Molecular Epidemiology of Diphtheria in Russia,

Transcription

1 1064 Molecular Epidemiology of Diphtheria in Russia, Tanja Popovic, Svetlana Y. Kombarova, Michael W. Reeves, Hiroshi Nakao, Izabella K. Mazurova, Melinda Wharton, I. Kaye Wachsmuth,* and Jay D. Wenger Division ofbacterial and Mycotic Diseases. National Center for Infectious Diseases, and Epidemiology and Surveillance Division, National Immunization Program, Centers for Diseases Control and Prevention. Atlanta, Georgia; G. N. Gabrichevsky Institute for Epidemiology and Microbiology. Moscow, Russia The largest diphtheria outbreak in the developed world since the 1960s began in the Russian federation in One hundred fifty-six Corynebacterium diphtheriae strains from throughout Russia, selected for temporal and geographic diversity, were assayed by ribotyping and multilocus enzyme electrophoresis (MEE). These tests showed significant genetic diversity within the C. diphtheriae species, and ribotyping and MEE data generally correlated well with epidemiologic data. A distinct clonal group of C. diphtheriae isolates (ET 8 complex) emerged in Russia in 1990 as the current outbreak began, and as the outbreak has progressed, these organisms have made up increasingly larger proportions of the strains that are isolated. Furthermore, the main characteristic of the epidemic strains is a specific combination of ET 8 and ribotypes G1 and G4. This study confirms the epidemiologic utility of the molecular subtyping methods that detected the epidemic clone and addresses the clone's origin and relation to C. diphtheriae from throughout Russia. The largest diphtheria outbreak in the developed world since the 1960s is in progress in the Russian federation. It began in 1990, when 1211 cases were reported, a 2-fold increase over the previous year. In 1993, > 15,000 cases were reported [1 3], and in 1994,39,907 cases were reported [4]. The epidemic has spread throughout the country, and only in 1995 did reported cases decrease to a provisional total of 35,716 [1]. By the end of 1994, diphtheria had spread to all 15 new independent states of the fanner Soviet Union [4]. Incidences have been highest in the Moscow and St. Petersburg areas, where the outbreak first began. A notable feature of the epidemic is the predominance of cases among adolescents and young adults, in contrast to the usual pattern of childhood disease seen in the prevaccine era [1]. About 3% of reported cases have been fatal. In the Russian federation, school-age children are well-vaccinated (90% have received four or more doses of vaccine containing diphtheria toxoid); however, seroprevalence data indicate that large numbers of adults remain susceptible [1,4]. A similar seroprevalence profile is seen in other developed countries [5-11]. Because immunity induced by diphtheria toxoid wanes over time, booster doses must be administered to maintain protective antibody Received 23 April 1996; revised 1 July Presented in part: 96th General Meeting of the American Society for Microbiology, Washington, DC, May 1995 (abstract 21). Reprints or correspondence: Dr. Tanja Popovic, Diphtheria Research Project, Childhood and Respiratory Diseases Branch, DBMDINCID, 1600 Clifton Rd., Atlanta, GA * Current affiliation: Department of Agriculture, Science and Technology Cotton Annex, Washington, DC. The Journal of Infectious Diseases 1996; t 74: by The University of Chicago. All rights reserved /96/ $01.00 levels. Since 1994, public health authorities in the Russian federation have initiated aggressive efforts to vaccinate adults, and in 1995 the number of reported cases decreased. The factors contributing to the epidemic are thought to be massive importation of toxigenic Corynebacterium diphtheriae into a susceptible population combined with social factors (including population movements) that facilitated its spread [1]. Laboratory characterization of C. diphtheriae isolates may be useful in identifying epidemic strains, defining their virulence, and understanding potential modes of transmission. Previous molecular epidemiologic studies have been based on restriction enzyme digests of the C. diphtheriae genome [12, 13] and on hybridization of genomic Southern blots to a variety of probes for the following: rrna [14, 15], an insertion element [16, 17], diphtheria toxin, and corynephage fj and its attachment site [18, 19]. These studies have shown diversity among the biotypes and toxigenic and nontoxigenic strains [12, 13, 16-19], and they have proven to be more discriminating than phage typing [12, 13, 16]. Ribotyping and multilocus enzyme electrophoresis (MEE) have been used successfully in numerous epidemiologic studies of other bacterial pathogens to estimate the genetic diversity and structure in their natural populations and to characterize bacteriologic aspects of endemic and epidemic disease [20 22]. Therefore, to define the outbreak strain(s) causing the current diphtheria epidemic in Russia and to determine the genetic relatedness of C. diphtheriae isolates from geographically diverse areas within Russia, we assayed strains by MEE and ribotyping. Methods Bacterial Strains Isolatesof C. diphtheriae from throughout Russia were referred to the National Diphtheria Laboratory, Gabrichevsky Institute for

2 lid 1996; 174 (November) Molecular Epidemiology of Diphtheria 1065 Epidemiology and Microbiology, for culture confirmation and biotype and toxin determinations. Methods of isolation and identification are standardized throughout Russia [23]. One hundred fiftysix microbiologically well-defined isolates were selected for temporal and geographic diversity and for availability of epidemiologic information (table 1). Detection of Diphtheria Toxin and Toxin Genes Diphtheria toxin was initially assayed by immunodiffusion by use of a modified Elek technique [24]. At the Centers for Disease Control and Prevention, all isolates were biotyped and tested by the Elek assay, as recently recommended by the World Health Organization [25], and by polymerase chain reaction (PCR), which amplified the A subunit of the diphtheria toxin gene (tox) as previously described [26, 27]. Ribotyping DNA extraction by phenol. Isolates were grown overnight on trypticase soy agar with 5% sheep blood (Becton Dickinson, Cockeysville, MD) at 37 C. The entire culture was collected from the plate with a sterile cotton swab and transferred to l5-ml polypropylene tubes (Becton Dickinson, Lincoln Park, NJ) containing 10 ml ofste (100 mmtris-hci [ph 8],10 mmedta [ph 8], 150 mn! NaCI) and centrifuged at 8000 g for 5 min at 4 C. Supernatant was discarded, and the pellet was resuspended in 2 ml of lox TE (100 mm Tris-HCl [ph 8], 10 mm EDTA [ph 8]) before being centrifuged again at 8000 g for 5 min at 4 C. Supernatant was discarded again, and the pellet was resuspended in 2 ml of lox TE, and 20 f.ll of lysozyme (100 mg/ ml; Boehringer Mannheim, Indianapolis) was added to achieve a final concentration of 2 mg/ml. Samples were incubated at 3T'C for 1-2 h, and 100 f.ll of 20% SDS was added to a final concentration of 1%. Incubation in a water bath at 65 C for 10 min followed; 100 f.ll of proteinase K (10 mg/ml; Boehringer Mannheim) was then added to achieve a final concentration of 500 f.lg/ml. Samples were incubated again in a water bath at 60 C for 2 h. Extraction with phenol was done with Gel Lock Tubes 2b (3' -+ 5', Boulder, CO) according to the manufacturer's instructions. DNA was precipitated as described earlier [28] and resuspended in 200 f.ll of sterile water. Restriction endonuclease digestion ofchromosomal DNA. The amount of DNA in each sample was determined by the minigel Table 1. Designations and origin of 156 Corynebacterium diphtheriae strains isolated in Russia from 1985 to Strains in European Russia Year Vladimir, region of focus Other European regions Strains in Asian Russia Total 708, 709, 710, 711,712, 713, , 716, 717, 718, 719, 720, 721, ,* 724, ,751,752, ,756,* 757, 759,* 760* G4192,* 763, 764,* 765, 766, 767* G4172, 855, 856, 857* G4182, 858, 859, 860, 1743 G4197, G4199, G4202, 861, 862, 863, 864, 1737, 1738,1751, ,1725,1727,1728,* 1734, 1735, 1739, 1740, ,1703, 1704,* 1705* 726,748,* 1706,1744, ,* G4168, G4169, G4170,+ G4171,+ G4208,* G4209,* G4215, 762* G4178, G4179, G4190, G4193* G4181, G4185, G4186, G4187, G4188, G4196,* G421O, G4211, G4212 G4173, G4174, G4175, G4177, G4189, G4200,11 G4201, , 1736, 483, 484, 485, 486,489,490, ,1714,1715,1716, 1717, 1718,1719,1720,1724,1726, 1749, 1750, 1754, 1755, 1756, 865, 866, 867, G4203,t G4204 t G4205, G ,1884, 1885, 1886,~ 1887, 1888,f 1889, 1890, 1891, 1892, 1893, 1894, 1895, 1896, 1897, 1898,** 1899,** 1900,1901,1903, 1904, *Nontoxigenic strain. 'School. Family ta, -c, lib, ~D, **E.

