
Legionella pneumophila Serogroup 1 Subtyping
183
correspond to mab 1, mab 2 and mab 4 respectively, of this panel. Mab 814 reacted like mab 6 of the standard panel in the IFA and in the dot-blot assay with the type strains. Nevertheless, it was included in this study because the intensity of the fluorescence was sometimes different when using these rwo mabs. Using 7 mabs and 10 type strains, we were able to define 9 subgroups of Legionella pneumophila serogroup 1. Subgroup aida and Heysham could not be differentiated due to the lack of mab 3 of the standard panel or another discriminating antibody. 7 out of the known 9 subtypes of Lp I-strains were found among clinical and environmental strains isolated in Germany (Tables 2 and 3). Subtypes Allentown and Camperdown were not found at all. Clinical isolates belonged to the Philadelphia, Benidorm, Knoxville, France, Olda-Heysham and Bellingham subgroups. Philadelphia, Benidorm and Bellingham being the most common among them. The distribution of subtypes among environmental strains was different. They were subtyped as Bellingham, Oxford, France, Philadelphia and Knoxville-like strains. Of the 118 environmental isolates, 86 (73%) belonged to the Bellingham subgroup. Legionellae of subtypes Benidorm and Olda-Heysham were only isolated from patients, whereas Oxford-like
Table 2. Reaction pattern of monoclonal antibodies to Legionella pneumophila serogroup 1 strains isolated from patients from various regions of Germany in the indirect immunofluorescence test Strain
Isolated in
Reaction pattern
Mii1 Mii2 Mii3 Mii4 Mii5 Mii7 Mii8 Berlin 1 Ulm2 Kassel 1 Augsburg 1 Ingolstadt 1 S-594* S-9596 S-686'} S-734* S-763 28191 1156 L 11 686/88 123/89 281/89 1180/89 387/91 533/91
Munich Munich Munich Munich Munich Munich Munich Berlin Ulm Kassel Augsburg Ingolstadt Liibeck Liibeck Liibeck Liibeck Itzehoe Hamburg Hamburg Halberstadt Dresden Dresden Dresden Dresden (Restock)' * Dresden Dresden
Benidorm-like Bellingham-like Benidorm-like Philadelphia-like Philadelphia-like Benidorm-like Bellingham-like Benidorm-like Olda-Heysham-like Knoxville-like Benidorm-like Olda-Heysham-like Bellingham-like Bellingham-like Bellingham-like Bennidorm-like o lda-Heysham-like Philadelphia-like Philadelphia-like Bellingham-like Knoxville-like Bellingham-like Philadelphia-like Philadelphia-like Bellingham-like France-like
* isolated during an outbreak of nosocomial legionellosis *" isolated in Dresden from a patient hospitalized in Rostock
184
P. C. Luck,
J. H. Helbig,
W. Ehret, R. Marre, and W. Witzleb
Table 3. Reaction pattern of monoclonal antibodies to environmental Legionella pneumophila serogroup 1 strains isolated in various regions of Germany in the indirect immunofluorescence test Water system
Number of water sample s tested
Number of individual colonies tested
Reaction pattern
Hospital water supply Munich GH
5
5
Bellingham-like
Hospital water supply Liibeck"
2
2
Bellingham-like
Hospital water supply Hamburg ED
3
3
Philadelphia-like (1) Knoxville-like (2)
Hospital water supply Halberstadt *
7
8
Bellingham-like
Hospital water supply Dresden F1
2
5
Bellingham-like Bellingham -like
Hospital water supply Dresden Fe Hospital water supply Dresden FH
5
10
Bellingham-like
Hospital water suppl y Dresden DH
4
12
Bellingham-like
Hospital water supply Dresden N S
2
6
Oxford-like
Water supply outpatient Department Bautzen
2
4
France-like
Hospital water supply Radebeul NS
5
10
Bellingham -like
Dental unit Dresden PO
2
5
Bellingham-like
Dental unit Dresden AB
1
1
Philadelphia-like
Water supply in pol ice barracks Rud olsradt " *
6
18
Philadelphia-like
Municipal water suppl y Dresden GU
4
28
Bellingham-l ike
'f
associated with nosocomial legionellosis
** associated with a cluster of community-acquired legionellos is
strains were not detected among clini cal isolates, but isolated from water. Still un described reaction patterns were not found. In this limited survey of L. pneumopbila sero gro up 1 strai ns isolated in Germany, we did not find an y clustering.
