LIi‘iRARY Michigan State University This is to certify that the thesis entitled EVAL'UATLN (“r AK} INDI‘lC-k‘r HLIVIP cpl, INMi ’-v ThtiT "a"; 77—114“ch .w‘ AN ific‘m TC 6:-.r26‘fgypm 1'4 5;— L ‘ "mt/(‘3: 5”) ‘_/1_tl LL" AN”) 'r‘; ; » IN Q“- ., L4 . x r" ,4" A .9EI\C£,(')‘[7L;"’HULL.CJIC {OILVLV C: ’JCRICAJV {-L.>1()c:v ' § presented by CHERVL A YCNKE has been accepted towards fulfillment of the requirements for N19 is . dggree in P} I: 03¢. {3 "i L I: I; 7 /./("1/ “brim \jajor professor I ’- fi . g‘ ’1'. ,". Date N5. V/Q'J.frt/l IV” I fix, 0-7639 MSU LIBRARIES ” RETURNING MATERIALS: Piace in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. EVALUATION OF AN’INDIRECT BEHAGGLUTINATION TEST FOR DETECTION OF ANTIBODY TO SEROGROUPS 1-4 OF Legionella pneulophila AND ITS USE IN A SEROEPIDEHIOLOGIC SURVEY OF MICHIGAN RESIDENTS. Dy Cheryl A. Yonke A THESIS thnitted to Michigan State University in partial fulfillment of the require-euro for the degree of MASTER OF SCIENCE Depart-out of Microbiology and Public Health 1980 ABSTRACT EVALUATION OF AN INDIRECT NEHAGGLUTINATION TEST FOR DETECTION OF ANTIBODY TO SEROGROUPS [-4 OF Legionella pneungphila AND ITS USE IN A SEROEPIDEMIOLOGIC SURVEY OF MICHIGAN RESIDENTS. 37 Cheryl A. Yonke Parallel teating of 89S eera by the indirect heaagglutination (INA) and indirect fluoreecent antibody (TIA) uethode for legionelloaie ehoqad 97.31 agreeaent between the techniquee. Although the IRA ehoead lore croee reactivity between eerogroupe than the IPA, the etiological aero- group could eaeily be defined. Since the IEA.uae shown to detect both 133 and IgG claee antibodiee and uae found to be rapid, einple, and in- expeneive, it appeare to be an excellent alternative to the IPA teat for eerodiagnoeia and eeroepideaiologic etudiee of legiouelloaie. when eaployed in a aurvey of 1200 apparently healthy Michigan reeidenta, the IRA ehoved eerogroup 1 antibody the neat prevalent; 71 (11.82) of 600 eera collected during the winter (January-April, 1980) and 131 (21.82) of 600 eera collected during the eu-aer (July-Septeaber, 1980) denou- etrated eerogroup l titere. Prevalence of antibody to eerogroupe 2, 3, and 4 vae significantly lower, all showed less than 12 prevalence regard- leee of eeaaon of the year. ABSTRACT EVALUATION or AN INDIRECT NENAGGLUTINATION TEST FOR SEROGROUPS 1-4 0? Legionellajpneunophila. By Cheryl A. Yonke, Harlan E. Stiefel, David L. Nilson, and Berttina Hentworth Parallel testing of 895 eera by indirect henagglutination and ins direct fluorescent antibody techniques showed 97.32 agree-eat. Although the INA usually showed lore cross reactivity between eerogroupe than the INA using fornalin fixed antigens, heterologous serogroup reactions were significantly lower than honologous serogroup titers and the etiological eerogroup could easily be defined. The INA showed no cross reactivity with a crude extract of 35 £91.; 013:K92:H4. Since the INA was shown to detect both IgN and IgG class antibodies and was found to be rapid, einple, and inexpensive, it appears to be an excellent alternative to indirect fluorescent antibody testing for serodiagnosis of legionellosis. ABSTRACT vacuum: or mxsonr 'ro smocnours H. or ‘Legionells pagyhils: A smosrrnmrowcrc sum! usnn nu: unmet amccwnmrrou us'r By Cheryl A. Yonke, Narlan E. Stiefel, David L. wilson, and Berttina Neutworth An indirect henagglutination test was used to deter-inc the prevalence of antibody to serogroups l-& of Legionelle‘pneunophila in sera free 1200 apparently healthy Michigan residents. Serogroup 1 antibody was the nest prevalent; 71 (11.8!) of 600 eera collected during the winter nonths (Januarqupril, 1980) and 131 (21.82) of 600 eera collected during the sun-er period (July-Septeaber, 1980) deaonstrated serogroup 1 titers. This seasonal difference was independent of sex and was statistically significant in four of six age groups studied. A trend towards decreasing prevalence in the 50-59 and 60 or older age groups was noted in the winter sanple and was statistically significant in the sunset study. Prevalence of antibody to serogroupe 2, 3, and A was significantly lower; all showed less than 11 prevalence, regardless of season. ACKNOWLEDGEMENTS My deepest appreciation goes to Dr. Berttina Nentworth for her patience, encouragement, and guidance throughout this project. I an especially grateful to Dr. Naria Patterson for serving as ny acadenic adviser and for serving on ny graduate eo-ittee. I owe each to Dr. Nor-an NcCullough for his extra devotion to our infectious disease seni— nars and for serving on ny co-ittee. Special thanks go also to Harlan Stiefel for his technical assistance, and to David Wilson for preparation of the antigens used in this study. I an also indebted to Dalice Henge, Donna Bottoeley, Dr. I.C. Chang, Carlton Evans, Patricia Poye, Marge Niegarth, [eh-Hing Pan, and countless others at the Michigan Departuent of Public Health for their help. Finally, I would like to thank Char Currie and Dace Valduss for their friendship and support during our years at N.S.U. ii PART I. PART II. TABLE OF CONTENTS Llrmm mIWOOOOOOOOOOCOOOOOOOOOOOOOOOOO0.00.00.00.00 ImowcrIonOOOOOOOOOOOOO00......00.00.00.000...0.0.0.... m DlsusEOOOOOO.COOOOOOOOC.0.000000000000000QOOO0...... Clinic‘l “fie.utmeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee P‘thOlogyCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.00.0.0...O. Dum.1BOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.000...O. “thicrObi‘l mer‘p’OOOOOOOOOOOOOOOOOOOOOOOOO00.0.0... In mmmOOOOOOOOOOO0.0.0000....00....0.00.00.00.00... Tmc Ch..1f1at10n000000000000OOOOOOOOOOOOOOOOOOO mrpholozyOOOOOOI..0...OO....0.000000000000000000000.CO BiOChuiatry ‘nd Ph’.1°1°8y00OOOOOOOOOOOOO0.00.0.0.0... “rahgyOOOOOOOOO0.0...00.0.00...OOOOOOOOOOOOOOOIOOOOO. "mmIOMeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee lemz CIMOOOOOOOOOOOOOOOIOOOOOOOOOOOOO0.0.0.0.0... EVALNATION OF AN INDIRECT NEMAGGLDTINATION TEST FOR SEROGROUPS l-fi OP Legionella pneunophila................. ImowCTIONOO...0......OOOOOOOOOOOOOOOOOOOOO0.00.0.0.0. “mm m MODSOOOOOOOOOOOOOOOOOOOOOOOO0.000...... Sera for parallel testing of INA and IPA techniques... Sera selected for co-parison of INA and IRA........... Sera for cross absorption and blocking fluid .‘uinOOOOOOOOOOIOOOOOOOOOOOOOOOO0.00.00.00.00... Sera for sucrose density gradient fractionation....... “ct“ul .tr‘inBOOOOOOOOOOOOOOOOOOOOOOOOOOOO000...... IPA ‘ntig‘n prep‘r‘tionOOOOOOOOO0.0000000000000000000. I“ .ntisen prmr‘tionOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 1" tOChniqu.OOOOOOOOOOOO00.0.0000...OOOOOOOOOOOOOOOO. I“ tcchniunOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. ct... .bwrptionOO0.00000000000000000000000000000000.0 Preparation of hyperin-une rabbit sera................ iii Page N 0‘ UIUUN @GNO 13 21 21 22 22 22 22 22 23 23 23 24 26 25 25 l . . O a s e e v 0 e . . a I . e I O . I a Q r a I s . e a Q I D a e . a w . . \ . n u 0 a I a . I a | . . . O . . o e I . O O . O a O a e e O . . . . e s D E U n O n . e e I n O s e O O a o c O O i e O l e a e e O c i I e O . s a s e e e e I O a . s e e e I v 0 O . a a w e . . e I w e l r C I I I a I I a s p I s e e s e . O o d O a Q 0 O a O I O O - Q I I A e e o e e . e e I r . . e C C O \ s v e e a . I . I p a p . o a o . a I e . e . O i I . I v D O a I C I o I n s u s I a v . a 0 e O C O . . . a s o e . I . e a e s I e e v s O . . e b w n p s A O a e u I I l . O u . e I e e e e a. O . . a w . w A C . 0 e I l O O s O a Page Conparison of in-une eera prepared against IPA ‘nd INA ‘ntisuaeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 25 BlOCking fluid prep.rat1°u000000000000000000.0000... 25 Sucrose density gradients........................... 26 waiffmionoO0.00.0.0...OOOOOOOOOOOOOOOOOOOOOOOO 26 RESULTSOOOOOOOOOOOOO00......OOOOOOOOOOOOOOOOO0.0.0.0... 26 Parallel testing of 895 eera by the INA and IFA tectmtque.0000OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 26 Conparison of IPA and INA on selected eera fro- 55 ”tientQOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0 28 Conparison of in-une sera prepared against IPA .nd m antingOOOOOOOOOOOOOOOOOOOOOOOOOOIOO 31 Blocking fluid and cross absorption studies......... 