3 1066 Popovic et al. lid ]996; 174 (November) method [29]. Approximately 6 fll of DNA was digested with BstEIl (10 U/mL) for 4 h at 65 C in a waterbath. Electrophoresis and Southern blotting ofrestricted DNA fragments. Following restriction, samples were heated at 65 C for 10 min. DNA fragments were separated by horizontal electrophoresis through 0.8% agarose gels at 60 V for 16 h. Molecular size standards, ladder and A (GIBCa BRL, Gaithersburg, MD), were included in the last lane of each gel. Gels were stained with ethidium bromide (l,ug/ml) for 30 min and photographed with type 55 Polaroid film (ISO 50; Polaroid, Cambridge, MA). DNA was transferred to a nylon membrane (20 X 20 em) by the method of Southern as described previously [28]. Digoxigenin-labeled edna probe for 16S and 23S RNA genes and digoxigenin-labeled molecular size standards were prepared, followed by the hybridization with digoxigenin-labeled cdna probe and detection as described previously [28]. MEE Strain extracts were prepared and MEE analysis was done as previously described [20], except that the cells were disrupted by sonication for 3 min with a 10-s-on/20-s-off pulse and in an ice bath with a cup hom sonicator (model XL ; Heat Systems Ultrasonics, Farmingdale, NY). Gel slices were stained for enzyme activities by the method of Selander et al. [21], and the following activities were observed: alcohol dehydrogenase, shikimate dehydrogenase, malate dehydrogenase, isocitrate dehydrogenase, 6 phosphogluconate dehydrogenase, glucose-6-phosphate dehydrogenase, NAD glyceraldehyde 3-phosphate dehydrogenase, NADP glutamate dehydrogenase, NADH diaphorase, catalase, indophenol oxidase, aspartate carbamoyltransferase, nucleoside phosphorylase, glutamate oxaloacetic transaminase, carbamylate kinase, adenylate kinase, phosphoglucomutase, leu-gly-gly peptidase, phe-leu peptidase, aconitase, mannose phosphate isomerase, and phosphoglucose isomerase. MEE Data Analysis Electrophoretic variants of each enzyme were considered to represent alleles of that enzyme and were each assigned different allele numbers; each strain was characterized by a list of allele numbers for the different enzymes. Each unique list of alleles was designated as an electrophoretic type (ET) and assigned an ET number [21]. The genetic relatedness of the ETs was illustrated as a dendrogram, which was generated by the average-linkage method of clustering the ETs, as described by Selander et al. [21 J, using an SAS macro program described by Jacobs [30]. Results A total of 156 isolates collected from 1985 to 1994 from 75 patients and 74 carriers (clinical data were not available for 7 isolates) were selected for study. Sixty-two isolates were from the Vladimir region (figure 1), where the long-standing and comprehensive diphtheria surveillance system, established there by the Russian Ministry ofhealth, provided strains continuously between 1985 and The remaining 94 strains were isolated in 11 different cities and regions throughout Russia (figure 1). Of the 156 isolates, 41 were isolated in the preepidemic period ( ), and 115 were isolated during the current diphtheria epidemic. Among the studied strains, 99 were of biotype gravis and 57 were ofbiotype mitis. All strains were analyzed by the Elek and Pf'R assays, which detect regions of the tox gene A subunit: 138 strains were toxigenic and 18 were nontoxigenic. No discrepancies between the results obtained by the Elek and per assays were recorded. The proportion of toxigenic strains among gravis and mitis biotypes was 92% and 82%, respectively. Ribotyping. Among 156 strains, 20 different ribotypes were identified, and only 2 strains were considered nontypeable (figure 2). Most ribotypes contained strains ofeither gravis or mitis biotypes. Four ribotypes, designated G1, G3, G4, and G4v, were seen only in strains ofthe gravis biotype. Fourteen ribotypes, designated Ml, M1v, M2, M3, M4, M5, M7, M8, M9, MI0, MIl, MI2, M14, and M15, were characteristic only for strains of mitis biotype. Ribotypes M6 and M13 were exceptional in that they had equal numbers (1 and 3, respectively) ofgravis and mitis strains. Most strains isolated in the Vladimir region belonged to ribotypes Gl (15 strains), Ml (18 strains), and G4 (11 strains). Four of 11 ribotype G4 strains were isolated in Vladimir before 1990; all 15 ribotype G1 strains were isolated since Ribotype Ml strains were isolated continuously before the beginning of the epidemic (10 strains) and during the epidemic (8 strains). The same pattern held for the strains isolated throughout Russia: Most analyzed strains belonged to ribotypes Gl (48 strains, 31%), G4 (40 strains, 26%), and Ml (37 strains, 22%). The remaining 29 strains were distributed among 15 other ribotypes, each represented by 1-6 strains. As in the Vladimir region, ribotypes MI-MI5 were isolated continuously throughout Russia during the study period. In the Vladimir region, ribotypes Gland G4 were first observed in 1991 and 