JOURNAL OF CLINICAL MICROBIOLOGY, Jan. 1991,
p.
Vol. 29, No. 1
27-32
0095-1137/91/010027-06$02.00/0 Copyright © 1991, American Society for Microbiology
Molecular Analysis of Variant Plasmid Forms of a Bivalent Salmonella typhi-Shigella sonnei Vaccine Strain ANTOINETTE B. HARTMAN,'* MARIA M. RUIZ,2 AND CLYDE L. SCHULTZ' Department of Biologics Research' and Department of Cellular Immunology,2 Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, Washington, D.C. 20307-5100 Received 8 June 1990/Accepted 11 October 1990
The Salmonella typhi-Shigella sonnei hybrid vaccine strain 5076-1C was constructed to express the S. sonnei form I antigen, which may play an important role in producing protective immunity. Three clonal variants which existed in preparations of the vaccine could be distinguished phenotypically by lactose utilization, S. sonnei form I antigen expression, and restriction enzyme analysis of large plasmid DNA. Since expression of the form I antigen was lost in two of the clonal variants, genetic instability of the 120-MDa vaccine plasmid appeared to be a potential problem. To examine the molecular basis of this genetic instability, we hybridized large plasmid DNA isolated from the clonal variants to a variety of DNA probes encoding virulence-associated antigens and to Escherichia coli lacZ DNA. Results indicated that DNA rearrangements accompanied by deletions of plasmid material occurred in the vaccine plasmid. In addition, the vaccine plasmid did not contain some S. sonnei genetic material encoding antigenic polypeptides necessary for virulence. construct, an analysis of the molecular changes that had occurred in the vaccine plasmid might elucidate some of the potential problems involved in designing vaccine constructs. We have recently reported that ipaH gene sequences are found in multiple copies on the Shigella invasion plasmid (12); it has also been noted that changes in ipaH hybridization patterns often accompany rearrangements and/or a loss of plasmid material (4a). ipaH DNA was therefore used to probe large plasmid DNA isolated from the clonal variants to detect any DNA rearrangements. On the basis of these results, additional probes were used to determine whether genetic material encoding other virulence-associated antigens was absent from the hybrid vaccine plasmid and whether the rearrangements observed in this plasmid involved deletions of any genetic material.
Shigellosis is a significant cause of serious gastrointestinal illness in the world, both in developing countries and in industrialized nations. Although parenteral vaccines of killed Shigella organisms have not proved effective (9, 11, 13), some attenuated Shigella vaccine strains administered orally have provided serotype-specific protection against shigellosis in humans and monkeys in laboratory and field trials (3, 7, 8, 21, 22, 31). However, various problems, such as residual virulence, genetic instability, and varying immunogenicity, have prevented the widespread use of these vaccines (18-20, 26). Some evidence has indicated that the serotype-specific somatic antigens may play an important role in the immunity conferred by the attenuated Shigella vaccine strains (7, 20, 23). The bivalent Salmonella typhi-Shigella sonnei vaccine strain 5076-1C was constructed by conjugal transfer of the large invasion plasmid of S. sonnei 53G, which encodes the form I somatic antigen found in virulent bacteria, into the galE S. typhi Ty2la oral vaccine strain by use of the mobilizing plasmid F' lac(Ts)::Tn3 (6). It has been demonstrated that this hybrid strain expresses the somatic antigens from S. typhi and the form I somatic antigen from S. sonnei (6). Although two different lots of the vaccine strain (lots 2 and 5) conferred significant protection in volunteers against a homologous challenge, a lot prepared for a large-scale field trial (lot 8) was not protective (3). Additional tests suggested that a lack of observable pili and Salmonella flagellar H antigen in the lot 8 vaccine organisms might have been responsible for the failure of this vaccine lot to protect volunteers (26). During testing of the different vaccine lots, it was determined that three different clonal types, defined by restriction enzyme analysis of the corresponding large plasmids, could be found in each vaccine lot (26). Two of the variants appeared to have undergone rearrangements in the large plasmid DNA which were correlated with the loss of expression of the S. sonnei form I somatic antigen. Since genetic instability can seriously reduce the usefulness of a vaccine *
MATERIALS AND METHODS