31 Fractionation of sera on sucrose density gradients to deternine the in-unoglobulin class reactive in the INA and IPA tests........... 36 DISCUSSIOlOOOOOOOOOOOOOO0.0000......OOOOOOOOOOOOOO0.0.. 36 lem CImOOOOOOOOOUOOO...OOOOOOCOOOOOOOOOOOOO... ‘1 PART III. PREVALENCE OP ANTIBODY SEROGROUPS l—b OP Legionella pneunophila: A SEROEPIDENIOLOGIC STUDY USING THE INDIRECT HEHAGGLUTINATION TEST.............. 63 ImODUCHOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO... ‘3 “mules m “mmSeeeeeeeeeeeeeeeeeeeeeeseeeeeeeee “ S“.OOOOOOOOOOOOOOO0......0.00.00...OOOOOOOOOOOOOO. ‘4 m t..tOOOOOOOOOOOOOOOOOOOOOOOQOOOOOOOO0.0.00.0... “ 8t.t1.t1c.eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee ‘5 RESULTSOOOOOOOOOIOOOOOOOOOOOOFUIOOOOOOIOIOOOOOO‘OOOI... ‘5 Prevalence of Antibody to Serogroup 1.............. ‘5 Prevalence of Antibody to Serogroups 2. 3’ “a ‘OOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0.0... ‘7 DIWSSIOIOOOOOOOOCOOOOOOOOOOOOOOOIOOOOOOOOOOOOOOO... 50 lem CImOOOO0......0.0.0.0...OOOOOOOOOOOOOOOO 53 iv LIST OF TABLES Table Page Part I Part II 1 Results of parallel testing by INA and IPA on 895 sera......27 2 Diverse serogroup reactivity of the huaan antibody response as neasured by the INA and IPA tests.....................29 3 Use of the cross absorption technique to define the etiological serogroup....................................3O A Conparison of IPA and INA results on 55 patients with IPA confirned or presunptive legionellosis...............32 5 Coaparison of innune sera prepared against IPA and INA anti'mOOOO0..0.0.0.0...0......OOOOOOIOOOOOOOOOOOOCOO0.033 6 Results of blocking fluid and cross absorption studies 1‘ th. I“ m ”A t..t.000000000000000000.0.0000000000003‘ 7 Innunoglobulin reactivity of 15 sera fractionated on sucrose density gradients as neasured by INA and IPA.....37 8 Conparison of INA and IPA results on ten fractions of a convalescent phase serun...............................38 Part III 1 Nunber and percent of sera with INA titers equal to or greater than 1:16 to serogroup l of Legionella EMhihOOOOOOOOOOOOOOOO0.0000000000000000.00.000.0000‘6 2 Geonetric nean titers to Laggonella pnequphila, serogroup 1, by age and sesson...........................A8 3 Distribution of INA titers to serogroup 1 of Legionella ppeunophila by season and sex.................49 LITERATURE REVIEW Introduction The Anerican Legion Convention at Philadelphia's Bellevue Stratford Hotel in July, 1976, captured world-wide attention when an outbreak of pneumonia occurred anong conventioneers. The pneuaonia, subsequently designated "Legionnaires' disease," was diagnosed in 182 persons who had been in or near the hotel (35). There were 29 deaths, a 162 nortality rate overall. A.nunber of possible etiologies were investi- gated, including those of toxic and infectious nature, but it was not until January, 1977, that the causative agent becane known. At that tine, NcDade and associates (59) at the Center for Disease Control succeeded in isolating a grsnvnegative bacillus after intraperitoneal injection of lung tissue fro- four fatal cases into guinea pigs and subsequent passage of the infected guinea pig spleen into enbryonated hens' eggs. Indirect fluorescent antibody (IPA) testing on sera fron surviving patients clearly defined the etiologic role of this infectious agent. Purther investigation proved that neither the organise nor the disease had gone entirely unrecognised in the past; IPA testing on stored sera fron four previous epidenics (40, 52, 76, 77) of pneunonia identified the Legionnaires' bacillus as the causative agent. In addition, an unclassified agent, OLDA, isolated by Jackson‘stugl. (45) in 1967 was shown by DNA.hybridisation studies and G+C (guanine + cytosine) content to be identical to the Philadelphia isolates (58). l TEE DISEASE Clinical Nanifestations Legionellosis has been shown to nanifest itself in one of two for-s: 1. An acute pulnonary forn (Legionnaires' disease) observed in several epidenics and a large nunber of sporadic cases, or 2. A.nilder, self-liniting fora, terned "Pontiac fever,” seen in only two outbreaks to date (3‘, 80). The pneunonic fora begins with nalaise, nyalgia, and a nonproductive cough, following a two to ten day incubation period. Nithin two to three days of onset, a high, unrenitting fever develops, frequently acconpanied by relative bradycardia and pleuritic pain. Radiographic studies reveal patchy infiltration early in the disease, with increasing consolidation as the disease progresses. Routine laboratory exaninations con-only denonstrate a noderate leukocytosis with a left-shift, proteinuria, hyponatr-ia, hypophosphatenia, asoteaia, increased aninotransf erases, and an increased erythrocyte sedinentation rate. Although largely pul- nonary, the disease features several extrapulnonary effects including gastrointestinal (1, 35, £9, 72), renal (68), and central nervous systen anonalies (53). A classical gran-negative endotoxin has been proposed to explain these nultisystea effects (23, 38). The organise caused gelatin of Linulus anebocyte lysates (AA, 91), but showed only low pyrogenicity in rabbits (91). No investigation has clearly denonstrated the existence of such a toxin. Pontiac fever presents as a adlder, non-pneunonic fora of the disease with a significantly shorter incubation period (seen 36 hours). Pever, npalgia, and headache characterise the disease. Cough, diarrhea, voniting, chest pain, and a sore throat were described in s snsll nunber a" of patients, but did not appear to be proninent features of the syndrone. Hepatic and renal involve-cut or fatality has not been observed. Patholggz The only consistent pathological feature in fatal cases of Legion- naires' disease is the finding of lesions in the lungs. The pneunonia is lobar, with no lobes preferentially involved (90). Consolidation is usually bilateral. Only recently has the Legionnaires' disease organise been denonstrated outside the thoracic cavity (86). Microscopic exanination of lung tissues fron several Legionnaires' disease patients has revealed an acute fibrinopurulent pneunonia (4). Alveolar spaces contain an exudate conposed of neutrophils, necrophages, and large anounte of fibrin. Lysis of these inflannatory cells has been associated with an increased nunber of the Legionnaires' bacillus (4). Coagulative necrosis has been observed in a few cases (90). The bacteriun stains poorly with routine histologic stains, including henatoxylin-eosin, Brown-Brenn, Brown-Nopps , and NcCallun—Goodpae ture; visualisation of the organise in tissue sections or inprints is best achieved with the nodified Dieterle (13, 86) or other silver impregnation stains (26). Diagggsis Diagnostic criteria for legionellosis include: 1. Direct ig_vitro isolation of the organise. The first direct iggvitro isolation of the Legionnaires' bacillus was nade in 1976 by Dunoff (21). The organise, recovered fron pleural fluid, was cultured on GC base chocolate agar containing 11 hemoglobin and 11 Isovitslex. I. Since that tine, a nuaber of prinary isolation nedia have been developed, including Nueller—Ninton—Isovitalex-Nenoglobin (MN-IN) (31), Peeley- Cor-an (P—G) (31), and charcoal yeast extract (CYE) agars (29). P-G and CPR are supple-anted with ferric pyrophosphate and L-cysteine hydrochloride, the cysteine being an essential growth require-ant and the iron serving as a stinulatory nutrient (85). ‘Eggionella pneunophila has been isolated fron blood (23, 56), pleural fluid (21), transtracheal aspirates (22, 51), and lung tissue. On the whole, direct in 11552 culture has not been highly successful, probably due to the fastidiousness of the organise, overgrowth of contaninant organisna, antibiotic inhibition, or loss of viability. 2. Staining of the organise by the direct fluorescent antibody technique (9, 14, 15). This is, at present, the only rapid diagnostic technique available to the physician. The test appears to be highly specific. Cherry and associates (15) tested 37‘ strains of bacteria representing 25 genera and 59 species by the direct in-unofluorescent technique; only one strain of Pseudo-ones fluorescens was found to cross react with Eggionella specific confugate. A strain of Pseudo-onas alcaligenes has since been reported to cross react in the test (9). 