1987, respectively, and their proportion among all ribotypes has been rising continuously since then in both Vladimir region and throughout the Russian federation (figure 3). Since the beginning of the epidemic, 80 (71%) of 115 isolates were ofribotypes Gland G4. In contrast, before the epidemic, only 8 (20%) of 41 isolates belonged to these two ribotypes. ME. Among 156 C. diphtheriae strains, 76 distinct ETs were identified; 6 strains were not assayed. The specific enzyme profiles (reactions) for each ET are available upon request. The genetic relatedness of ETs is represented as a dendrogram in figure 4. Most ETs contained only 1 strain (62 ETs). The remaining 88 isolates were distributed among 14 ETs. The ETs for toxigenic and nontoxigenic strains were different, with the exception ofet 173, which contained 1 toxigenic gravis bio-

4 lid 1996; 174 (November ) Molecular Epidemiology of Diphtheria 1067 Norwegian Sea Russia o National Capital BOOkm BOO Miles Figure 1. Cities and regions in Russia where Corynebacterium diphtheriae isolates were obtained. type strain and I nontoxigenic mitis strain, and ET 22, which contained I nontoxigenic gravis strain and 5 toxigenic gravis strains (table 2, figure 4). Most ETs (75/76) contained only strains of one biotype, either gravis or mitis. Again, the only exception was ET 173, which contained 2 strains, one each of gravis and mitis biotype, but both were of the M13 ribotype. Ribotype MI was associated with 27 different ETs. A significant proportion of strains (42%) was clustered in 13 ETs, within a genetic distance of < Despite the significant diversity of ETs seen throughout the study period, 79 strains clustered within a distinct clonal group containing 14 ETs; since ET 8 predominated, this group was defined as the ET 8 complex (figure 4, table 2). ET 8 contained 41 strains, ET 154 contained 8 strains, and the remaining 12 ETs of this complex contained from I to 4 strains each. All ETs of the ET 8 complex were very similar: 1-4 enzyme differences were observed among individual ETs. The most frequently affected enzymes were NADH diaphorase [ and peptidase II (phenylalanine-leucine). With a single exception (C diphth eriae 720, ribotype M3, biotype mitis), all strains of this complex were of the gravis biotype. Among all other ETs, mitis biotype strains significantly predominated (75%). Also, with a single exception (C diphtheriae G4208, G4 ribotype, gravis biotype), all strains of the ET 8 complex were toxigenic. The proportion of toxigenic strains among all other ETs was 78% (60/77). Of 79 strains within the ET 8 complex, the great majority (71/79) were isolated during the epidemic period, and only 8 strains were isolated before 1990, the first one, dating back to 1986, was isolated in Vladim ir. The remaining 7 isolates were from the Vladimir area (3), Murmansk (2), and Krasnoyarsk (2). The proportion ofthe ET 8 complex strains has been continuously rising in the Vladimir area since 1990, and in 1994 strains ofthis complex made up 67% of all strains. An almost identical trend was apparent in other regions throughout Russia: The proportion of ET 8 complex strains rose from 0% in 1985 to 73% in 1994 (figure 5). Strains of ET 8, ET 22, ET 24, and ET 165 were isolated throughout Russia. Differences were observed regarding the distribution of other ETs of the ET 8 complex. ET 7, ET 20, ET 31, ET 81, ET 122, and ET 124 were seen only in the Central European region of Russia. ET 154 and ET 157 were seen only in the Vladivostok and Magadan regions, which arc miles cast of Vladimir. Comparison of ribotype and MEE data for the strains of the ET 8 complex revealed that 21 (95%) of 22 strains of this complex in the Vladimir area and 56 (98%) of 57 strains isolated throughout Russia were also of Gland G4 ribotypes. The 2 remaining strains within the ET 8 complex were C diphtheriae 720 (M3 ribotype; isolated in Vladimir in 1986) and C diphtheriae 865 (M I ribotype ; isolated in Murmansk in 1994).

5 1068 Popovic et at..tid 1996; 174 (November) A ~ E?C D E FGHIJKLM NO PO R a school cluster showed that they belonged to the same ET and to the same ribotype (table 2). Discussion Figure 2. BstEll ribotypes of 156 Corynebacterium diph theriae strains. Ribotypes by lanes : A, M14, strain 762, isolated in Moscow, 1990; B, 0 I, strain G4174, St. Petersburg, 1993; C, G3, stra in 748, Moscow, 1987; D, G4, strain 749, Vladimir, 1987; E, 04v, strain 722, Vladimir, 1986; F, M5, strain 713, Vladimir, 1985; G, Mlv, strain 1899, Vladivostok, 1994; H, MI, strain 04212, Murmansk, 1992; I, M4, strain 718, Vladimir, 1986; J, M 13, strain 750, Vladimir, 1988; K, M8, strain 724, Vladimir, 1987; L, M7, strain 723, Vlad imir, 1987 ; M, M6, strain 765, Vladimir, 1990; N, MIl, strain 1709, Penza, 1988 ; 0, M12, strain 767, Vladimir, 1990; P, M9, strain 04182, Vladimir, 1992; Q, MIO, strain 760, Vladimir, 1989; R, M3, strain 711, Vladimir, Among the study strains, 12 were from infected individuals in 6 