Bacterial strains and growth conditions. As described previously (6), the S. typhi-S. sonnei hybrid vaccine strain 5076-1C was constructed by conjugal transfer of the large invasion plasmid of S. sonnei 53G into S. typhi Ty2la by use of an F' lac(Ts)::Tn3 mobilizing plasmid. The large plasmid (120 MDa) found in this vaccine strain, although derived from the invasion plasmid of S. sonnei 53G, does not produce the invasive phenotype in the hybrid vaccine strain. Different lots of the S. typhi-S. sonnei hybrid vaccine strain 5076-1C were grown, harvested, and freeze-dried at the Department of Biologics Research, Walter Reed Army Institute of Research, Washington, D.C., as described previously (26). Vaccine lots 2, 5, and 8 were tested in human volunteers; lots 2 and 5 provided significant protection against homologous challenge, while lot 8 did not provide protection (3). Lot 10 was never released for human studies. Individual colonies from lots 2, 5, 8, and 10 of the S. typhi-S. sonnei hybrid vaccine strain 5076-1C were obtained by rehydrating lyophilized vials of the vaccine lots prepared at the Walter Reed Army Institute of Research and streaking the material onto MacConkey agar (Difco Laboratories, Detroit, Mich.) to determine colonies expressing the Lac' phenotype, which was used as a marker for the presence of
Corresponding author. 27
28
J. CLIN. MICROBIOL.
HARTMAN ET AL.
the conjugally transferred S. sonnei 53G invasion plasmid. After incubation at 37°C for 28 to 36 h, lactose-positive and -negative colonies were picked, inoculated into brain heart infusion medium (Difco), and grown overnight at 37°C. S. sonnei 53G, obtained from the Walter Reed Army Institute of Research collection, was streaked onto brain heart infusion agar plates. After overnight incubation at 37°C, form I colonies were selected from these plates and used to inoculate brain heart infusion medium for overnight growth at 370C. Plasmid preparations and DNA hybridizations. Small-scale preparations of large plasmid DNA from S. sonnei 53G and the vaccine strain were obtained by a modification of the procedure of Cassie et al. (5). Following the removal of the bacterial chromosomal DNA by salt precipitation, the plasmid DNA was extracted with phenol, extracted twice with chloroform, and ethanol precipitated. Restriction enzyme digestions of large plasmid DNA, transfer to nitrocellulose of the DNA fragments following agarose gel electrophoresis on 0.6% agarose gels, and hybridization conditions have been described previously (12). The ipaH probe used in these studies was obtained from plasmid pWR391 and contained a 2.9-kb EcoRI fragment that encodes the ipaH structural gene cloned into pBR322 (12). We also used (i) an ipaBCD probe excised from plasmid pHC17 and containing a 4.7-kb HindIII fragment cloned into pUC18 (29) and (ii) a virG probe containing a 6.1-kb EcoRI-SalI fragment cloned into pUC18 (pHS3192, obtained from L. Hale). For the virG probe, we excised a 1.2-kb HindIII-EcoRV fragment containing a significant portion of the structural gene (17). The lacZ probe was prepared by excising the E. coli lacZ gene from pUC19 with BglI and EcoRI. The S. sonnei form I antigen gene was obtained from pHS2-14 (obtained from D. Kopecko). This clone synthesizes the form I antigen, as determined by a serological assay, and contains three HindlIl fragments (11 kb, 2.1 kb, and 900 bp) cloned into pHC79; the 11-kb fragment is responsible for form I antigen synthesis (15a). The 11-kb insert as well as the entire pHS2-14 plasmid was used to probe large plasmid DNA from the three clonal variants. As a control for the whole plasmid probe, pBR322 DNA was also hybridized to vaccine plasmid DNA. Serological studies. Qualitative determinations for the presence of the form I somatic antigen were made by slide agglutination assays with serotype-specific rabbit antisera as described previously (6, 26). RESULTS
Three clonal types found in vaccine lots. As reported
previously (26), three clonal types were isolated from lyophilized vials of vaccine strain 5076-1C, lots 2, 5, and 8. These variants differed in both restriction fragment patterns of enzyme-digested large plasmid DNA (Fig. 1) and expression of the form I antigen and the Lac phenotype (Table 1). Examination of the large plasmid restriction profiles of each clonal type indicated that DNA rearrangements had occurred in the variants (Fig. 1). Comparisons of phenotypic traits for each type (Table 1) suggested that these rearrangements involved changes in expression of the Lac phenotype and the form I antigen. Correlation of variant type with form I antigen production was also demonstrated when analyses of the predominant variant type and form I antigen production in different vaccine lots were done. Large plasmid DNA from 10 to 15 randomly selected lactose-positive colonies from lots 2, 5, 8, and 10 was analyzed by restriction enzyme digestion; the colonies were also scored for form I antigen
1
-
2kb
!