3. Detection of antibody by the indirect innunofluorescent test (IPA). The indirect fluorescent antibody test has been used to diagnose the najority of legionellosis cases. A four-fold rise in titer on paired sera or a single serun titer of at least 1:128 is considered diagnostic (59). The IPA is apparently quite specific; however, four- fold rises in titer in hunan cases subsequently diagnosed as lepto- spirosis, plague, and tudarenia have been reported (83). The test is disadvantageous in that diagnosis can he ends only retrospectively, 5 since significant titers do not generally develop until at least 16 days after onset of the disease (75). Antieicrobial therapy Present reconnendations for antibiotic therapy are based largely on a review of the effectiveness of these agents in treating past cases of Legionnaires' disease. Fraser 25_gl;_(35) observed that case fatality rates were highest for those treated with cephalosporins, internediate in those treated with aninoglycosides, chloranphenicol, anpicillin, and penicillin, and lowest in those treated with tetracycline and erythro- nycin. In one of the earliest susceptibility trials, Nash and associates (6‘) found erythronycin and ninocycline hydrochloride effective in an 13 13:2 guinea pig nodal. Fraser and co-workers (36) showed erythro- nycin end rife-pin effective using a sinilar guinea pig nodel, while penicillin, chloranphenicol, tetracycline and gentsnicin showed no sig- nificant effect. Lewis 35 _a_l_. (55), using an enhryonated hens' egg nodel, showed rifanpin and gentsnicin to have the highest prophylactic effective- ness, followed by streptonycin, erythronycin, sulfadiasine, chloranphen- icol, cephalothin, oxytetracycline, and chlortetracycline, in decreasing order. In _v_i_t_rg studies do not correlate well with 12.1122 results. A study by Thornsberry, Baker, and Rirven (79) showed Legionella ppeunophila susceptible to rife-pin, cefoxitin, erythronycin, the aninoglycosides, ndnocycline, doxycycline, chloranphenicol, anpicillin, penicillin G, carbenicillin, colistin, and sulfanethoxazole-triaethoprin, in that order. In the sane study, _I_._._ pneunophila showed internediate susceptibility to tetracycline, nethicillin, cephalothin, cephanandole, and clindanycin and was resistant to vanconycin as neasured by nininun inhibitory 6 concentration endpoints. Saravolatz and colleagues (71) found little difference in in zitrg susceptibility anong serogroups 1-4, the only notable differences being an increased susceptibility of serogroup 2 to penicillin and serogroups 3 and 4 to sulfanethoxasole-trinethoprin. Thus relatively little is certain in terns of the antinicrobial susceptibility of Legionella pneunophila. The lack of correlation between _i_n_ 3.1.99. and in 2.1.19. studies say be explained by relative differences in intracellular bactericidal action on Legionella-laden nacrophages (80). Cephalosporin ineffectiveness is undoubtedly due to a recently discovered e lactanase (81) which is unlike that of either the Enterobacteriaceae or Pseudo-ones (37). Although linited, existing data suggest erythronycin and rifanpin to be the best available therapeutics for treat-ant of Legionnaires' disease. Erythroaycin is considered the drug of choice, with rife-pin as a reserve for patients unresponsive to erythronycin, as widespread use of rifanpin.nay result in the eaergence of resistant tubercle bacilli. THE ORGANISM Taxononic classification The organise isolated by NcDade in 1977 (59) had growth and staining patterns as well as biochenical characteristics far different than any previously described species. DNA.hcnology studies convinced Brenner and codworkers (7) that they were indeed dealing with a previously unrecognised species. The Legionnaires' bacillus was Sound to have a gene-e size of 2.5 x 109 daltons and a G +1C content of 391. when Brenner and Steigerwalt (7) conpared the Legionnaires' agent with 20 7 bacterial species having conparable G + C ratios, none showed nore than 31 DNA relatedness. They felt there was sufficient evidence to declare the organisu a new species. Legionella pneuaophila sp. now. has been proposed as the type species of the genus flimella. Morphology Egionella pneuophila is a gran-negative bacillus neasuring 0.3 - 0.9/un by 2.0 ,«u or greater in length (12). Pilanentous forns are observed; length, however, is largely dependent on environment. Both "roller pin" and "spindle-shaped" form of the organise have been reported, though their frequency and significance is not clear (65). Transaission electron uicroscopy studies (11, 12) reveal well-defined riboseues, a double envelope enclosure conposed of a triple-unit nenbrane, and a pinching type of division, all typical of gran-negative bacilli. The detection of dianinopinelic acid in low concentration in the cell wall further substantiates its gran-negative character (Al). The bacillus contains large vacuoles which stain readily with Sudan Black B (11 , 12) . The ultrastructural appearance of these granules suggest that they are poly @ bydroxybutyrate in nature. Although legionella was originally thought to be non-stile, flagella have recently been denonstra- ted on strains of serogroups 1-6 (69,78). Colonial norpbology of 1.; Emphila varies sonewhat with the nedi- of isolation. Colonies of CC base agar are described as gray and glistening (21). Colonies on CTE, F-G, and NH-IN agars usually develop after 3-5 days of incubation and have a ground-glass appearance (30, 87). A brown pignent is observed on HH-IN and other tyrosine-containing nedia, possibly due to L-phenylalanine hydroxylase activity (3) . Colonies exhibit a yellow fluorescence under long wave ultraviolet light (366m) (30). Biocheuistry and Physiolggz. Legionella pneunophila is a strict aerobe, and, in fact, is sensitive to excesses of oxygen (67). The organise can grow under nicroaerophilic conditions, but a 2.5: CO: atnosphere is superior. Acid pH is preferred, with a pH of 6.9 as optinuu. Oleic acid is inhibitory and seleniun stinulatory to growth (66). Serine and threonine serve as the primary energy sources (39). An investigation by Tison 3t 3_l_._ (82) indicated that the tenperature, pH, and nutritional require-ants of E; pneuuophila in nature say not be as strict as those observed when the organise is cultured on couples nadia. Nhen grown in association with cyanobacteria, Legionella pneunophila grew well over wide pH and tenperature ranges. 'This association, the investigators felt, night account for the wide distribution of 1.; memphila in nature. Biocbeuically, the organise is relatively inert. It is weakly oxidase positive, liquifies gelatin, and hydrolyzes starch. Nitrates are not reduced, carbohydrates not utilized, and urea not degraded. Purther-ore,‘§;_pneuuophila lacks lysine and ornithine decarboxylases and arginine dehydrolase. Thus, laboratory identification based on routine biochenical tests alone would be virtually inpossible. The cellular fatty acid conposition of L=_pneunophila is unique and serves an an aid to identification. Gas-liquid chroaatography and less spectronetry indicate the cellular fatty acids are largely (81-901) branched chain. The iso C16:O acid is predouinant, followed by antesio 015:0, antesio 617:0, iso 016:0, and iso 016:1, in decreasing order (62, 63). Serology Six serogroups of Legionella pneunophila have been described to date by the direct fluorescent antibody technique. The Knoxville 1 strain . "‘ s 1 . n o " F " ‘r , if: , I . . . ' ‘ ‘ I Q . . . . \ ‘ . . . ' ‘ . I 4‘ g . . ' t . '- . I‘ V I . a ' v . " ‘ ‘ ' t. t p . ‘ . I ' t , - ‘x 1. - . .. . I , . , O ,, ‘ . ‘ . I ‘c ) ; s , ' e r , . ~ ‘ . . A ‘ ' ‘ J . .‘ , _ . . . . , ’. (91‘..