different disease clusters (5 families and a school), possibly representing a direct spread of the same strain (tabl e I). The rem aining 144 strain s were isolated from apparently unrelated diphtheria infections. In 4 of 5 family clusters, families A and E (figure 4) and families Band C (table 2), a single ET group and a single ribotype were identified among epidemiologically linked isolates. In the fifth family cluster, family D, I isolate wa s of ET 155 and ribotype G4 (figure 4), and the other was of ET 154 and ribotype G 1 (table 2). Analysis of isolates from After three decades of excellent control, epidemic diphtheria emerged in Russia. Between 1990 and the end of 1995, almost 100,000 cases were reported. By , epidemic diphtheria had spread to all 15 newly independent states of the former Soviet Union. In addition, at least 20 imported cases of diphtheri a were reported in neighboring and Western European countr ies and in 2 US citizens, and once again, diphtheria is a cause for global concern. Althou gh the reasons for the resurgence of diphtheria in the newly independent states are not fully understood, low vaccine coverage ofinfants and children, waning immunity in adults, and increased population migrations may be important factors [31]. However, biologic characteristics of the causative organism should also be considered. In addition, laboratory characterization and subtyping of C. diphtheriae from outbreak situations will provide invaluable information for epidemiologic studies attempting to determine source s and vehicles of transmission of the organism through the local community, the country, and beyond. Differentiation of C. diphtheriae strains has traditionally involved biotyping, serotyping, and phage typing. Early molecular analyses of C. diphtheriae focused on restriction fragment length polymorphisms and use of specific DNA probes for detection ofthe diphtheria toxin gene (tox), its attachm ent sites, and its insertion elements [16]. How ever, some of these methods are labor-intensive and time-consuming, while others lack discriminatory power. Recently DeZoysa et a1. [32] reported that ribotyping can be used to characterize C. diphtheriae strains isolated in 1993 in northwestern Russia, and they identified two predominant ribotype s. This study significantly extends the molecular characterization of C. diphth eriae isolates No. of i 30 s 0 I 20 a t e 10 s 0 '85 '86 '87 '88 '89 '90 '91 '92 '93 '94 Ribotypes G1 and G4 ~ Other ribotypes Year Figure 3. Yearly distribution of BstElI ribotypes of 156 Corynebacterium diphtheria e isolates in Russia

6 JID 1996; 174 (November) Molecular Epidemiology of Diphtheria 1069 Year City/Region No. ETa complex RT '87 Vladimir 72 3* M7 '85 Vladimir 712 Ml v '90 Vladimir 764* Mlv '94 Vladivostok G4 '86 Penza 1705* M8 '86 Vladimir 718 M4 '85 Vladimir 709 M5 '86 Penza 1702 M4 '86 Penza 1703 M4 '89 Vladimir 760* Ml0 '92 Vladimir 860 Ml '94 Vladivostok 1884 G4, 93 Vladimir 1753 G4 '90 Vladimir 766 Mlv '90 Vladimir 767* M12 '93 Vladimir 864 Mlv '85 Vladimir 710 Mlv '93 Vladimir 175 Mlv '93 Vladimir 862 Mlv '85 Vladimir 708 Mlv,86 Vladimir 715 Ml v ' ** M1v '89 Penza 1710 Mlv, 94 Vladivostok 1901 Mlv '94 Vladivostok 1883 Mlv '94 4** Mlv ' ** Ml '89 Vladimir 757 M1v '91 Moscow G4178 G1 '92 Murmansk G4196* G1 '88 Vladimir 750 M13 '85 Vladimir 714 M1v '86 Vladimir 716 M1v '86 Vladimir 721 M6 '87 Moscow 726 M1v '93 St. Petersburg 489 G3 '90 Vladimir 765 M6 '85 Vladimir 711 M3 '85 Vladimir 713 M5 '93 Murmansk 490 G4 '88 Moscow 753* NT '92 Murmansk G4181 M1 '92 Murmansk G4212 Ml '87 Vladimir 724 M8 '88 Penza 1709 M11 ' ** M13 '91 Penza G4193* M13 '86 Vladimir 719 M13 '89 Vladimir 756* M13 '86 Vladimir 722 G4v '87 Moscow 748* G3 '89 Vladimir 759* M1v '92 Vladimir 859 G4v '91 Vladimir 857* NT '94 2** M1 '90 Vladimir G4192* G4v '90 Penza G4209* G3 '94 Murmansk 866 M1 '94 Murmansk 867 M1v '86 Vladimir 717 M3 '92 Vladimir G4182 M9 '92 Vladimir 858 M3 ET 77:=============]- --, _ =====::J J 73 :========}- -, , _IL 31---_ _ , 79 I- JI J J J J _ , ' J l l :=======]-_ , :============:::Jf "1-1"'11...1"" '1'" ** Year Location No. RT '90 Moscow G4168 Mlv '90 Penza G4170t Mlv '90 Penza G417lt M1v '91 Yaroslavl G4179 M1v '94 Vladivostok 1890 Mlv '94 Vladivostok 1898' Mlv '94 Vladivostok 1899' Mlv '94 Vladivostok 1900 M1v '88 Vladimir 751 Ml '89 Vladimir 755 M1v '86 Penza 1704* M13 '87 Penza 1706 M13 '94 Vladimir 1728* M1 '94 Vladimir 1723 M1 11'""1i'-"'-'I-'I"TI"TI-r["T"['T'"['T'"I"'I"'1"'1Il"'T"1ijl"'1I1"'1I"'TI"TI"TI"TI"'I"ITI"T"''''I'''I'''11""1I""j'Ilr1jl"'1Ir"lI"'I"TI"TI"TI"TI"'I"I"'I'"TII""['Iill'r"lir1I"""T'Tl' GENETIC DISTANCE ET Figure 4. Dendrogram showing genetic relationships of 76 electrophoretic types (ETs) of 156 Corynebacterium diphtheriae isolates from Russia, RT, ribotype; v, variant of either M1 or G4. Details for ET 8 complex are in table 2. * Nontoxigenic C. diphtheriae strain; t family A; H family D; ~ family E; ** multiple isolates described in detail in insert.