9.4 -6.6
2
4.4 --
2.3
--
2.0
-
2
]
1_
i
FIG. 1. Restriction enzyme patterns of plasmid DNA isolated from individual colonies of three clonal types of vaccine strain 5076-1C (described in Table 1). Plasmid DNA samples were digested with BglII, electrophoresed on 0.6% agarose gels, and stained with ethidium bromide. Lanes: 1, type 3 plasmid DNA; 2, type 2 plasmid DNA; 3, type 1 plasmid DNA. Lambda HindIII DNA molecular size standards (unmarked lane) are indicated at the left (in kilobases).
production. In addition, form I antigen production of different vaccine lots was assayed by analysis of lyophilized vials of each lot. Results from screening of randomly selected individual colonies from lots 2, 5, and 8 indicated that these lots contained >90% type 1 organisms; agglutination tests performed on lyophilized vials of these vaccine lots produced a strong reaction against form I antisera. In contrast, lot 10, which contained >70% type 2 organisms, did not produce a strong agglutination reaction against form I antisera and was not released for human trials (unpublished results). ipaH hybridization patterns of clonal types 1 and 2. Probing of invasion plasmid DNA isolated from Shigella strains with the ipaH gene segment produced multiple hybridizing bands; variations in ipaH hybridization patterns often accompanied plasmid DNA rearrangements (4a, 12). Since restriction fragment analysis had indicated that DNA rearrangements might be occurring in the vaccine plasmid, large plasmid DNA isolated from the clonal variants was digested with a variety of restriction enzymes, subjected to agarose gel electrophoresis, transferred to nitrocellulose, and probed with ipaH DNA to determine whether plasmid rearrangements could be detected. Only two of the clonal variants showed hybridization to the ipaH probe; type 3 variants apparently had deletions of ipaH-hybridizing DNA. ipaH TABLE 1. Characteristics of clonal types of vaccine strain
5076-1C Clonaltypea
Lac
phenotype
1 2
+ +
3
-
ipaH
Form I antigen production
hybridization
+
+ +
-
_
a As determined by restriction enzyme analysis of plasmid DNA from each colony with BglII.
PLASMID VARIATIONS IN VACCINE STRAIN 5076-1C
VOL. 29 1991
kb
1
2
3
4
5
6
7
8
29
10 1 112 13 14
9
23 -9.4 -
miag
-iiiSg~~~~~
6.6 4.4 r.(
_
U
4
"
_b
_
2.3 2.0 -
FIG. 2. Southern blot analysis of plasmid DNA isolated from clonal types 1 and 2 hybridized with the ipaH probe. Lanes 1, 3, 5, 7, 9, 11, and 13 represent plasmid DNA from type 2; lanes 2, 4, 6, 8, 10, 12, and 14 represent plasmid DNA from type 1. Plasmid DNA digested with BglII (lanes 1 and 2), BglII and Sacl (lanes 3 and 4), Sacl (lanes 5 and 6), KpnI and Sacl (lanes 7 and 8), KpnI (lanes 9 and 10), Sacl and XbaI (lanes 11 and 12), and XbaI (lanes 13 and 14) was electrophoresed on 0.6% agarose gels, transferred to nylon filters, and hybridized to ipaH DNA. Lambda HindIll molecular size markers are shown at the left (in kilobases).