Legionella pneumophila Serogroup 1 Subtyping
185
Discussion A comparison of IFA and dot blot assay revealed a complete agreement of both tests. Therefore, the two tests can be used for subtyping Lp1 strains. IFA is easier to perform if only a few strains are to be tested, but a larger number of strains or individual colonies can be tested in a shorter period of time of in the dot-blot assay. It is another advantage of the dot-blot assay that the nitrocellulose membranes can be stored for a long time without any effect on the results. The subtypes were named by Joly et al. (12) according to a representative strain of the respective subgroups (Table 1). With the exception of mab 3 from this panel, we used the same reaction patterns for typing clinical and environmental Legionella pneumophila serogroup 1 strains isolated at different locations in Germany. We found 7 antigenic subtypes (Tables 2 and 3). Subtypes Camperdown and Allentown were not found in our study. This corresponds to the findings of Ruf et al. (19) in Berlin. Few subtypes were absent among patient or environmental strains. Because the number of strains tested by us was limited we may have missed these subtypes. Contrary to our results, they were found in a high percentage of isolated L. pneumophila serogroup 1 strains in Italy (1). 15 out of 26 clinical isolates from Germany (58%) reacted with mab 3/1, which detects Lp1 strains of the following subgroups: Philadelphia, Knoxville, Pontiac, Benidorm, France and Allentown. Mab 3/1 is comparable to mab 2 of the international standard panel. In several studies conducted in various regions of the world, it was shown that L. pneumophila strains isolated from patients carried an epitope which reacted with mab 2 of the standard panel and may have been virulence-associated (2, 6, 20, 21). Our findings are in accordance with this observation. We found that strains reacting with mab 3/1 (corresponding to mab 2) were significantly more frequent among clinical isolates. On the other hand, it should be noted that Legionella isolates from 8 patients (31%) were positive with mab 10/1, being specific to Bellingham-like strains and 2 (8%) reacted like Olda-Heysham strains. When summarizing these results, it may be suggested that virulence is associated with one epitope on the lipopolysaccharide of the Legionella outer membrane recognized by mab 3/1 (or mab 2 of the standard panel), but also with other as yet unknown factors. Contrary to this, 10 out of 15 (66%) hot water supplies contained Bellingham-like strains. In England and Italy, aida, Heysham and Oxford-like strains account for the majority of environmental isolates (1, 21). It is noteworthy that in our study, two of three water systems associated with human disease harboured Bellingham subtype strains. In a nosocomial outbreak caused by a water-borne Bellingham-like strain, two of three cases had a lethal outcome (17). Ruf et al. (19) also found the Bellingham subtype to have been responsible for two of 12 nosocomial Legionella associated pneumonia cases in a Berlin hospital. With the exception of one, only one subtype was found in all hot water systems studied. This is in contrast to other reports in which two, three or even four subtypes were isolated from hot water supplies (3, 11, 18, 19,20). This discrepancy may be in part attributed to the fact that the number of individual colonies tested from each water sample was too low. It seems to be possible, that we have missed other subtypes. Further studies should pay attention to this. On the other hand, it may be that indeed only one subtype was present in each of the water supplies. So Watkins et al. (21) found that in 85% of buildings where Lpl strains had been isolated from several sources, only one subtype was found. Possible phenotypic variations among Lp1 strains (11) also may have influenced these results, but further investigations are needed to elucidate this.
186
P. C. Luck,
J. H. Helbig, W. Ehret,
R. Marre, and W. Witzleb
Out of 2 6 = 64 possible combinations of 6 discriminating mabs, 12 have been found up to now (12,18,20). For clinical microbiological laboratories this ma y be useful and convenient. In many studies, ep idemiological linkage between environmental sites and infections in humans were made on the basis of monoclonal antibody subtyping. This wa s helpful in understanding the epidemiology of human legionellosis (3, 7, 10, 14, 15, 16, 18, 19,20). When further su bty p ing is required it might be useful to supplement the monoclonal subryping system with other subtyping methods, such as plasmid analysis, alloenzyme pattern, restriction fragment length pattern or pulsed field gel electrophoresis of cleaved genomic DNA (3 , 11, 14 , 17, 18, 20 ).
Acknowledgement. This study was supported by an FEMS fellowship to P.c.L. We are grateful to]. Joly, University of Quebec, Canada for supplying us with mabs 33G3, 32A12 and 144C2 and to the following colleagues for sending us Legionella strains : Dr. Elke Dinger, Wernigerode, Prof. Dr. F. Fehrenbach, Berlin, Dr. K. Zobel, Hamburg, and Dr. Seewald, Kassel. We would like to thank ]utta Moller and Sigrid Giibler for excellent technical assistance.