7 1070 Popovic et al. JID 1996; 174 (November) Table 2. Corynebacterium diphtheriae strainsof the ET 8 complex isolated in Russia from 1986 to ET RT Year(s) City/region Strain 81 G Vladimir Gl 1993 Vladimir G G Moscow G G Penza G4169 Gl 1992 Moscow G G Murmansk 1744 Gl 1994 Vladimir 1734, 1735, 1740 G Magadan G Krasnoyarsk G4203* 24 G Krasnoyarsk G4204* Gl 1993 Vladimir 1737 Gl 1993 St. Petersburg 485 G Vladimir G Murmansk 1752 G Vladimir 749 G Penza G4190 Gl Vladimir 855,856,861,863, 1727, 1738, 1741, G4172, G4202 G Vladimir 1743 Gl Moscow 1714,1716,1717,1720,1726, 1733, 1736, G4186, G4187, G4188, G4189 Gl 1993 Saratov G4200,t G4201 t G St. Petersburg 484, G4173, G4174, G4175, G4177 G Kaliningrad 1749, 1750, 1754 G Murmansk 1755,1756 G Moscow 1719 Ml 1994 Murmansk 865 G Penza 1713, 1715, G Moscow G4208 Gl 1991 Omsk G4205,:t G4207:t Gl 1992 Murmansk G4210 G Kaliningrad G4211 G Vladimir G G Magadan 1895, Gl 1994 Vladivostok G Vladivostok 1885, 1887, 1889 G Magadan 1891, 1892,1893, Gl 1994 Murmansk 868 G Vladivostok 1903, 1904, M Vladimir G Vladimir 754 G Vladimir 763 NOTE. ET, electrophoretic type; RT, ribotype. *School. Family 'n, 'c, 110. Nontoxigenic strain. associated with the current epidemic. We analyzed strains from throughout Russia by ribotyping and MEE, with a particular focus on the Vladimir region. The surveillance system established there by the Russian authorities in the 1980s led to continuous collection of isolates from that region from 1985 to 1995, allowing us to document expansion of the epidemic clones over a 10-year period. Our study indicates that there is significant genetic diversity within the C. diphtheriae species and that ribotyping and MEE data generally correlate well with biologic properties, such as toxigenicity and biotype. Only two ribotypes (M6 and M13) contained strains of both mitis and gravis biotypes, and one ET (ET 173) was an exception in that it contained both a toxigenic and a nontoxigenic strain of different biotypes. During the pre-epidemic period, a diversity of ribotypes was seen, with the predominant ribotype being MI. Even after 1990, this ribotype has appeared to be present at a continuous level, except in 1994, when a slight increase was observed. In contrast, strains of Gland G4 ribotypes were rarely seen before 1990, but their proportion among all ribotypes began to increase in 1990, and in 1994 they accounted for > 80% of all identified ribotypes. Diversity of C. diphtheriae strains was even greater when the strains were analyzed by MEE: 76 different ETs were observed for Despite this diversity, there was an excellent correlation between MEE, biotyping, and toxigenicity. With a few exceptions, strains of different biotypes and toxigenic properties had different ETs and ribotypes. Comparative analysis of MEE and ribotyping showed that MEE was more discriminatory but that both methods showed general patterns in endemic and epidemic periods. Before the epidemic, strains of the predominant M 1 ribotype were distributed within 27 ETs, forming a cluster with a genetic distance of <0.17. A smaller group of strains of M 13 ribotype formed a cluster with a genetic distance of <0.22. During the epidemic, a significant association of ribotypes and ETs was identified among strains, in particular, association ofribotypes Gland G4 and ETs of the ET 8 complex. This clonal group or complex of C. diphtheriae isolates emerged in Russia at approximately the same time the outbreak began. Only 2 strains of this complex were not of ribotypes Gland G4. This complex now predominates in the Central European region and in the Vladivostok region. Rappuoli et al. [17] suggested that the introduction of a single epidemic strain of C. diphtheriae, which then spread from person to person, led to the diphtheria outbreak in Sweden and that the epidemic strain had some selective advantage, such as increased virulence or enhanced ability to colonize and spread. Among the strains in our study, 1 strain of ET 8 (the predominant ET of the ET 8 complex) was first isolated in 1987 from an I8-year-oldman with diphtheria in the Vladimir region, miles southeast of Moscow. However, strains of this complex were sporadically seen in Vladimir and other regions of the country before In the Vladimir region, 3 additional ET 8 complex strains (ET 122, ET 124, and ET 81) were isolated in 1986 and Over the past 5 years, strains of this complex were identified with increasing frequency, and by 1994 they accounted for 72% of all epidemic strains included in this study. It is possible that the Russian epidemic strain(s) was introduced by travelers, particularly those returning from Afghani-