‘ .. ,‘ . . ' o ‘ n . ' . I t V J ‘ ‘ v ' 3r' “ ; - J 'l ' " I . . ’ ' " w l h ‘ h . , 4 . '. , ‘ I . ‘ . . . . . I I .U, ‘- ‘ ‘ , serves as the prototype strain for serogroup 1. This serogroup con- tains the original Philadelphia isolates and the vast eajority of clinical strains, as well as nine of eleven‘uichigan isolates. Togus 1 serves as the prototype of serogroup 2, a group which includes patient isolates free the Togns, Heine outbreak and two environeental isolates free the Atlanta, Georgia outbreak of 1978 (60). For eany eonths, serogroup 3 contained only the Blooeington, Indiana creek water isolate (61), the prototype strain, but recently hueen serogroup 3 isolates have been recovered (32), including one free Michigan (Saravolatz, Chang, and Hentworth, unpublished). Serogroup 6 includes Los Angeles 1 (60), the prototype strain, as well as several isolates free the wadsworth V.A. Hospital epideeic. Berogroup 5 was identified after an outbreak of legionellosis at a Veterans of Foreign wars convention in Dallas, Texas in the fall of 1978 (25). The prototype strain, Dallas 18, was recovered free water saeples froe a nearby cooling tower. Serotype 6 of h pneueo- philg_wes recovered free pleural fluid or lung tissue froe three renal transplant patients in Chicago (16). Isolates were also recovered free 3 of 6 shower heads in the ward where these patients had stayed. It has recently been proposed that three Legionella-like bacteria be placed in the genus Legionella: HIGA, first recovered by Bose-an (5) in 1959, is to be designated l:_boseeenii (6): TATLOCK (76), which has been found to be identical to the Pittsburgh pneueonia agent (66), is to beddesignated Legionella eicdadei (63). Legionella dueoffii (6) is the proposed uses of the NY 23 strain (19). It appears that the Legionnaires' disease patient can develop an ieeunological response against serogroup specific antigenic detereinants as well as against detereinants coeeon to serogroups 1—6 (89). we are 10 only beginning to characterise these antigenic determinants. Gong g£_£l=_ (92) isolated a serospecific antigenic fraction consisting of a lipid- protain-carbohydrate coeplex which protected guinea pigs and nice against challenge with lethal doses of §:_pneueophila. SDS acrylsnide gel electrophoresis resolved this into 6 protein and 1 glycoprotein band. Johnson and codworkers (67) identified soluble high eolecular weight coeplexes of serogroups 1-6, designated Fraction-1 (F-l). Ieeunodiffusion studies deeonstrated serospecificity; P—l showed single precipitin lines with hoeologous serogroup sntisera and no precipitin lines with heterolo- gous serogroup sntisera. Both the coeplexes of Hong and Johnson elute in the void voluee of Sepharose 68 coluens and contain carbohydrate, but their exact relationship has not been assessed. EPIDEMIOLOGY A review of epideeic, sporadic, and nosocoeial cases of Legionnaires' disease reveals soee salient epideeiologic features of the ealady (2, 8). Infection is greater in sales than feeales by a ratio of approxieately 2.6 to 1. Cigarette seeking and alcohol abuse appear to be high risk factors. leeenocoeproeised and ieeunosuppressed patients seee to be especially susceptible to the disease (62). A late sueeer-early fall seasonality is apparent, both in Legionnaires' disease cases and in anti- body levels in a healthy population (26). The node of transeission in legionellosis is uncertain, but coepelling evidence points to an airborne route. A 1965 epideeic of Legionnaires' disease at St. Elisabeth's Hospital in washington, 0.0., the earliest docueented outbreak, indicated wind-blown dust as the source of infection (77). Sporadic cases of Legionnairea' disease have also been associated 11 with excavation and construction sites (2). The organise has been isolated froe evaporative condensers and air cooling towers in at least nine epideeics (17, 18, 20, 25, 36, 60, 62, 57, 68). Person-to-person spread has not appeared to be an ieportant asses of transeission. There were no secondary cases of legionellosis aeong rooe- sates and fseily eeebers of the Philadelphia conventioneers (35). In con- trast, the increased prevalence of antibody to E; pneneophila in hospital eeployees is consistent with this node of disseeination (70). Personrto- person transeission appeared likely in the 1976 case of a Scottish physician who acquired Legionnaires' disease after treating a patient with this disease (10); this has not been unequivocally proven. Accueulating data show an association between infection and reservoirs of the Legionnaires disease bacillus in the inanieate environeent and indicate that 1.; maesphila say be a canon inhabitant of our surroundings. The organise was isolated free.weter and soil of a nearby streee after one outbreak of legionellosis (61); Flier-ans and associates (33) showed fluorescing cells of §=_pneueophila in 902 of water seeples collected froe 23 lakes in Georgia and South Carolina. Although fastidious in the lab- oratory, Legionella seees to survive well in stringent environeental con- ditions: strains of the organise survived 139 days in distilled water and 369 days in tap water (73). With the rapidly-expanding genus of Legionella, serodiagnosis by IRA is bacoeing exceedingly coeplex. A test which is rapid and inexpensive, yet sensitive and specific, is needed. Hicroagglutination (28), ieeune adherence (56), card agglutination (50), and ELISA (27, 28) techniques are currently under developeent. It was the purpose of this study to define sore clearly the sensitivity and specificity of a eicroheeagglutination 12 test for legionellosis and to evaluate further its usefulness in sero- epideedologic studies. LITERATURE CITED 5. 6. 10. LITERATURE CITED Baine, W.B. 1979. Clinical eanifestations of Legionnaires' disease and recon-ended therapy, p. 3-8. lg_G.L. Jones and G.A. Hebert (ed.), "Legionnaires'": the disease, the bacteriue, and eethodology. U.S. Department of Health, Education, and welfare, Center for Disease Control, Atlanta, Ga. Baine, w.n. 1979. The epideeiology of Legionnaires' disease, pe 13-16e Refs ls Seine, ".8. and J .K. Rasheed. 1979. Aroeatic substrate specific- ity of browning by cultures of the Legionnaires' disease bac- Blackeon, J .A., 11.1). Hicklin, rm. Chandler, and the Special Expert Pathology Panel. 1978. Legionnaires' disease: pathological and historical aspects of a new disease. Arch. Path. Lab. Med. ‘122: 337-363. Bozeean, F.M., J.W. Huephries, and J.M. Caepbell. 1968. A new group of ricksetta-like agents recovered froe guinea pigs. Brenner, D.J., A.C. Steigerwalt, c.w. Gorean, R.E. weaver, J.C. Peeley, C. Patton, 8.11. Thoeason, and K.R. Lewallen- Sasseville. 1980. Legionella bozeeanii species nova and Legionella dueoffii species nova: classification of two addi- tional species of Le ionella associated with huean pneueonia. Curr. Microbiol. ‘5; 115-120. Brenner, D.J., A.C. Steigerwalt, G.A. Robert, and J.E.‘HcDade. 1979. Legionella pneueophila sp. nov.: The Legionnaires' disease bacteriue, p. 59-66. Ref. 1. Brooee, C.V. and D.U. Fraser. 1979. Epideeiologic aspects of legionellosis. Epideeiol. Rev; 1; 1-16. Brooee, C.V., v.3. Cherry, w.c. Winn, and B.R~ HacPherson. 1979. Rapid diagnosis of Legionnaires' disease by direct ieeuno- fluorescent staining. Ann. Intern. Med. ‘22: 1-6. Ceepbell-Love, Hi, A.K.R. Chaudhuri, and 1.0. Chin. 1978. Possible case to case transeission of Legionnaires' disease. Lancet. 1.3.: 1269. 13 11. 12. 13. 16. 15. 16. 17. 18. 19. 20. 21. 16 Chandler, E.W., J.A. Black-on, M.D. Hicklin, R.M. Cole, and 0.8. Callaway. 1979. Electron eicroscopy of the Legionnaires' disease bacteriue, p. 61-66. Ref. 1. Chandler, P.W., R.M. Cole, M.D. Hicklin, J.A. Black-on, and 0.8. Callawey. 1979. Ultrastructure of the agent of Legion- naires' disease in the huean lung. Ae. J. Clin. Pathol. _7_1: ‘3‘50e Chandler, E.W., M.D. Ricklin, and J.A. Black-on. 1979. Demonstra- tion of the agent of Legionnaires' disease in tissue. N. Engl. J. Med. 297: 1218-1220. Cherry, 11.8. and 11.11. McKinney. 1979. Detection of Legionnaires' disease bacteria in clinical specieens by direct i-unofluo- rescence, p. 91-103. Ref. 1. Cherry, W.R., B. Pittean, P.P. Barrie, G.A. Hebert, 8.11. Thomson, L. Thacker, and R.E. Weaver. 1978. Detection of Legionnaires' disease bacteria by direct inunofluorescent staining. .7. Clin. Microbiol. _8_: 329-338. Cordes, L.C. 1980. Serogroup 6 legionellosis in renal transplant patients. Possible transeission froe contaeinated shower water. Abstract. EIS Officers Conference, Center for Disease Control, Atlanta, Ga. Cordes, L.G., D.W. Eraser, P. Skaliy, C.A. Perlino, W.R. Elsea, 0.1". Hallison, and P.S. Reyes. 