References 1. Castelani Pastoris, M., M. Mclntyre, and P. Goldoni: Legionella pneumophila serogroup 1 population in Ital y by monoclonal subtyping. Epidem. Infect. 105 (1990) 169-1 74 2. Dournon, E., W. F. Bibb, P. Rajagopalan, N. Deplaces, and R. M. McKinney: Monoclonal antibody reactivity as a virulence marker for Legionella pneumophila serogroup 1 strains. ]. infect. Dis. 157 (1988) 496-501 3. Edelstein, P. H., C. Nakahama, J. O 'H. Tobin, K. Calarco, K. B. Beer, J. R. Joly, and R. K. Selander: Paleoepidemiologic investigation of Legionnaires' disease at Wadworth Veterans Administraton Hospital by using three typing methods for comparison of legionellae from clinical and environmental sources. ]. Clin. Microbiol. 23 (1986) 1121-1 126 4. Edelstein, P. H.: Laboratory diagno sis of infection cau sed by legionellae . Eur. J. Clin . Microbiol. 6 (1987) 4-10 5. Ehret, W. and G. Ruckdeschel: Species specific membrane proteins of Legionellaceae. Zbl. Bakt. Hyg. A 255 (1983 ) 33-38 6. Ehret, W ., B. U. von Specht, and G. Ruckdeschel: Discrimination between clinical and environmental strains of Legionella pneumopbila by a monoclonal antibody. Israel J. Med . Sci. 22 (1986) 715- 723 7. Garbe, P. L, B. ]. Davis, ]. S. Weisfeld, L. Markowitz, P. Miner, F. Garrity, J. M. Barbaree, and A.L. Reingold: Nosocomial Legionnaires' disease - Epidemiologic demonstration of cooling towers as a source.]. Am. Med. Ass. 254 (1985 ) 521-524 8. Go rman, G. W., [. M. Barbaree, and ]. C. Feeley: Procedures for the recovery of Legionella from water. Developmental manual. U.S. Dept . Health and Human Service, Cente r for Disease Control, Atlant alUSA (1983) 9. Graham, R. C. and R. J. Karnousky : The early stages of absorption of injected horseradish pero xida se in the prox imal tubules of mouse kidney : ultrastructural cytochemistry by a new technique.]. Histoch em. Cytochern. 14 (196 6) 291-302 10. Hand rahan, J. P., D. L. Morse, V. B. Scharf, J. G. Debbie, G. P. Schmid, R. M. Mckinney, and M. Shayegani: A community hospital outbreak of legionellosis. Am. ]. Epidem. 125 (1987 ) 639-649 11. Harrison, T. G., N. A. Saunders, A . Hatbthotuuia, G. Hallas, R. J. Birtles, and A. G. Taylor: Phenotypic variation amongst genotypically homogenous Legionella pneumophila serogroup I isolate s: implications for the investigations of outbreaks of Legionna ires' disease. Epidem . Infect. 104 (1990) 171-180
Legionella pneumophila Serogroup 1 Subtyping
187
12. j oly, J. R., R. M. McKinney, j . O 'H. To bin, W. F. Bibb, I. D. Watkins, and D. Ramsa y: Development of a standardized subgrouping scheme for Legionella pneum opbila serogroup 1 using monoclonal ant ibodies. J. Clin. Microbiol . 23 (1986) 768-771 13. Luck , P. Ch., J. H. Helbig, Ch. Pilz und W. Wit zleb : Nachweis von Legionellen in klinischen Untersuchungsmaterialien mit FITC-mark ierten Antikorpern, Z. ges. Hyg. 35 (1989) 601- 604 14. Maher, W. E., M. F. Para, and]. F. Plouffe: Subryping of Legionel/a pneumophila serogroup 1 isolates by monoclonal antibody and plasmid techniques. ] . Clin. MicrobioI. 25 (1987) 2281-2284 15. Mitchel, E., M . O 'H. Mahony, j . M . Watson, D. Lynch, C. j oseph, C. Quigley, R. Aston, G. N. Constable, J. R. Farrand, S. Maxwell, D. N. Hut chinson, and]. Craske: Two outbreaks of Legionn aires' disease in Bolton Health District. Epidem. Infect. 104 (1990) 159-1 70 16. Ne ill, M. A., G. W. Gorman , C. Gibert, A. Roussel, A. W. Hightower, R. M. McKinney, and C. V. Broome: Noso comial legionellosis, Paris, France - Evidence of transmission by potable water. Am. J. Med. 78 (1984) 581-588 17. O tt, M., L. Bender, R. Marre, and j. Hacker: Pulsed field electrophoresis of genomic restriction fragments for the detection of nosocomial Legionella pneum opbila in hospital water supply. J. Clin. Microbiol. 29 (1991) 813-815 18. Pfaller, M., R. Hollis, W. j ohnson, R. M. Massanari, C. Helms, R. Wenzel, N. Hall, N. Moyer, and ]. joly: The application of molecular and immunologic techniques to study the epidemiolgy of Legionella pneum ophila serogroup 1. Diagn. Microbiol. Infect. Dis. 12 (1989) 295-302 19. Ruf, B., D. Schiirmann, I. Horbach, K. Seidel, and H. D. Pohle: Nosocomial Legionella pneum ophila: demon stration of potable water as the source of infection. Epidem. Infect. 101 (1988) 647-654 20. Stout, J. E., J. joly, M. Para, J. Plou ffe, C. Ciesielski, M. J. Blaser, and V. L. Yu : Comparison of molecular methods for subtyping patients and epidemiologically linked environmental isolates of Legionella pneum oph ila. J. Infect. Dis. 157 (1988) 486--495 21. Watkins, I. D., J. O'H. Tobin, P. J. Dennis, W. Brown, R. Newnham, and ]. B. Kurtz: Legionella pneum oph ila serogroup 1 subgrouping by monoclonal antibodies - an epidemiological tool.]. Hyg. (Camb.) 95 (1985) 211-21 6 22. Wilk inson, H. W.: Hospital-Laboratory diagnosis of Legionella-infection s. Center for Disease Control, AtiamalUSA (1988) 23. Winn, W. c.: Legionn aire Disease: Historical perspective. Clin. Microbiol. Rev. 1 (19 88 ) 60-8 1
Dr. Paul Christian Luck, Institut fiir Medizinische Mikrobiologie, Medizinische Akademie Dresden, Giintzstr . 32, 0 -8019 Dresden, Germany
Subtyping of Legionella pneumophila Serogroup 1 Strains Isolated in Germany Using Monoclonal Antibodies PAUL CHRISTIAN LOCK l , JORGEN HERBERT HELBIG!, WERNER EHRET 2 , REINHARD MARRE 3 , and WOLF WITZLEB l 1
2
3
Institut fiir Medizinische Mikrobiologie, Medizinische Akademie Dresden, 0-8019 Dresden Max-von-Pettenkofer-Institut fur Hygiene und Mikrobiologie. Ludwig-Maximilian-Universitat Munchen, W-8000 Munchen Institut fur Medizinische Mikrobiologie, Medizinische Universitat Lubeck, W-2400 Lubeck, Germany
With 1 Figure· Received August 16, 1991 . Revision received January 27, 1992· Accepted March 10, 1992
Summary Legionella pneumopbila serogroup 1 (Lpl ) strains isolated from patients and hot water supplies in different locations of Germany were subtyped using seven monoclonal antibodies (mabs) in the indirect immunofluorescence test (IFA) and in part, using a dot blot assay. Four of these mabs were produced in Dresden. Three mabs (mab 33G3, mab 32A12 and mab 144c2) were kindly supplied by]. ]oly, Quebec, Canada. Altogether, seven antigenic variants were found among Lpl strains isolated in Germany. Patient strains belonged to the Philadelphia, Benidorm, Knoxville, France, Olda-Heysham and Bellingham subgroups, whereas environmental isolates reacted like the Bellingham, Oxford, Philadelphia, Knoxville and France strains. The majority of patient strains (15 out of 26, 58 %) reacted with our mab 3/1 (corresponding to mab 2 of the standard panel), but only 26 out of 118 environmental strains (22% ) isolated from 4 of 15 hot water supplies did so (p < 0.05). The majority of water-borne Lp1 strains reacted with a mab specific of the Bellingham subgroup. Three water systems under study were associated with human legionellosis. Two of them contained Bellingham-like strains, one Philadelphia-like legionellae.