8 JID 1996; 174 (November) Molecular Epidemiology of Diphtheria ,,...---r-----,-----,-----,----,, , r----,...,.,,.,...,..,..,.,.,..., Figure 5. Yearly distribution of different electrophoretic types (ETs) of 156 Corynebacterium diphtheriae isolates in Russia, No. of 40--lf------lf----r----t----t-----l----lf-----lI----I-----lH 30---l1----r ~,..._---I i s o I a t e S 20--f t l1----H O---lf-.J.:o:w.:.:.::.:.:L L '86 '87 '88 '89 '90 '91 '92 ' I ET8complex DOtherETs Year stan, where diphtheria is endemic. Diphtheria may also have been introduced into Russia in the late 1980s with demobilization of Soviet military forces. Population movements, especially into the large metropolitan areas of Russia since 1990, may have contributed to epidemic disease in these areas. The number of isolates evaluated in this study limits the extent to which we can address issues such as when the outbreak strain was introduced or whether donation of toxin genes played a role in this outbreak. The Swedish epidemic isolates and the 1983 Danish strain [17] belonged to ETs that were clearly distinct from the Russian isolates, suggesting that there was no relationship between the Scandinavian outbreaks of the early 1980s and the current series of outbreaks in the former Soviet Union (data not shown). Incidentally, even though Danish and Swedish isolates were indistinguishable by restriction fragment length polymorphism analyses [17], MEE clearly differentiated these strains. It was suggested, as a result of studies during the diphtheria outbreak in Manchester, United Kingdom, in 1977, that a nontoxigenic strain, which was resident in the community, acquired the toxin gene (corynephage) from a newly introduced toxigenic strain [18]. A similar event may have taken place in Russia. It is also possible that the nontoxigenic isolate from this outbreak may have resulted from the loss of a corynephage or toxin gene by a strain ofthe ET 8 complex. The spontaneous loss of a toxin gene is thought to be unusual [18], but there is abundant documentation that toxigenic and nontoxigenic isolates are often present in individual throat cultures; presumably, some of these represent isogeneic pairs with respect to toxin (phage) [33-37]. During the past 5 years, changes within the epidemic ET 8 complex were apparent. ET 8 strains were the predominant clone within the ET 8 complex in central Russia (Moscow and Vladimir regions) as well as in the Northwest region (St. Petersburg). In Vladivostok, which is situated in the far east of Russia, no ET 8 strains were identified during We can only speculate on the events in that region prior to 1994, since no strains from that time were available to be studied. It may be possible that ET 8 or other ET 8 complex strains evolved into two new ET 8 complex ETs (ET 154, ET 157) that are now typical for the Vladivostok and Magadan regions but have been identified in no other strains from other parts of Russia. Those two ETs differ from ET 8 at only 2 or 3 of 27 assayed enzyme loci. Gradual changes in bacterial populations have been reported for Listeria monocytogenes, another gram-positive rod, with significant ET diversity and changes of ETs over time [38]. In contrast, the Vibrio cholerae strains associated with the recent Latin American epidemic that has affected> 1,000,000 persons was characterized by a single ET that was detected in > 99% of available isolates. To date, ribotyping and MEE have successfully been used to identify clones with increased potential to cause epidemic disease within serogroups of other important bacterial pathogens, such as Salmonella species, Neisseria meningitidis, Streptococcus pneumoniae, and V cholerae 01 [20,22, 39, 40]. In this study, we found a specific ET complex (ET 8) that contains C. diphtheriae isolates of particularribotypes to be significantly associated with the emergence of diphtheria in the Russian federation over the past 3 years. Now that an epidemic complex has been identified, future molecular studies can address differences in toxin and other potential virulence factors within and between ET and ribotype groups. A global collection of C. diphtheriae isolates could provide information about the geographic origin of the outbreak strain(s) and their relation to endemic and epidemic strains isolated worldwide. References I. Hardy IRB, Dittmann S, Sutter RW. Current situation and control strategies for resurgence of diphtheria in newly independent states of the fanner Soviet Union. Lancet 1996; 347: World Health Organization. Expanded Program on Immunization. Outbreak of diphtheria, update. Weekly Epidemiol Rec 1993; 68:

9 1072 Popovic et a!' JID 1996; 174 (November) 3. Markina SS, Maksimova NM, Bogatyureva AJ, Jilina NJ, Kotova EA. Update on diphtheria in Russia, In: Monisov AA, Podunova LG, Tyasto AS, Emeljanov OV, Churchill RE, eds. The health of the population and the environment. Moscow: Russian Federation State Committee on Sanitary Epidemiologic Surveillance, 1993:3-8. April (no. 1). 