1980. Legionnaires' disease outbreak at an Atlanta, Georgia, country club-evidence for spread froe an evaporative condenser. An. J. Epidemiol. 111: 625-631. Cordes, L.G., W.D. Gold-en, J.S. Herr, 8.11. Eriedxan, J.D. Band, E.O. Rothschild, R. Rravetz, J .0. Peeley, D.W. Fraser, and the Field Investigation Teas. 1980. Legionnaires' disease in New York City, August-SeptaIber, 1978. Bull. 1LT. Acad. Med. _5_6_: 667-681. Cordes, L.G., R.W. Wilkinson, G.W. Cor-an, B.J. Pikes, and D.W. Eraser. 1979. Atypical Legionella-like organises: fastidious water-associated bacteria pathogenic for nan. Lancet. .151: 927-930. Dondero, D., R.C. Rendtorff, 1'. Hallison, 11.11. Weeks, J.S. Levy, E.W. Wong, and W. Schaffner. 1980. Outbreak of Legionnaires' disease associated with a contaeinated air conditioning cooling tower. N. Engl. J. Med. _3_(_)_2_: 365-370. Dueoff, H. 1979. Direct in vitro isolation of the Legionnaires' disease bacteriue in two fatal cases. Cultural and staining characteristics. Ann. Intern. Med. 9_0: 696-696. 22. 23. 26. 25. 26. 27. 28. 29. 30. 31. 32. 33. 36. 15 Rdelstein, P.H. and S.M. Pinegald- 1979. Isolation of Legianells pneueaphila froe a transtracheal aspirate. J. Clin. Microbiol. _9_: 657-658. Edelstein, P.H., R.D. Meyer, and S.M. Finegald. 1979. Isolation of Legionella maeueaphila froe blood. Lancet. _i.: 750-751. Edson, D.C., H.E. Stiefel, 8.8. Wentworth, and D.L. Wilson. 1979. Prevalence of antibodies to Legionnaires' disease: a sera- apid-ialagic survey of Michigan residents using the heeagglu- tination test. Ann. Intern. Med. 20: 691-693. England, A.C., R.M. McKinney, P. Skaliy, and 0.W. Gar-an. 1980. A fifth serogroup of Legionella pneuaphila. Ann Intern. Med. 93: 58-59. Raine, 8., P. Edelstein, 8.D. Kirby, and S.M. Pinegald. 1979. Rapid presueptive bacteriological diagnosis of Legionnaires' disease. J. Clin. Microbial. 15;: 106-105. Parahy, 0.8. and J.C. Peeley. 1979. p. 126-129. ELISA for Legionnaires' disease antibody and antigen systees. An interim report. Ref. 1. Parshy, 0.8., 0.0. Klein, and J.0. l'eeley. 1978. Detection of antibodies to Legionnaires' disease organise by eicraagglu- tinstian and eicra-enryee-linked i-unasarbent assay tests. Jo C1111. 111011313101. 1: 327-3310 Faeley, J.C., R.J. Gibson, G.W. Gar-an, 11.0. Langfard, J.R. Rasheed, D.C. Msckal, and W.8. Baine. 1979. Charcoal yeast extract agar: prieary isolation eediu for Legianella pneueaphila. J. Clin. Microbial. 12: 63 l. Feeley, J.C., G.W. Gorean, and R.J. Gibson. 1979. Prieary isola- tion eedia and eethads. p. 77-86. Ref. 1. Peeley, J.C., 0.W. Gar-an, 11.8. Weaver, D.0. Mackel, and H.W. Seith. 1978. Prieary isolation eedia for the Legionnaires' disease bacteriue. J. Clin. Microbial. 3: 320-325. Fisher-Roch, 8., M.J. Hudson, and M.H. Thaepson. 1979. Identifi- cation af a clinical isolate as L_e_gionella pnetnaphila by gas chroeatagraphy and ease spectroeetry of cellular fatty acids. Lancet. 1.13 323-325. Flier-ans, 0.8., 17.8. Cherry, L.H. Orriaan, and L. Thacker. 1979. Isolation of ngianella nan-aphila frae nan-epideeic-related aquatic habitats. Appl. Environ. Microbiol. 31: 1239-1262. Fraser, D.W., 'D.0. Deubner, D.L. Hill, and D.R. Gillian. 1979. Hanpneueonic, short incubation period legionellosis (Pontiac fever) in use who cleaned a steae turbine condenser. Science 393: 690-691. 35. 36. 37. 39. 60. 61. 62. 63. 66. 65. 66. 67. 16 Fraser, D.W., T.E. Tsai, W. Orenstein, and others. 1977 Legionnaires' disease: description of an epideeic. H. Engl. Jo Mo £2} 1189-1197e Eraser, D.W., 1.x. Wachseith, C. Hopp, J.C. Feeley, and ‘1‘.P. Tsai. 1978. Antibiotic treateent of guinea pigs infected with agent of Legionnaires' disease. Lancet. _i_: 175-177. Pu, R.P. and 8.0. Men. 1979. Inactivation of beta-lactae anti- biotics by _L_e_gianella pneuaphila. Antisicrob. Agents J Chuather. _1_6_: 561-566. Euarola, D. 1979. Pathogenic eechsnises in Legionnaires' disease. 8r. Med. J. _2_: 796. George, J.R., L. Pine, M.W. Reeves, and W.K. Harrell. 1980. Alina-acid require-ants of legionella pneueaphila. J. Clin. Microbiol. _1_1_: 286-291. Click, T.H., 11.8. Greg, 8. Harlan, 0. Mallisan, W.W. Rhodes, and I. Rassanaff. 1978. Pontiac fever: an epideeic of un- known etiology in a health depart-eat. Ae. J. Epidaeial. 121: 169-160. Guerrant, 0.0., 11.8. Laebert, and 0.W. Mass. 1979. Identification of a diaeinopieelic acid in the Legionnaires' disease bacteriue. J. Clin. Microbial. 10: 815-818. Haley, 0.8., M.L. Cohen, J. Halter, and R.D. Meyer. 1979. Hasocaeial Legionnaires' disease: a continuing caeean source epideeic at Wadsworth Medical Center. Ann. Intern. Med. 29; 583-586. Hebert, G.A., A.G. Steigerwalt, and D.J. Brenner. 1980. Legianella eicdadei species nova: classification of a third species of La ionella associated with husn pnetnania. Curr. Microbial. 3: 2553257. Highseith, A.R., 8.0. Mackat, W.8. Raine, R.L. Anderson, and D.W. Fraser. 1978. Observations of endotaxin-like activity associated with the Legionnaires' disease bacteriue. Curr. Microbial. 1_ 315-317. Jackson, 8.8., T.T. Cracker, and J.E. Seadel. 1952. Studies on two rickettsia-like agents probably isolated frae guinea pigs. 8act. Prac. p. 119. Jacques, P.A., D.L. Sealley, and J.S. Payne. 1980. Enhanced growth of Legionella pneueaphils in presence of seleniue. J. Ae. Med. Assoc. 26 : 27. Johnson, 17., J.A. Elliot, C.M. Hales, and 8.8. Renner. 1979. A high ealecular weight antigen in Legionnaires' disease bacteria: isolation and partial characterisation. Ann. Interns Med. 29} 638’6‘1 e a: o 1. 0-‘, l r o x I c 68. 69. 50. 51. 52. 53. 56. 55. 56. 57. 58. 59. 17 Kerr, D.N., R.A.L. 8rewis, and A.D. Macrae. 1978. Legionnaires' disease and acute renal failure. Br. Med. J. 2: 538-539. Kirby, 8.D., R.M. Snyder, R.D. Meyer, and S.M. Einegald. 1978. Legionnaires' disease: clinical features of 26 cases. Ann. Intern. Med. _82: 297-309. Rleger, 8. and R.A. Hartwig. 1980. Develop-cut and evaluation of a card agglutination titer (CAT) test for serodiagnosis of Legionnaires' disease. Public Health Lab. 38: 267-256. Lattiear, 0.L., 0. McCrone, and J. Calgon. 1978. Diagnosis of Legionnaires' disease froe transtracheal aspirate by direct fluorescent antibody staining and isolation of the bacteria. 8. mgl. J. Med. _2_9_9: 1172-1173. ~ Lawson, J.H., 11.x. Grist, 1). Reid, and 1.3. Wilson. 1977. Legionnaires‘ disease. Lancet. ii: 1083. Less, A.W. and W.8. Tyrell. 1978. Severe cerebral disturbances in Legionnaires' disease. Lancet. .i_i_: 1136-1137. Lennette, D.A., E.T. Lennette, 8.8. Wentwarth, M.L.V. French, and 0.L. Lattieer. 1979. Serology of Legionnaires' disease: caeparisan of indirect fluorescent antibody, iaune adherence haagglutinatian, and indirect heeagglutinatian tests. J. Clin. Microbial. _1_Q: 876-879. Lewis, V.J., W.L. Thacker, 0.0. Shepard, and J.E. McDade. 1978. _Ig viva susceptibility of the Legionnaires' disease bacteria to ten antieicrabial agents. Antieicrab. Agents Cheeather. 13: 619-622. Macrae, A.D., P.W. Graves, and P. Platte. 1979. Isolation of Legionnella pneueaphila frae blood culture. Br. Med. J. .2_: 1189-1190. Marks, J.S., T.E. Tsai, W.J. Martone, 8.0. Heron, J. Rennicat, R.J. Holtshauer, I. Baird, D. Fay, J.0. reeley, 0.17. Mallison, D.W. Fraser, and T.J. Halpin. 1979. Naeocoeial Legionnaires' disease in Calabus, Ohio. Ann. Intern. Med. 9_0-: 565-569. McDade, J.S., D.J. Brenner, and P.M. Hose-en. 1979. Legionnaires' disease bacteria. isolated in 1967. Ann Intern. Med. _9_0: 659- 661. McDade, J.8., 0.0. Shepard, D.W. Fraser, T.E. Tsai, M.A. Redus, W.R. Dawdle, and the Laboratory Investigation Teas. 1977. Legionnaires' disease: isolation of a bacteria and deeanstra- tion of its role in other respiratory disease. N. Engl. J. Med. _2_9_7_: 1197-1203. it. 60. 61. 62. 63. 66. 65. 66. 67. 68. 69. 70. 71. 72. 18 McKinney, R.M., L. Thacker, P.P. Harris, R.R. Lewallen, 0.A. Hebert, P.H. Edelstein, and 8.M. Thoeasan. 1979. Four serogroups of Legionnaires' disease defined by direct i-unofluarescence. Ann. Intern. Med. _9_9_: 621-626. Morris, 0.8., C.M. Patton, J.C. Feeley, 8.8. Johnson, 0. Canaan, W.'1‘. Martin, P. Skaliy, 0.8. Mallison, 8.8. Paliti, and D.C. Mackel. 