Zusammenfassung Legionella pneumophila Serogrouppe 1 (Lp l ) Stamrne, die von Patienten und aus Wassersystemen in verschiedenen geographischen Regionen Deutschlands isoliert worden waren, wurden mit sieben monoklonalen Antikorpern (mab) im indirekten Immunofloureszenztest (lFA) und teilweise in einem Dot-Blot-Assay subtypisiert. Vier der verwendeten mab wurden in Dresden produziert. Drei (mab 33G3, mab 32A12 und mab 144C2) wurden freundlicherweise von j. ]oly, Quebec, Kanada, uberlassen. Insgesamt wurden sieben antigene Varianten unter den Isolaten aus Deutschland gefunden. Patientenstamme gehorten zu
180
P. C. Luck,
J. H. Helb ig, W. Ehret, R. Marre, and W. Witzleb
den Subgruppen Philadelphia, Benidorm, Knoxville, France , Olda-He ysham und Bellingham, wahrend Umweltisolate wie die Bellingham, Oxford, France , Knoxville und Philadelph ia-Stamrne reagierten. Die Mehrzahl der Patienten isolate, (15 von 26, 58%) reagierte mit unserem mab 311 (er entspricht dem mab 2 des Standardpanels), aber nur 26 von 118 Urnweltstammen (22 %) aus 4 von 15 Warmwassersystemen taten dies (p < 0,05). Die meisten der Wasserisolate reagierten mit einem mab , der fur die Bellingham-Subgruppe spezifisch ist, Drei der unter suchten Wassersysteme waren mit Erkrankungen bei Menschen assoziiert. Zwei davon enth ielten den Subryp Bellingham, eins Stamrne, die wie Philadelphia reagierten.
Introduction Bacteria of the genus Legionella are ub iquitous water inhabita nts. They can be isolated from various man-made aqua tic systems such as hot water supplies in hospitals, hotels and other buildings, coo ling towers, whirlpools, show ers, dental units or from natural water environments. Infections in susceptible humans can occur following aeros olisation and transm ission of Legionella-contaminated water to the lungs. Legionella infections include pneumonia (legionnaires' disea se), influenza-like respirator y illness (Pontiac fever ) and asympto ma tic seroconversion (4) . Up to now , 37 different spe cies with at least 56 sero groups h ave been described. Legionella pneumopbila is still the mo st common spec ies. Using absorbed rabbit antisera, it can be divided int o 14 sero gro ups . Serogroup 1 (Lp 1) acco unts for th e majoriry o f clinical and envir onmental isolates (4, 17). Using monoclonal anti bo dies, th is sero gro up can be divided into 12 different antigenic subgro ups (12, 18, 20 ). Introduced by [oly et al. (12), the se subgroups we re named according to representative strains of Legionnella pneumopbila serog ro up 1. The distribution of different subgroups among clinical and environmental isolates in Euro pe w as studied in Gre at Britain (2 1), France (2) and Italy (1). In Germany, one study w as carried out in a Berlin ho spital (19). It ha s been the aim of the pr esent study to investigate for the first time the distribution of monoclonal an tibody su bgroups among clinical and environmental Lp 1 str ains iso lated in vario us geographical region s of Germany.
Materials and Methods
Legionella strains. Legionella strains were isolated on BCYE agar supplemented with alpha-ketoglutarate and antib ioti cs with or without acid pretreatment of samples (4, 8) from clinical and environmental specimens or supplied by colleagues (Tables 2 and 3). All strai ns reacted with FITC-Iabelled rabbit serum prepared against Lpl strain Philadelphia-I but not with absorbed sera to serog roups 2 to 14 (13). To determ ine th e reactiviry of our mon oclona l antibodies, 10 Lpl strains from the standard panel of Ioly et al. (12) were used (Table 1). Monoclonal antibodies . Monoclon al anti bodies from Dresden were selected from fusions of P3X630-Ag8/635 mouse myelom a cells with spleen cells from Balblc mice immun ized with form alin-killed Lpl strains : Philadelph ia-l for the production of mab 311, a mixture of Olda, Philadelphia-I , Knoxville-I , Bellingham-l for the production of mab 8/4 and mab 8/5) and Bellingham-1 for the production of mab 1011. Hybridomas were screened in an ELISA using sonicated Lp1 stra ins and in the IFA using heat (15 min; 100 °C) or formalinkilled (10% , 15 min) who le cell antigens. Briefly, ELISA microt iter plates were coated with sonicated Lp 1 strains (1 ug/rnl), blocked with Tween-Gela fusal-PBS, incubated with undiluted hybr idoma cell superna tants and incubated for 90 min at 37 °C. After carefull wash-
Legionella pneumophila Serogroup 1 Subtyping Table 1. Reactivity of monoclonal antibodies
to
181
Legionella pneumophila serogroup 1 type strains
Indirect immunofluorescence test with Mab Strain Philadelphia Allentown Benidorm Knoxville France OLDA Oxford Heysham Camperdown Bellingham
ATCCNr.
