A convenience sample of 47 Corynebacterium diphtheriae isolates was available for analysis from a collection of isolates obtained during 1957-1987, before the onset of the recent diphtheria outbreak. These isolates were collected from both carriers (n=37) and patients (n=10) in different regions of Russia. All isolates were kept freeze-dried at the G. N. Gabrichevsky Institute for Epidemiology and Microbiology, Moscow, Russia, and were transported on silica gel packages to the Centers for Disease Control and Prevention, Atlanta, Georgia, for molecular characterization.

All isolates were biotyped by using the commercial API Coryne kit (Biomerieux, Lyon, France). Toxigenicity status was determined by the Elek test, as recommended by WHO (9), and by the polymerase chain reaction (PCR), which targeted both A and B subunits of the tox gene (10).

All the strains were characterized by ribotyping as previously described (11). The hybridization was done by using five oligonucleotide probes according to Regnault et al. (12). Ribotyping pattern designations were based on the scheme established by Popovic et al. (7). A difference in one band was defined as an individual ribotype (RT).

MEE was carried out as previously described (7,11). The electromorphs of the same enzyme were visualized in a starch gel matrix as bands of different migration rates. Each electromorph was considered to represent a distinct allele of the same enzyme. By testing 27 different enzymes, a profile of electromorphs, defining the electrophoretic type (ET) of each strain, was obtained. The genetic relatedness of ETs was illustrated as a dendrogram, which was generated by the average-linkage method of clustering ETs described by Selander et al. (13).

We examined 47 C. diphtheriae isolates collected in the pre-epidemic period (1957-1987) from 10 patients and 37 carriers in different areas of Russia. Thirty-nine strains were of the gravis biotype, 7 were the mitis biotype, and 1 was of the intermedius biotype. All the mitis biotype strains were toxigenic. Among the gravis biotype strains, 36 were toxigenic, and 3 were nontoxigenic. No discrepancies between the results obtained by traditional identification, the API Coryne test, or toxigenicity testing by the Elek test and PCR were detected.

Figure 1

Figure 1. Twelve BstEII ribotypes identified in 47 Corynebacterium diphtheriae isolates collected in the Russian Federation between 1957 and 1987. The figure is composed of ribotype gels exemplifying the different patterns observed in...

In the 47 isolates, 12 different RTs were identified (Figure 1). Twenty-two (47%) were of the M11e RT; all were toxigenic and of the gravis biotype. They were collected from 1957 to 1985. RT G4, characteristically seen in the recent epidemic clonal group, was identified in 6 (13%) isolates, all of which were collected from 1984 through 1987. Four isolates had two new ribotype patterns, not previously described. They were collected from 1977 through 1981. RT G4v, a variant of the G4 RT, was identified in a single strain isolated in 1977. Three of the isolates (all toxigenic and of the mitis biotype) had a unique RT not previously described; they were collected from 1977 through 1981.

Figure 2

Figure 2. Dendrogram showing the genetic relatedness of 85 electrophoretic types (ETs) of Corynebacterium diphtheriae isolates collected in different countries around the world. Arrows indicate the different ETs identified among the 47 C....

Sixteen (6 isolates of RT pattern G4 and 10 isolates of different RT patterns) of the 47 isolates were analyzed by MEE; 13 different ETs were identified. Of the six isolates with the G4 patterns, four also belonged to the ET8 complex (Figure 2). An additional isolate (strain designation B533 in Table) collected in 1957 belonged to the ET8 complex but had a different RT.

In the pre- and early vaccine era, diphtheria incidence was high in the Soviet Union. After the diphtheria vaccine was introduced, a decrease in incidence was seen in the 1950s. During the mid-1970s, immunization programs resulted in control of diphtheria throughout the country. However, an increase in incidence was noted at the end of 1970 and during the 1980s, and a peak was observed in 1983. This resurgence was associated with a change in the biotype of the circulating C. diphtheriae strains from gravis, which had been dominating for several decades, to mitis (14).

To allow better monitoring of the global spread of diphtheria, the WHO ribotyping database for C. diphtheriae was established at the Pasteur Institute in Paris, France. The institute demonstrated that C. diphtheriae RTs are quite diverse worldwide but remain stable over time (15). Both ribotyping and MEE have provided a significant level of differentiation and reliability and subsequently have been accepted as the standard for molecular subtyping of C. diphtheriae. Thus, we used these molecular methods to characterize our archival isolates.

Twelve different RTs were found in our 47 isolates. Our data show that nine C. diphtheriae isolates from the 1950s and 1960s had an RT pattern (M11e) that was very similar to ribotype M11, which was only seen occasionally in the FSU in the 1990s. Epidemic RT G4 was seen in six toxigenic C. diphtheriae isolates collected from 1984 through 1987 in four distant regions of Russia (Moscow and Moscow region, Anapa, Smolensk, and Sverdlovsk) from both diphtheria patients and carriers; four of these isolates were also members of the ET8 complex.

Our investigation of the origin of the epidemic clonal group determined that, in our strain collection, the earliest reported strain of this clonal group was identified in Smolensk in 1985, and that strains of this clonal group were simultaneously present in several geographically distant areas in Russia from 1985 through 1987. These findings suggest that the current epidemic clone was an integral part of the endemic reservoir that existed in the FSU at least 5 years before the epidemic began. Further studies that would include a large number of gravis biotype strains from throughout the Soviet Union isolated from 1980 through 1985 might unveil where and when strains of the epidemic clone were first associated with disease or carriage.