1978. Isolation of the Legionnaires' disease bacteriue froe environeental seeples. Ann Intern. Med. 29; 666-666. Mass, 0.W., R.E. Weaver, 8.8. bees, and W.8. Cherry. 1977. Cellular fatty acid coepasition of isolates froe Legionnaires' disease. J. Clin. Microbial. _6_: 160-163. Mass, 0.W., 8.8. Weaver, S.8. Bees, and W.8. Cherry. 1979. Cellular fatty acid coepasitian of the Legionnaires' disease 13.08.21“. Po ‘7-5‘e Refs ls Hash, P., L. Sideean, V. Pidcoe, and 8. Rleger. 1978. Minacy- cline in Legionnaires' disease. Lancet. 3;: 65. Heblett, T.R., J.M. Riddle, and M. Dueaff. 1979. Surface tapa- grapby and fine structure of the Legionnaires' disease bac- teria: a study of six isolates in hospitalized patients. Ann. Intern. Med. _9_0_: 668-651. Pascuelle, A.W., R.L. Myerowitz, and 0.8. Rinaldo. 1979. New bacterial agent of pneueania isolated froe renal transplant patients. Lancet. _i_: 58-61. Pine, L., J.R. George, M.W. Reeves, and W. Knox Harrell. 1979. Physiology: characteristics of the Legionnaires' disease bac- teria in saisynthetic and cheeically defined liquid eedia. pa 27“0e his ls Paliti, 8.D., D.W. Eraser, 0.8. Mallison, J.V. Mahatt, 0.11. Morris, C.M. Patton, J.C. Feeley, R.D. Telle, and J.V. Bonnet. 1979. A esjor focus of Legionnaires' disease in Bloaeington, Indiana. Ann. Intern. Med. _9_0: 587-591. Rodgers, 17.0., P.W. 0reaves, and A.D. Macrae. 1979. Plagella and finbrae an Legionella organises. Lancet. _i_: 753-756. Sarsvolatz, L., L. Arking, 8. Wentworth, and 8. Quinn. 1979. Prevalence of antibody to the Legionnaires' disease bacteria in hospital eeployees. Ann. Intern. Med. 22: 662-663. Saravolatz, L.D., D.J. Pohlod, and 8. Quinn. 1979. _Igvitra susceptibility of Legionella pneaaphila serogroups 1-6. J. Infect. Dis. 160: 251. Schwartz, M.H. 1979. Clinical aspects of Legionnaires' disease. Ann. Intern. Med. _9_(_)_: 692-695. 73. 76. 75. 76. 77. 78. 79. 80. 81. 82. 83. 86. 85. 86. 19 Skaliy, P. and H.V. McEachern. 1979. Survival of the Legionnaires' disease bacteria in water. Ann. Intern. Med. 9_0_: 662-663. Tatlock, H. 1966. A rickettsia-like organise recovered froe guinea pigs. Proc. Soc. Exp. Dial. and Med. 27: 95-99. Taylor, A.0. and '1'.0. Harrison. 1979. Tieing of antibody response in Legionnaires' disease. Lancet. _i_i_: 697. Terranova, W., M.L. Cohen, and D.W. Fraser. 1978. 1976 outbreak of Legionnaires' disease diagnosed in 1977: clinical and epi- deeiological features. Lancet. _i_i_: 122-126. Thacker, 8.8., J.V. Bonnet, T.F. Tsai, D.W. Fraser, J.8. McDade, 0.0. Shepard, R.H. Williaes, Jr., W.H. Stuart, 8.8. Dull, and '1'.C. Rickhoff. 1978. An outbreak in 1965 of severe respira- tory illness caused by the Legionnaires' disease bacteria. Jo Infecto D1.o 129;: 512-519o Tho-soon, 8.M., P.W. Chandler, and D.0. Hollis. 1979. Flagella on Legionnaires‘ disease bacteria: an interie report. Ann. 10:011. Med. 21: 22“”225o Thornsberry, 0., C.M. Baker, and L.A. Rirven. 1978. _Igvitro activity of antieicrobial agents on Legionnaires' disease bacteria. Antieicrab. Agents Cheeother. 1_3_: 78-80. Thornsberry, C. and L.A. Rirven. 1979. Antieicrobial suscepti- bility of the Legionnaires' disease bacteria. p. 55-58. Ref. 1. Thornsberry, C. and L.A. Rirven. 1979. Blactaease of the Legionnaires' bacteria. Curr. Microbial. _1_: 51-56. Tison, D.L., D.H. Pope, W.8. Cherry, and 0.8. Fliereans. 1980. Growth of Legionella pneueophila in association with blue- green algae. Appl. Environ. Microbiol. _3_9_: 656-659. Tsai, ‘1'.1". and D.W. Fraser. 1978. The diagnosis of Legionnaires' disease. Ann. Intern. Med. _82: 613-616. Van Orden, A.E. and P.W. Greer. 1979. Deeonstration of the bacteria in tissues by a eadification of the Dieterle silver iepregnation stain. p. 105-109. Ref. 1. Warren, W.J. and R.D. Miller. 1979. Growth of Legionnaires' disease organise in cheeically defined eedia. J. Clin. Micro- bial. _1_9: 50—55. Watts, J.C., M.D. Hicklin, 8.M. Thoeason, 0.8. Callaway, and A.J. Levine. 1980. Fatal pneueonia and bactereeia caused by Le ionella Enemaphik, serogroup 3: demonstration of the bacifi' i in extrathoracic organs. Ann Intern. Med. 2_2_: 186-188. 87. 88. 89. 90. 91. 92. 20 Weaver, R.8. and J.C. Feeley. 1979. Cultural and biochemical characterization of the Legionnaires' disease bacteriue. p. 19-25o Refs lo Wilkinson, H.W., D.D. Cruse, 8.J. Pikes, L.P. Yealy, and 0.8. Farshy. 1979. Indirect ieeunafluorescence test for Legionnaires' disease. p. 111-116. Ref. 1. Wilkinson, H.W., 8.J. Pikes, and D.D. Cruce. 1979. Indirect ieeunofluorescence test for serodiagnosis of Legionnaires' disease: evidence for serogroup diversity of Legionnaires' disease bacterial antigens and for eultiple specificity of huean antibodies. J. Clint Microbiol. ‘2; 379-383. Winn, W.C., Jr., R.L. Glsvin, D.P. Perl, J.L. Keller, T.L. Andres, T.M. Brown, C.M. Coffin, J.R. Sensecqua, L.H. Roean, and J.R. Craighead. 1978. The pathology of Legionnaires' disease: fourteen fatal cases froe the 1977 outbreak in Vereont. Arch. Pathol. Lab. Med. 192: 366-350. Wong, K.H., C.W. Moss, D.H. Hochstein, R.J. Arka, and W.O. Schalla. 1979. Endatoxicity of the Legionnaires' disease bacteriue. Ann. Intern. Med. 29: 626-627. Wong, K.H., W.O. Schalla, R.J. Arka, J.C. Ballard, and J.0. Foeley. 1979. Ieennacheeical, serologic, and ieeunologic properties of eajor antigens isolated froe the Legionnaires' disease bacteriue. Ann. Intern. Med. 22; 636-637. Eweluation of an Indirect Heeagglntination Test for Serogroups 1-6 of Logionella pneaophila 8y Cheryl A. Yonkal, Harlan 8. Stiefelz, David L. Wilsana, and 8erttina 8. Wentworth2 1 Departesnt of Microbiology and Public Health, Michigan State University, East Lansing, Michigan, 68826 2 Huroau of Disease Control and Laboratory Services, Michigan Departeent of Public Health, Lansing, Michigan, 68909 3 Difco Laboratories, Inc., Detroit, Michigan, 68201 INTRODUCTION The indirect fluorescent antibody (IFA) test for legionellosis (9, 16) has becoee an ieportant diagnostic tool in the three years since its introduction. Perforeing the test on a routine basis, however, is becoeing an increasingly coeplex task; ‘with the recent addition of the agents TATLOCK (12), WICA (1), and HT 23 (6) to the genus Legianella (proposed designations L_._ eicdadei (8), I; bozeeanii, and _1_._._ dueoffii (2), respectively), proper serodiagnosis of Legionnaires' disease (LD) involves the use of nine separate antigens. The use of polyvalent antigens (7, 16) can sieplify testing soeewhat, but even this does not discount sees in- herent disadvantages of the technique, i.e. that it is technically diffi- cult and expensive to perfore. As a practical alternative, Edson and associates (5) introduced an indirect heeagglutination technique (IRA) for legionellosis. The test is rapid, inexpensive, sieple to perfore, and showed a 96.92 agreeeent with IFA sergroup 1 (81) results. The IHA test was recently expanded to include serogroups 2, 3, and 6 (82, S3, and S6) of Legionella pneueaphila, and this study was eade to evaluate the sensitivity and specificity of the eethod. 21 MATERIALS AND METHODS Sara to: parallel testing of mi and IFA‘techniques Sera were received at the Michigan Departeent of Public Health for routine IFA legionellosis testing between March and October, 1979, All sera received during this period were included in the study except for those where the quantity of serue was insufficient for testing by both techniques. The saeple tested included a total of 895 sera. There were single serue specieens free 683 patients and paired or serial sera free 188 patients. Sara selected for caeparison of IHA and IFA Sara froe 55 patients with IRA confireed (four-fold or greater rise in titer) or presaptive (titer‘21:128) legionellosis were used. Single serue spocieens were available froe 8 patients and paired or serial sera free 67. These cases represent 61! of the 136 cases of legionellosis diagnosed by serology between October, 1977, and August, 1980. The only additional criterion for inclusion in the study was sufficient serue values for testing. Sers for cross absorption and blocking_fluid studies The thirteen sera with serogroup 1 titers used were froe IFA con- fireed or presueptiva cases. 