Mab 1"** Mab 2***
33152 43016 43108 33153 43112 43109 43110 43107 43113 43111
+ + + + + + + + + +
8/5*
3/1 *
+ + + + + 0 0 0 0 0
8/4*
33G3** 32A12** 144C2** 10/1* Mab 4*** Mab 5*** Mab 6*** Mab 7***
+
0 0 0 0 0 0 0 0 0
0 0 v****
0
+ + + 0 0
+
+ + + 0 0 0 0 0 0 0
+ 0 0
+ 0
+ + + 0 0
0 0
+ 0 0
+ 0
+ 0
+
- produced in Dresden -* kindly supplied by J.Joly, Quebeck, Canada according to the international standard panel variable results
---
ing, bound monoclonal antibodies were detected using peroxidase-conjugated rabbit anti mouse immunoglobulin (90 min; 3rC) and a-phenylenediamine as chromogen. The IFA was carried out as described below. Clones were obtained by the dilution method. Four of them were chosen for further studies. Immunoglobulin isotypes were determined in the radial immunodiffusion test using antisera from Sigma (Miinchen, Germany). Further characterisation of our mabs was done by Western blot studies with purified outer membranes. Briefly, SDS-PAGE of sonicated Legionella cells treated with N-Iauroylsarcosine (Serva, Heidelberg, Germany), was performed using a vertical electrophoresis system (GE 2/4, LKB, Freiburg, Germany). Separated Legionella antigens were transferred to nitrocellulose (0.45 urn, Bio-Rad, Munich, Germany) using a LKB 2005 Transphor unit (Pharmacia, Freiburg, Germany) (5). Nitrocellulose sheets were blocked with 10% BSA-PBS and incubated with 10 ml of undiluted cell culture supernatant of hybridoma cells for 60 min at 37°C. Following intensive washing in PBS-Tween, bound mabs were detected by POD-conjugated anti-mouse immunoglobulin (60 min; 37°C) and diaminobenzidine hydrochloride (Chemapol, Prag, CSFR) (9). Mabs 33G3, 32A12 and 144C2 were kindly supplied by JeanJoly, University of Quebec, Canada. These antibodies were named mab 5, mab 6 and mab 7 of the standard panel (12), respectively. Mab 4/4 produced after fusion of mouse myeloma cells and spleen cells of Balbi c mice immunized with 1. pneumophila serogroup 6 strain Chicago-2 as described above and reacting with lipopolysaccharide epitopes of 1. pneumophila serogroups 2-6, 8-10, 12-14 strains (n = 208) served as a negative control. Subtyping of Legionella pneumophila serogroup 1 strains. The indirect immunofluorescence test (IFA) was performed with Legionella cells scraped from BCYE-agar and fixed in 10% formalin for 10 min or in 1% formalin overnight. Antigens were further diluted in 0.5% normal yolk-sac buffer according to Wilkinson et al. (22). The bacterial suspensions (10 ul) were ploud on Teflon-coated microscope slides and air-dried at room temperature. 10 ul of undiluted hybridoma cell culture supernatant was added to each well and slides were incubated in a moist chamber for 45 min at 3rc. Next, they were washed with PBS and FITC-conjugated goat anti-mouse immunoglobulin (Staatliches Institut fur Immunpraparate und Nahrmedien, Berlin) was added at a dilution of 1 : 30. Following additional
P. C. Liick, J. H. Helbig, W. Ehret, R. Marre, and W. Witzleb
182
incubation for 30 min at 3rC, the slides were washed, airdried and examined with an epilight fluorescence microscope at a magnification of 400. All tests were done at least twice and read by two persons independently. Only bright (3+ ) or good (2+) fluorescence of Legionella isolates were considered positive. Dot-Blot-assay. To compare the IFA with a dot blot assay, a few strains were tested in both tests. Briefly, 20 ul of Legionella IFA antigens were spotted on nitrocellulose filters (0.45 urn, Schleicher and Schuell, Dassel, Germany) in a dot blot chamber (von Keutz, Reisingen, Germany), dried overnight at room temperature and incubated with 20 111 of monoclonal antibod y. Filters were incubated for 60 min at room temperature, carefully washed and horseradish-peroxidase-conjugated goat anti-mouse immunoglobulin was added for 60 min. After additional washing steps of 3 X 20 min, a colour reaction developed by using 0.005 % HzO z in diaminobenzidine-hydrochlorid (Chemapol, Prague, CSFR) (10 mg in 20 ml 0.05 M Tris-HCI, pH = 7.6) (9).