hybridization patterns obtained when plasmid DNA from type 1 and 2 variants was digested with BglII, BgIII and Sacl, Sacl, and XbaI indicated that DNA rearrangements had occurred within these restriction sites (Fig. 2). On the other hand, DNA rearrangements had occurred outside the KpnI and HindIII restriction sites, since no restriction fragment size changes were observed in the ipaH-hybridizing bands when plasmid DNA from type 1 and 2 variants was digested with these enzymes (Fig. 2; see Fig. SA). Hybridization of lacZ and form I antigen probes to large plasmid DNA from clonal variants. ipaH hybridization patterns from the three clonal types indicated that plasmid rearrangements and, in the case of clonal variant 3, deletions of plasmid material had occurred in the vaccine plasmid. To determine whether changes in the expression of the Lac phenotype and the form I somatic antigen were due to DNA rearrangements or to deletion of the locus responsible for the expression of the phenotype, we used the E. coli lacZ gene segment and DNA expressing the form I somatic antigen of S. sonnei to probe large plasmid DNA from the clonal variants. The lacZ gene probe only hybridized to plasmid DNA from clonal types 1 and 2 (Fig. 3), indicating that the lac locus had been deleted from the large plasmid of type 3 variants. When the 11-kb fragment from recombinant clone pHS2-14, which expresses the form I antigen, was used to probe HindIII-digested S. sonnei 53G invasion plasmid DNA, a strongly hybridizing band at 11 kb and a more weakly hybridizing band at 6.5 kb were observed (Fig. 4). When HindIII-digested large plasmid DNA from the three clonal variants was hybridized to this probe, type 1 DNA displayed the S. sonnei hybridization pattern, type 2 DNA only hybridized to the weaker 6.5-kb band, and type 3 DNA did not hybridize at all (Fig. 4A). Hindlll-digested type 1 DNA also hybridized to 2.1-kb and 900-bp fragments when
2
3
4
23 9.4 6.64.4-
2.0 -
-
p.
Vol. 29, No. 1
27-32
0095-1137/91/010027-06$02.00/0 Copyright © 1991, American Society for Microbiology
Molecular Analysis of Variant Plasmid Forms of a Bivalent Salmonella typhi-Shigella sonnei Vaccine Strain ANTOINETTE B. HARTMAN,'* MARIA M. RUIZ,2 AND CLYDE L. SCHULTZ' Department of Biologics Research' and Department of Cellular Immunology,2 Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, Washington, D.C. 20307-5100 Received 8 June 1990/Accepted 11 October 1990
The Salmonella typhi-Shigella sonnei hybrid vaccine strain 5076-1C was constructed to express the S. sonnei form I antigen, which may play an important role in producing protective immunity. Three clonal variants which existed in preparations of the vaccine could be distinguished phenotypically by lactose utilization, S. sonnei form I antigen expression, and restriction enzyme analysis of large plasmid DNA. Since expression of the form I antigen was lost in two of the clonal variants, genetic instability of the 120-MDa vaccine plasmid appeared to be a potential problem. To examine the molecular basis of this genetic instability, we hybridized large plasmid DNA isolated from the clonal variants to a variety of DNA probes encoding virulence-associated antigens and to Escherichia coli lacZ DNA. Results indicated that DNA rearrangements accompanied by deletions of plasmid material occurred in the vaccine plasmid. In addition, the vaccine plasmid did not contain some S. sonnei genetic material encoding antigenic polypeptides necessary for virulence. construct, an analysis of the molecular changes that had occurred in the vaccine plasmid might elucidate some of the potential problems involved in designing vaccine constructs. We have recently reported that ipaH gene sequences are found in multiple copies on the Shigella invasion plasmid (12); it has also been noted that changes in ipaH hybridization patterns often accompany rearrangements and/or a loss of plasmid material (4a). ipaH DNA was therefore used to probe large plasmid DNA isolated from the clonal variants to detect any DNA rearrangements. On the basis of these results, additional probes were used to determine whether genetic material encoding other virulence-associated antigens was absent from the hybrid vaccine plasmid and whether the rearrangements observed in this plasmid involved deletions of any genetic material.