0f the twelve sera with E; coli titers, four were froe LD patients and eight froe patients without legionellosis. Sara for sucrose densitygradient fractionation Twelve sera selected froe the 55 patients described above were used. An additional three sera, which showed four-fold or greater rises, or titers 2:1:128, by IHA, which were IFA negative, were fractionated for 22 23 analysis. Bacterial strains Bacterial strains were grown in Roux bottles an enriched chocolate agar (00 eediue base with heated sheep blood and vitaedn.suppleeents). Flint 1, Flint 2, and Detroit 1 strains of Legionella pneueophila were eaintainod as stock cultures at the Michigan Depart-ant of Public Health. Strains of Knoxville 1, Togus 1, Blooedngtan 2, Las Angeles 1, and E; coli 013: K92:H6 were obtained froe the Center for Disease Control (Atlanta, Georgia) . IFA antigen preparation Bacterial growth was washed froe the surface of the Roux bottle after 72 hours of incubation with 0.62 forealinized phosphate buffered saline (FPBS), filtered through a fine eesh wire screen to reeove agar debris, than refrigerated (6° 0) for one week. The suspension was then washed three tiees in FPBS and the opacity standardized to 300 International Opacity Units per el (IOHAel) by coeparison to W.H.O. opacity standards. Suspensions of forealinized whole organises, adjusted to 2 IOU/e1, served as antigen. IHA.antigengpreparation IHA.antigens consisted of turkey erythrocytes stabilized by glutsraldahydo treat-ant and sensitized with bailed, sonicatod antigen froe agar-grown isolates by the bis-diasotized benzidina technique. Anti- gens were propared free the following strains: 81 antigen as a eixture of Flint 1, Flint 2, and Detroit 1; S2 froe Togus 1; S3 froe Blooeington 2; and 86 free Loo Angsles 1. The polyvalent antigen was eade froe a eixture 26 of Knoxville 1, Togus 1, 8looeington 2, and Los Angelas 1. All anti- gens were supplied as exporieental antigens by Difco Laboratories, Inc. (Detroit , Michigan). IFA technique A standard eicrotitor technique using 0.005 el antigen seounts on 12 well acetone-resistant glass eicroscope slides was eeployed. Con- jugates were diluted in phosphate buffered saline (PBS) with 22 Tween 80. Rabbit anti-haan globulin conjugate was purchased froe Beckean Instru- eents, Inc. (Fullerton, California). Cappel (Cochranville, Pennsylvania) anti-rabbit globulin conjugate was used to detereine titers on hyperi-une rabbit sntisera. Anti-haan IgM conjugate was supplied by Burroughs Wellcoea Co. (Research Triangle Park, North Carolina). IHA technique The IHA technique was sieilar to that of Edson _e_t_ _a_l_._ (5). Unless otherwise noted, sera were screened at an initial 1 :16 dilution against polyvslsnt sensitized cells and unsensitized control cells. Any sera showing a reaction with the test, but not control, cells was subsequently screened with the four eonovslont (SI-S6) antigens and corresponding control cells. Sera were titrated to endpoint in serial two-fold dilu- tions with appropriate eonovslont antigens. Any sera showing agglutination of both antigen sensitized and un- sensitized control cells was treated in the following eannor: 0.02 II of packed, unsensitized (control) cells were added to 0.12 el of sera. After overnight incubation at 60 0, the cell suspension was centrifuged at 3,000 rpe for 15 ein., the absorbed sera reeoved, and the IRA test repeated. Sera with high titers of anti-call antibody required two or 25 three successive absorptions for coeplete removal of these agglutinins. Cross absorption, For absorption, 0.1 e1 of serue was added to approxieately 3.7 x 1010 organises which had been previously washed three tines in PBS to reeove forealin. The absorption eixture was incubated overnight at 6° C, then centrifuged at 3,000 rpe for 15 ein., and the absorbed serue raeaved. Preparation of hyperieeune sera Rabbits, greater than 2 kg in weight, were ieennized with an initial intraenscular injection.of 1‘el of antigen nixed with an equal values of Freund's coeplets adjuvsnt, followed by five intravenous boosters of l‘el of antigen at days 22, 26, 26, 28 and 30. Rabbits were exsanguinated nine days after the last booster. Coeparison of ieeuno seragpreparedagainst IPA and IHA antigens Rabbits were divided into four groups of two rabbits each and ieeunized as described above. Groups 1 and 3 received injections of IFA forealinized antigens at 16 IOU/e1. Groups 2 and 6 received injections of IHA antigen (boiled, sonicatod) at 16 IOU/e1. Groups 1 and 2 were inocu- lated with the Flint 1 strain Legionella pneueophila. Groups 3 and 6 were ieeunized with Detroit 1 strain of serogroup 1. Sera froe two rabbits in each ieeunization group were pooled, then IFA and IHA titers detereined on each of the pools. Blocking fluid preparation Blocking fluids of Flint 1 and §g_coli 013:K92:H6 were prepared in the eanner described by Wilkinson gt'al. (13). Sera were initially 26 diluted in IHA diluent or PBS as a control and in both blocking fluids. Subsequent dilutions were eade in diluent or P88 and IHA.or IFA titers detereined. Sucrose density gradients Sera were fractionated on 10-502 sucrose density gradients (11) by centrifugation at 33,500 rpe for 17 hours in a Spinco ultracentrifuge with a SW 50.1 rotor. Ten fractions of 0.5 el were collected for analy- sis. Ieeunodiffusion The presence of IgM and/or IgG in each of the fractions was con- fireed with Malay (Springfield, Virginia) radial ieeunodiffusion kits. These kits eeploy the Fahey (6) technique to eeasure ieeunoglobulins. RESULTS Parallel testingrof 895 sera by the IHA and IFA techniques Results of parallel testing of 895 sera by the IHA and IFA techniques are shown in Table 1. There were 796 nonreactive and 66 reactive sera by both techniques in serogroup 1. There were 16 acre reactive in the IRA which were nonreactive in the IHA at a 1:16 dilution. There were 39 sera reactive in the IHA that failed to react in the IFA at a 1:16 dilution. Discrepancies in serogroups 2, 3, and 6 were due largely to the cross reactive nature of 81 antibody and not to sera with specific 82, 83, or 86 antibody. Agreeeent (: sera positive by both techniques plus 2 sera negative by both techniques) was 96.11, 98.32, 99.62, and 97.52 for 81, 82, S3, and 86, respectively. There was an overall .IJ 27 TABLE 1 : Results of parallel testing by IHA and IFA on 895 sera Ho. of sera that were: _81 _82 _83 _fl IHA-l IFA- 796 879 885 858 IBA+IIFA+ 66 1 5 15 m-IIFA-I- 16 0 2 2 IBA+IIFA- 39 15 3 20 TOTAL 895 895 895 895 ( - - titer (1:16, «I» - titer 21:16) 28 agree-ant for serogroups 1-6 of 97.32. Along the 188 paired or serial sera, there were nine instances of four-fold or greater increases in titer by both IRA and IFA techniques and four discrepancies. One patient with a four-fold rise in IFA titer showed a two-fold rise by IHA and one patient had the reverse (i.e., a four-fold rise by IHA.and a two-fold rise by IFA). Another patient with a greater than four-fold increase in IFA had a high stable titer (1:2068) by IHA. Finally, serue froe a patient with successive IFA titers of 1:256 and 1:128 in 81 was IHA negative. Sons of the diverse reactivity between serogroups of the huean anti- body rosponsa noted by Wilkinson :5 El; (16) in the IFA.was observed with the IRA. However, the IFA, using forealin fixed antigen, did not appear to be so cross reactive in this study. Representative results are shown in Table 2. Where this cross reactivity was seen, it was found that the etiologi- cal serogroup could best be defined by cross absorption. Absorption of a serue with a serogroup specific antigen reeoved the cross reacting as well as hoealogous antibody. Absorption with a heterologous serogroup antigen reeoved only the cross reacting antibody, but did not decrease the serogroup specific titer. Ten cases of serogroup 1 and one case of sera- group 6 were identified by cross absorption. Two exaeples are shown in Table 3. Patient 1 appears to have 81 legionellosis and patient 2 to have LD of 86 etiology. Coeparison of IHA and IFA on selected sera froe 55 patients Of the 55 patients with IFA confireed or presueptive legionellosis, 56 were identified as serogroup 1 and the other as serogroup 3. The IFA detected four-fold or greater rises in titer in 62 patients (61 81, 1 S3), '1‘.