Results
Isorype determination using the Ouchterlony technique showed that mabs 311, 8/4, 10/1 and 8/5 were IgG 3 , IgG 3 , IgG3 and IgM, respectively. Results of immunoblots with selected strains of the standard panel revealed a ladder-like reaction pattern (data not shown). So it can be concluded that our mab s recognize lipopolysaccharide structures of the cell wall of Legione/la strains. A comparison of IFA and dot-blot assa y showed complete agreement of both tests for 23 arbitarily selected Legione/la strains isolated in Germany and the 10 type strains listed in Table 1. The results of a typical dot-blot assay with 14 strains are shown in Fig. 1. When comparing the reacti vity of our mabs with the 10 type-strains of the standard panel reported by jol» et a!' (12), it become s obvious that mab s 8/5, 311 and 10/1
Lpl strains Mabs 8/5 3/1
10/1 33 G3
Dr Mil Ro Ha Bh Ca He Ox 01 Fr Kn Bd AI Ph
•
•••
•
I I I
• • ••
•
32A12 144C2
8/4
•
Fig. 1. Reactivity of monoclonal antibodies with Legionella pneumophila serogroup 1 strains in the dot blot assay (Dr = Dresden 123/89; Mii = Miinchen 1; Ro = Restock/ Dresden 1180/89; Ha = Hamburg 1156; Bh = Bellingham; Ca = Camperdown; He = Heysham; Ox = Oxford ; 01 = Olda; Fr = France; Kn = Knoxville; Bd = Benidorm; Al = Allentown; Ph = Philadelphia)
Legionella pneumophila Serogroup 1 Subtyping
183
correspond to mab 1, mab 2 and mab 4 respectively, of this panel. Mab 814 reacted like mab 6 of the standard panel in the IFA and in the dot-blot assay with the type strains. Nevertheless, it was included in this study because the intensity of the fluorescence was sometimes different when using these rwo mabs. Using 7 mabs and 10 type strains, we were able to define 9 subgroups of Legionella pneumophila serogroup 1. Subgroup aida and Heysham could not be differentiated due to the lack of mab 3 of the standard panel or another discriminating antibody. 7 out of the known 9 subtypes of Lp I-strains were found among clinical and environmental strains isolated in Germany (Tables 2 and 3). Subtypes Allentown and Camperdown were not found at all. Clinical isolates belonged to the Philadelphia, Benidorm, Knoxville, France, Olda-Heysham and Bellingham subgroups. Philadelphia, Benidorm and Bellingham being the most common among them. The distribution of subtypes among environmental strains was different. They were subtyped as Bellingham, Oxford, France, Philadelphia and Knoxville-like strains. Of the 118 environmental isolates, 86 (73%) belonged to the Bellingham subgroup. Legionellae of subtypes Benidorm and Olda-Heysham were only isolated from patients, whereas Oxford-like
Table 2. Reaction pattern of monoclonal antibodies to Legionella pneumophila serogroup 1 strains isolated from patients from various regions of Germany in the indirect immunofluorescence test Strain
Isolated in
Reaction pattern
Mii1 Mii2 Mii3 Mii4 Mii5 Mii7 Mii8 Berlin 1 Ulm2 Kassel 1 Augsburg 1 Ingolstadt 1 S-594* S-9596 S-686'} S-734* S-763 28191 1156 L 11 686/88 123/89 281/89 1180/89 387/91 533/91
Munich Munich Munich Munich Munich Munich Munich Berlin Ulm Kassel Augsburg Ingolstadt Liibeck Liibeck Liibeck Liibeck Itzehoe Hamburg Hamburg Halberstadt Dresden Dresden Dresden Dresden (Restock)' * Dresden Dresden
Benidorm-like Bellingham-like Benidorm-like Philadelphia-like Philadelphia-like Benidorm-like Bellingham-like Benidorm-like Olda-Heysham-like Knoxville-like Benidorm-like Olda-Heysham-like Bellingham-like Bellingham-like Bellingham-like Bennidorm-like o lda-Heysham-like Philadelphia-like Philadelphia-like Bellingham-like Knoxville-like Bellingham-like Philadelphia-like Philadelphia-like Bellingham-like France-like
* isolated during an outbreak of nosocomial legionellosis *" isolated in Dresden from a patient hospitalized in Rostock
184
P. C. Luck,
J. H. Helbig,
W. Ehret, R. Marre, and W. Witzleb
Table 3. Reaction pattern of monoclonal antibodies to environmental Legionella pneumophila serogroup 1 strains isolated in various regions of Germany in the indirect immunofluorescence test Water system
Number of water sample s tested
Number of individual colonies tested
Reaction pattern
Hospital water supply Munich GH
5
5
Bellingham-like
Hospital water supply Liibeck"
2
2
Bellingham-like
Hospital water supply Hamburg ED
3
3
Philadelphia-like (1) Knoxville-like (2)
Hospital water supply Halberstadt *
7
8
Bellingham-like
Hospital water supply Dresden F1
2
5
Bellingham-like Bellingham -like
Hospital water supply Dresden Fe Hospital water supply Dresden FH
5
10
Bellingham-like
Hospital water suppl y Dresden DH
4
12
Bellingham-like
Hospital water supply Dresden N S
2
6
Oxford-like
Water supply outpatient Department Bautzen
2
4
France-like
Hospital water supply Radebeul NS
5
10
Bellingham -like
Dental unit Dresden PO
2
5
Bellingham-like
Dental unit Dresden AB
1
1
Philadelphia-like
Water supply in pol ice barracks Rud olsradt " *
6
18
Philadelphia-like
Municipal water suppl y Dresden GU
4
28
Bellingham-l ike
'f
associated with nosocomial legionellosis
** associated with a cluster of community-acquired legionellos is
strains were not detected among clini cal isolates, but isolated from water. Still un described reaction patterns were not found. In this limited survey of L. pneumopbila sero gro up 1 strai ns isolated in Germany, we did not find an y clustering.