4. Centers for Disease Control and Prevention. Diphtheria epidemic-new independent states of the former Soviet Union, MMWR Morbid Mortal Weekly Rep 1995;44: Cellesi C, Zanehi A, Michelangeli C, Giovannoni F, Sansoni A, Rossolini GM. Immunity to diphtheria in a sample adult population from central Italy. Vaccine 1989;7: Christenson B, Bottiger M. Serological immunity to diphtheria in Sweden in 1978 and Scand J Infect Dis 1986; 18: Crossley K, Irvine P, Warren lb, Lee BK, Mead K. Tetanus and diphtheria immunity in urban Minnesota adults. JAMA 1979;242: Koblin BA, Townsend TR. Immunity to diphtheria and tetanus in innercity women of child-bearing age. Am J Public Health 1989;79: Hennekens CH, Saslaw MS. A diphtheria outbreak in Dade County, Florida. South Med J 1976;69: Popovic T, Wharton M, Wenger J, McIntyre L, Wachsmuth IK. Are we ready for diphtheria? A report from the Diphtheria Diagnostic Workshop, Atlanta, 11 and 12 July J Infect Dis 1995; 171: Simmonsen 0, Kjeldsen K, Bentzon MW, Heron 1. Susceptibility to diphtheria in populations vaccinated before and after elimination of indigenous diphtheria in Denmark. Acta Pathol Microbiol Immunol Scand [C] 1987; 95: Hallander HO, Haeggman H, Lofdahl S. Epidemiological typing of Corynebacterium diphtheriae isolated in Sweden Scand J Infect Dis 1988; 20: Bobkova MR, Kombarova SI, Lipis SV, Bobkova AF, Mazurova IK. The use of DNA fingerprint analysis for the differentiation of populations oftoxigenic Corynebacterium diphtheriae. Zh Mikrobiol Epidemiol Immunobiol 1989; 7: Efstratiou A, Tiley SM, Sangrador A, et a1. Invasive disease caused by multiple clones of Corynebacterium diphtheriae [letter]. Clin Infect Dis 1993; 17: Efstratiou A, George RC, Begg NT. Non-toxigenic Corynebacterium diphtheriae val' gravis in England [letter]. Lancet 1993; 341: Rappuoli R, Perugini M, Ratti G. DNA element of Corynebacterium diphtheriae with properties of an insertion sequence and usefulness for epidemiological studies. J Bacteriol 1987; 169: Rappuoli R, Pergini M, Falsen E. Molecular epidemiology of the outbreak of diphtheria in Sweden. N Engl J Med 1988;318: Pappenheimer AM, Murphy JR. Studies on the molecular epidemiology ofdiphtheria. Lancet 1983;2: Coyle MB, Groman NB, Russell JQ, Harnisch JP, Rabin M, Holmes KK. The molecular epidemiology of three biotypes of Corynebacterium diphtheriae in the Seattle outbreak, J Infect Dis 1989;159: Reeves MW, Evins GM, Heiba AA, Plikaytis BD, Fanner 11 III. Clonal nature ofsalmonella typhi and its genetic relatedness to other salmonellae as shown by multilocus enzyme electrophoresis, and proposal of Salmonella bongori comb. nov. J Clin MicrobioI1989;27: Selander RK, Caugant DA, Ochman H, Musser JM, Gilmore MN, Whittam TS. Methods of multilocus enzyme electrophoresis for bacterial population genetics and systematics. Appl Environ Microbiol 1986;51: Wachsmuth IK, Evins GM, Fields PI, et al. The molecular epidemiology of cholera in Latin America. J Infect Dis 1993;167: Mazurova, IK, Melnikov VG, Kombarova S'Y,Manual for laboratory diagnostic of diphtheria infection. Moscow: Russian Federation State Committee on Sanitary Epidemiologic Surveillance, Russian Ministry of Health. Bacteriological research in diphtheria infection. Moscow: Russian Ministry of Health, Efstratiou A, Maple PA. WHO manual for the laboratory diagnosis of diphtheria. Geneva: World Health Organization, 1994, reference no. ICP-EPI 038(C). 26. Pallen MJ. Rapid screening for toxigenic Corynebacterium diphtheriae by the polymerase chain reaction. J Clin PathoI1991;44: Mikhailovich V, Melnikov V, Mazurova I, et al. Application of PCR for detection of toxigenic Corynebacterium diphtheriae strains isolated during the Russian diphtheria epidemic, 1990 through J Clin MicrobioI1995;33: Popovic T, Bopp C, Olsvik 0, Wachsmuth K. Epidemiologic application of a standardized ribotype scheme for Vibrio cholerae 01. J Clin Microbiol 1993;31: Sambrook JE, Fritsch EF, Maniatis T. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, Jacobs D. SAS/GRAPH software and numerical taxonomy. In: Proceeding of the 15th annual Users Group International Conference. Cary, NC: SAS Institute, 1990: Galazka AM, Robertson SE, Oblapenko GP. Resurgence of diphtheria. Eur J Epidemiol 1995; II : DeZoysa A, Efstratiou A, George R, et a1. Molecular epidemiology of Corynebacterium diphtheriae from northwestern Russia and surrounding countries studied by using ribotyping and pulsed-field gel electrophoresis. J Clin Microbiol 1995;33: Simmons LE, Abbott JD, Macaulay ME, et al. Diphtheria carriers in Manchester: simultaneous infection with toxigenic and nontoxigenic mitis strains. Lancet 1980; 1: Chang ON, Laughren GS, Chalvardjian NE. Three variants of Corynebacterium diphtheriae subsp. mitis (belfanti) isolated from a throat specimen. J Clin Microbiol 1978; 8: Freeman VI. Studies on the virulence of bacteriophage-infected strains of Corynebacterium diphtheriae. J Bacteriol 1951;61: Okell CC. The relationship of virulent to avirulent diphtheria bacilli. J Hyg 1929;29: Moore PS, Reeves MW, Schwartz B, Gellin BG, Broome CV. Intercontinental spread of an epidemic group A Neisseria meningitidis strain. Lancet 1989;2: Bibb WF, Gellin BG, Weaver RE, et al. Analysis of clinical and foodborne isolates of Listeria monocytogenes in the United States by multilocus enzyme electrophoresis and application of the method to epidemiologic investigations. Appl Environ Microbio11990;56: Wang JF, Caugant DA, Morelli G, Koumare, Achtman M. Antigenic and epidemiologic properties of the ET-37 complex of Neisseria meningitidis. J Infect Dis 1993; 167: Versalovic J, Kapur V, Mason EO Jr, et a1. Penicillin-resistant Streptococcus pneumoniae strains recovered in Houston: identification and molecular characterization of multiple clones. J Infect Dis 1993; 167:850-6.