‘ a 3 v I - 1 l r. I ., C l I 29 TABLE 2 : Diverse serogroup reactivity of the ban antibody response as eoasured by the IHA and IFA tests Titer of convalescent phase sera by: Legionella pneaophila antigen: IFA IHA IFA IHA IFA IHA __n__z_8m on .5. _A_ .2. .3. _c_ .9. 1 1:66 1:512 1:66 1:16 1:128 1:2068 2 < 1:8 (1:16 (1:8 1:256 (1:8 1:32 3 < 1:8 (1:16 1:512 1:512 <1:8 (1:16 6 (1:8 1:32 1:16 1:128 1:8 1:128 30 33V meow: meow: 33V ouuuv con.— cm 33V 33V odzv 33V emu—V cuuuv mm "cows,senoa ounce asawmouoa Hsoweououue sea onuuou ou essences» oowueuoaes snows sea «a so: 33v 23v caudv 33V 33V 33v «a 33V 23V onus emu»— o—uu v emu: do. uoooaonssnoo no need» «In a 819.4 .3 .33 63.3.5 a one: 58 nee—85 88883.. e .885 .3 Eu. 388.2 a one: Sr. 888:: 63883: a.” no.8... n53 31 37 ( 36 81, 1 83) of when also showed significant rises by IHA (Table 6). In three of the renaining five patients, the IHA showed a rise froe 1:16 to 1:16 and in two cases the IHA was nonreactive. Seven of 13 patients with presueptive legionellosis showed IHA titers 21:128, one showed a rise froe <(1:16 to 1:16, two showed titers between 1:16 and 1:128, and three were nonreactive in the IHA (Table 6). It was considered that the differences between IHA and IFA results eight be explained by: 1. eeasureeent of two different antigen-antibody systees; 2. reactions with cross reacting antigon(s); or 3. differences in ieeunoglobulin specificity. To explore each of these possibilities, the following studies were conducted: 1. Coeparison of ieeune sera prepared agginst IFA.and IHA antigens Results of IFA and IHA testing of hyperieeune rabbit sera, prepared with IRA and IHA antigens, are shown in Table 5. If two different antigen- antibody systees are involved, titers with hoealogous antigens eight be expected to be higher than those with heterologous antigens. However, there were no significant differences in titers with IFA and IHA antigens, regardless of the antigen used to prepare the antisera. 2. Blockigg fluid and cross absorption studies Wilkinson and associates reported that a crude extract of E, 521; 013:K92:H6 could block 972 of IFA positive reactions with a variety of gran-negative bacterial species, and blocked 62 of Legionella pneuegphila titers (13). This strain of 5. 52}; was obtained froe Dr. Wilkinson and blocking fluids of this and Flint 1 strain were prepared. Twelve sera with anti-serogroup 1 IHA titers were tested as shown in Table 6. Dilution in E. ggli_blocking fluid had no significant effect on any of the 12 sera tested, while Flint 1 blocking fluid caused a 32 mm 2 no Hack ”Nu « "N. NIH—"u. ea nouns aw eeasuoaw nausea» no waounusoh a m N a h N o n o 3: v eat—v u...— 2: 8 33w 3.3m- uoud. 23A 839 #3.... 5 83 .833 . eeepauelssewm an mm eweoHHe Lacuna.— .elanaou 33.30 1.33 garden-seen.— 33.5.0 sesame.— eve-.33 (NH swsodaeflouued sane-seen:— wo assuage «nu new: 383!— nn so auasseu a was «mu no usual—loo u c 53 33 TABLE 5 : Coaparison of issune sera prepared against IPA and IHA antigens Iaannisation llaunising Titer by* Group Antigen '32; 3:5_ 1 Flint 1, IRA 1:6096 1:2048** 2 Flint 1, IHA 1:4096 1:4096** 3 Detroit 1, IRA 1:4096 1:2048*** 4 Detroit 1, IHA 188192 1:6096*** * Results shown are the average value of triplicate detereinations ** Ilint 1 served as the test antigen *** Detroit 1 served as the test antigen TABLE 6 : 34 Results of blocking fluid and cross absorption studies in the IDA and IFA tests Effect on Titer Blocking no Initial Fluid or Significant Serua Absorbing Decrease in 22; £23.52 £232.29; mm IHA Serogroup 1 Flint 1 B.F. 0/12 n " 5. 5.2;; 3.17. 12/12 n " Flint 1 Abs. 0/11 .. " 5. 59;; Abs. lllll IRA Flint 1 Flint 1 D.P. 0/13 « " _15. 59_1_1_ 3.1. 13/13 H " Flint 1 Abs. 0/9 .. " 59 52;;_Abs. 9/9 IFA E. 5_o_1_._i_._ Flint 1 B.F. 12/12 .. " _s_. 53;; 3.1!. 0/12 " " Flint 1 Abs. 7/3 .. n _z_. 53}; Abs. 0/8 Significant Decrease in Titer** * lo change in titer or two-fold decrease in titer ** Four-fold or greater decrease in titer D.P. - Blocking fluid used for initial dilution of serue Abs. - Whole cell antigen used to absorb serue 12/12 0/12 11/11 0/11 13/13 0/13 9/9 0/9 0/12 12/12 1/8 8/3 l v . - .‘ , I ' a .. I s a f‘ s e . ‘ . e . s '\ v . e e l r a: .\, . . . . \ ‘7 ‘ . e 1 . V U . I s a I ‘ n, V . e A a -. 35 four—fold or greater decrease in all. Blocking fluid studies in the IFA showed sieilar results. §_‘_ gel; blocking fluid had no effect on the 13 sera tested. Dilution in Flint 1 blocking fluid. reduced serogroup 1 titers by at least a four-fold de- crease in 1002 of the cases tested. Twelve sera with initial IFA _g_._ 521;; titers were sieilarly tested (Table 6). Flint 1 blocking fluid caused no significant decreases in titer in any of the 12 sera, while '_!_._ 321$ blocking fluid significantly decreased all titers. There was one patient with a concurrent four-fold rise in titer to both 5:. is}; and Flint l in which blocking was specific, i.e. Flint 1 blocking fluid reeoved only the Flint 1 titer and EL 5211' blocking fluid reduced only the §_._ 9311 titer. Another patient who showed a four-fold rise in SI titer to 1:128 by IFA and a eaxieI- IHA titer of 1:16 was blocked by Flint 1 but not _§_._ _c_o_l_:_1_ blocking fluids. Cross absorption was felt to be a note conclusive test for M antigens than the use of crude blocking fluid extract. Absorption of 11 sera with Sl IHA titers with whole cell _E_._ 53}; antigen caused no signifi- cant decrease in any 81 titers, while absorption with whole cell Flint 1 caused a four-fold or greater decrease in all ('hble 6). Sieilarly, in the IFA, absorption with whole cell _F; 52y; did not change Flint 1 titers on 9 sera, while absorption with Flint 1 organise significantly decreased all. Cross absorption studies on 8 sera with _B_=_ coli IFA titers proved specific on all but one serue. This serue had an initial 1!_._ £o_1_i_ titer of 1:32, and absorption with _F_._ 52}; organise decreased the titer to <1:8. Absorption with whole cell Flint 1 decreased the titer four-fold to 1:8. 36 3. Fractionation of sera on sucrose density gradients to detereine the ieeunoglobulin'class reactive in thelga and IFA tests A total of 15 sera were fractionated by the sucrose density gradient technique and the fractions analysed by both eethods. Results of IFA and IHA detereinations are sueearised in Table 7. Three sera free cases diagnosed by IHA but not IFA showed reactivity only in the IgM fractions. Ten sera showed reactivity in the IgM and IgG fractions in the IRA test, but showed reactivity only in the IgG fractions by IFA. Another serue showed reactivity in the IgG fractions by both techniques. The remaining serue showed reactivity in the IgG fractions with the IFA, but was non- reactive in all fractions with the IRA. This serue had an initial IFA titer of 1:128 (81) and was nonreactive in the IRA. It appeared that the standard antiglobulin conjugate used in the IFA did not detect IgM. To test this hypothesis, the fractions of 10 sera which showed the presence of Igfl by IHA.and not IFA were re-exaeined by IFA with an IgM specific conjugate. All sera now showed reactivity in the IgM fractions. Typical results are shown in Table 8. DISCUSSION Farallel testing of 895 sera showed the IRA and IFA to yield essentially the seas results for serogroups 1-4. There was approxieately 972 overall agreeeent between the two techniques and discrepancies were observed in both directions. IHA positive/IFA negative discrepancies were associated with the inability of the antiglobulin conugate used in this study to detect IgM. Nagingten and associates (10) reported that 3 of 22 LB patients showed a rise in IgM antibody without concurrent rise 37 TABLE 7 : Ieeunoglobulin reactivity of 15 sera fractionated on sucrose density gradients as eeasured by IRA and IFA Reactivity Reactivity Technique Friztiggs Friztiggs IFA 0* 12 IRA 13 11 * Rueber of sera 38 .mum .mmm c..ao«aaeeaeonanun Isuen cease uneoeedaenou e no enowuusuu use so nua:eou.