Legionella pneumophila Serogroup 1 Subtyping
185
Discussion A comparison of IFA and dot blot assay revealed a complete agreement of both tests. Therefore, the two tests can be used for subtyping Lp1 strains. IFA is easier to perform if only a few strains are to be tested, but a larger number of strains or individual colonies can be tested in a shorter period of time of in the dot-blot assay. It is another advantage of the dot-blot assay that the nitrocellulose membranes can be stored for a long time without any effect on the results. The subtypes were named by Joly et al. (12) according to a representative strain of the respective subgroups (Table 1). With the exception of mab 3 from this panel, we used the same reaction patterns for typing clinical and environmental Legionella pneumophila serogroup 1 strains isolated at different locations in Germany. We found 7 antigenic subtypes (Tables 2 and 3). Subtypes Camperdown and Allentown were not found in our study. This corresponds to the findings of Ruf et al. (19) in Berlin. Few subtypes were absent among patient or environmental strains. Because the number of strains tested by us was limited we may have missed these subtypes. Contrary to our results, they were found in a high percentage of isolated L. pneumophila serogroup 1 strains in Italy (1). 15 out of 26 clinical isolates from Germany (58%) reacted with mab 3/1, which detects Lp1 strains of the following subgroups: Philadelphia, Knoxville, Pontiac, Benidorm, France and Allentown. Mab 3/1 is comparable to mab 2 of the international standard panel. In several studies conducted in various regions of the world, it was shown that L. pneumophila strains isolated from patients carried an epitope which reacted with mab 2 of the standard panel and may have been virulence-associated (2, 6, 20, 21). Our findings are in accordance with this observation. We found that strains reacting with mab 3/1 (corresponding to mab 2) were significantly more frequent among clinical isolates. On the other hand, it should be noted that Legionella isolates from 8 patients (31%) were positive with mab 10/1, being specific to Bellingham-like strains and 2 (8%) reacted like Olda-Heysham strains. When summarizing these results, it may be suggested that virulence is associated with one epitope on the lipopolysaccharide of the Legionella outer membrane recognized by mab 3/1 (or mab 2 of the standard panel), but also with other as yet unknown factors. Contrary to this, 10 out of 15 (66%) hot water supplies contained Bellingham-like strains. In England and Italy, aida, Heysham and Oxford-like strains account for the majority of environmental isolates (1, 21). It is noteworthy that in our study, two of three water systems associated with human disease harboured Bellingham subtype strains. In a nosocomial outbreak caused by a water-borne Bellingham-like strain, two of three cases had a lethal outcome (17). Ruf et al. (19) also found the Bellingham subtype to have been responsible for two of 12 nosocomial Legionella associated pneumonia cases in a Berlin hospital. With the exception of one, only one subtype was found in all hot water systems studied. This is in contrast to other reports in which two, three or even four subtypes were isolated from hot water supplies (3, 11, 18, 19,20). This discrepancy may be in part attributed to the fact that the number of individual colonies tested from each water sample was too low. It seems to be possible, that we have missed other subtypes. Further studies should pay attention to this. On the other hand, it may be that indeed only one subtype was present in each of the water supplies. So Watkins et al. (21) found that in 85% of buildings where Lpl strains had been isolated from several sources, only one subtype was found. Possible phenotypic variations among Lp1 strains (11) also may have influenced these results, but further investigations are needed to elucidate this.
186
P. C. Luck,
J. H. Helbig, W. Ehret,
R. Marre, and W. Witzleb
Out of 2 6 = 64 possible combinations of 6 discriminating mabs, 12 have been found up to now (12,18,20). For clinical microbiological laboratories this ma y be useful and convenient. In many studies, ep idemiological linkage between environmental sites and infections in humans were made on the basis of monoclonal antibody subtyping. This wa s helpful in understanding the epidemiology of human legionellosis (3, 7, 10, 14, 15, 16, 18, 19,20). When further su bty p ing is required it might be useful to supplement the monoclonal subryping system with other subtyping methods, such as plasmid analysis, alloenzyme pattern, restriction fragment length pattern or pulsed field gel electrophoresis of cleaved genomic DNA (3 , 11, 14 , 17, 18, 20 ).
Acknowledgement. This study was supported by an FEMS fellowship to P.c.L. We are grateful to]. Joly, University of Quebec, Canada for supplying us with mabs 33G3, 32A12 and 144C2 and to the following colleagues for sending us Legionella strains : Dr. Elke Dinger, Wernigerode, Prof. Dr. F. Fehrenbach, Berlin, Dr. K. Zobel, Hamburg, and Dr. Seewald, Kassel. We would like to thank ]utta Moller and Sigrid Giibler for excellent technical assistance.
References 1. Castelani Pastoris, M., M. Mclntyre, and P. Goldoni: Legionella pneumophila serogroup 1 population in Ital y by monoclonal subtyping. Epidem. Infect. 105 (1990) 169-1 74 2. Dournon, E., W. F. Bibb, P. Rajagopalan, N. Deplaces, and R. M. McKinney: Monoclonal antibody reactivity as a virulence marker for Legionella pneumophila serogroup 1 strains. ]. infect. Dis. 157 (1988) 496-501 3. Edelstein, P. H., C. Nakahama, J. O 'H. Tobin, K. Calarco, K. B. Beer, J. R. Joly, and R. K. Selander: Paleoepidemiologic investigation of Legionnaires' disease at Wadworth Veterans Administraton Hospital by using three typing methods for comparison of legionellae from clinical and environmental sources. ]. Clin. Microbiol. 23 (1986) 1121-1 126 4. Edelstein, P. H.: Laboratory diagno sis of infection cau sed by legionellae . Eur. J. Clin . Microbiol. 6 (1987) 4-10 5. Ehret, W. and G. Ruckdeschel: Species specific membrane proteins of Legionellaceae. Zbl. Bakt. Hyg. A 255 (1983 ) 33-38 6. Ehret, W ., B. U. von Specht, and G. Ruckdeschel: Discrimination between clinical and environmental strains of Legionella pneumopbila by a monoclonal antibody. Israel J. Med . Sci. 22 (1986) 715- 723 7. Garbe, P. L, B. ]. Davis, ]. S. Weisfeld, L. Markowitz, P. Miner, F. Garrity, J. M. Barbaree, and A.L. Reingold: Nosocomial Legionnaires' disease - Epidemiologic demonstration of cooling towers as a source.]. Am. Med. Ass. 254 (1985 ) 521-524 8. Go rman, G. W., [. M. Barbaree, and ]. C. Feeley: Procedures for the recovery of Legionella from water. Developmental manual. U.S. Dept . Health and Human Service, Cente r for Disease Control, Atlant alUSA (1983) 9. Graham, R. C. and R. J. Karnousky : The early stages of absorption of injected horseradish pero xida se in the prox imal tubules of mouse kidney : ultrastructural cytochemistry by a new technique.]. Histoch em. Cytochern. 14 (196 6) 291-302 10. Hand rahan, J. P., D. L. Morse, V. B. Scharf, J. G. Debbie, G. P. Schmid, R. M. Mckinney, and M. Shayegani: A community hospital outbreak of legionellosis. Am. ]. Epidem. 125 (1987 ) 639-649 11. Harrison, T. G., N. A. Saunders, A . Hatbthotuuia, G. Hallas, R. J. Birtles, and A. G. Taylor: Phenotypic variation amongst genotypically homogenous Legionella pneumophila serogroup I isolate s: implications for the investigations of outbreaks of Legionna ires' disease. Epidem . Infect. 104 (1990) 171-180
Legionella pneumophila Serogroup 1 Subtyping
187
12. j oly, J. R., R. M. McKinney, j . O 'H. To bin, W. F. Bibb, I. D. Watkins, and D. Ramsa y: Development of a standardized subgrouping scheme for Legionella pneum opbila serogroup 1 using monoclonal ant ibodies. J. Clin. Microbiol . 23 (1986) 768-771 13. Luck , P. Ch., J. H. Helbig, Ch. Pilz und W. Wit zleb : Nachweis von Legionellen in klinischen Untersuchungsmaterialien mit FITC-mark ierten Antikorpern, Z. ges. Hyg. 35 (1989) 601- 604 14. Maher, W. E., M. F. Para, and]. F. Plouffe: Subryping of Legionel/a pneumophila serogroup 1 isolates by monoclonal antibody and plasmid techniques. ] . Clin. MicrobioI. 25 (1987) 2281-2284 15. Mitchel, E., M . O 'H. Mahony, j . M . Watson, D. Lynch, C. j oseph, C. Quigley, R. Aston, G. N. Constable, J. R. Farrand, S. Maxwell, D. N. Hut chinson, and]. Craske: Two outbreaks of Legionn aires' disease in Bolton Health District. Epidem. Infect. 104 (1990) 159-1 70 16. Ne ill, M. A., G. W. Gorman , C. Gibert, A. Roussel, A. W. Hightower, R. M. McKinney, and C. V. Broome: Noso comial legionellosis, Paris, France - Evidence of transmission by potable water. Am. J. Med. 78 (1984) 581-588 17. O tt, M., L. Bender, R. Marre, and j. Hacker: Pulsed field electrophoresis of genomic restriction fragments for the detection of nosocomial Legionella pneum opbila in hospital water supply. J. Clin. Microbiol. 29 (1991) 813-815 18. Pfaller, M., R. Hollis, W. j ohnson, R. M. Massanari, C. Helms, R. Wenzel, N. Hall, N. Moyer, and ]. joly: The application of molecular and immunologic techniques to study the epidemiolgy of Legionella pneum ophila serogroup 1. Diagn. Microbiol. Infect. Dis. 12 (1989) 295-302 19. Ruf, B., D. Schiirmann, I. Horbach, K. Seidel, and H. D. Pohle: Nosocomial Legionella pneum ophila: demon stration of potable water as the source of infection. Epidem. Infect. 101 (1988) 647-654 20. Stout, J. E., J. joly, M. Para, J. Plou ffe, C. Ciesielski, M. J. Blaser, and V. L. Yu : Comparison of molecular methods for subtyping patients and epidemiologically linked environmental isolates of Legionella pneum oph ila. J. Infect. Dis. 157 (1988) 486--495 21. Watkins, I. D., J. O'H. Tobin, P. J. Dennis, W. Brown, R. Newnham, and ]. B. Kurtz: Legionella pneum oph ila serogroup 1 subgrouping by monoclonal antibodies - an epidemiological tool.]. Hyg. (Camb.) 95 (1985) 211-21 6 22. Wilk inson, H. W.: Hospital-Laboratory diagnosis of Legionella-infection s. Center for Disease Control, AtiamalUSA (1988) 23. Winn, W. c.: Legionn aire Disease: Historical perspective. Clin. Microbiol. Rev. 1 (19 88 ) 60-8 1
Dr. Paul Christian Luck, Institut fiir Medizinische Mikrobiologie, Medizinische Akademie Dresden, Giintzstr . 32, 0 -8019 Dresden, Germany