Shigellosis is a significant cause of serious gastrointestinal illness in the world, both in developing countries and in industrialized nations. Although parenteral vaccines of killed Shigella organisms have not proved effective (9, 11, 13), some attenuated Shigella vaccine strains administered orally have provided serotype-specific protection against shigellosis in humans and monkeys in laboratory and field trials (3, 7, 8, 21, 22, 31). However, various problems, such as residual virulence, genetic instability, and varying immunogenicity, have prevented the widespread use of these vaccines (18-20, 26). Some evidence has indicated that the serotype-specific somatic antigens may play an important role in the immunity conferred by the attenuated Shigella vaccine strains (7, 20, 23). The bivalent Salmonella typhi-Shigella sonnei vaccine strain 5076-1C was constructed by conjugal transfer of the large invasion plasmid of S. sonnei 53G, which encodes the form I somatic antigen found in virulent bacteria, into the galE S. typhi Ty2la oral vaccine strain by use of the mobilizing plasmid F' lac(Ts)::Tn3 (6). It has been demonstrated that this hybrid strain expresses the somatic antigens from S. typhi and the form I somatic antigen from S. sonnei (6). Although two different lots of the vaccine strain (lots 2 and 5) conferred significant protection in volunteers against a homologous challenge, a lot prepared for a large-scale field trial (lot 8) was not protective (3). Additional tests suggested that a lack of observable pili and Salmonella flagellar H antigen in the lot 8 vaccine organisms might have been responsible for the failure of this vaccine lot to protect volunteers (26). During testing of the different vaccine lots, it was determined that three different clonal types, defined by restriction enzyme analysis of the corresponding large plasmids, could be found in each vaccine lot (26). Two of the variants appeared to have undergone rearrangements in the large plasmid DNA which were correlated with the loss of expression of the S. sonnei form I somatic antigen. Since genetic instability can seriously reduce the usefulness of a vaccine *
MATERIALS AND METHODS
Bacterial strains and growth conditions. As described previously (6), the S. typhi-S. sonnei hybrid vaccine strain 5076-1C was constructed by conjugal transfer of the large invasion plasmid of S. sonnei 53G into S. typhi Ty2la by use of an F' lac(Ts)::Tn3 mobilizing plasmid. The large plasmid (120 MDa) found in this vaccine strain, although derived from the invasion plasmid of S. sonnei 53G, does not produce the invasive phenotype in the hybrid vaccine strain. Different lots of the S. typhi-S. sonnei hybrid vaccine strain 5076-1C were grown, harvested, and freeze-dried at the Department of Biologics Research, Walter Reed Army Institute of Research, Washington, D.C., as described previously (26). Vaccine lots 2, 5, and 8 were tested in human volunteers; lots 2 and 5 provided significant protection against homologous challenge, while lot 8 did not provide protection (3). Lot 10 was never released for human studies. Individual colonies from lots 2, 5, 8, and 10 of the S. typhi-S. sonnei hybrid vaccine strain 5076-1C were obtained by rehydrating lyophilized vials of the vaccine lots prepared at the Walter Reed Army Institute of Research and streaking the material onto MacConkey agar (Difco Laboratories, Detroit, Mich.) to determine colonies expressing the Lac' phenotype, which was used as a marker for the presence of
Corresponding author. 27
28
J. CLIN. MICROBIOL.
HARTMAN ET AL.
the conjugally transferred S. sonnei 53G invasion plasmid. After incubation at 37°C for 28 to 36 h, lactose-positive and -negative colonies were picked, inoculated into brain heart infusion medium (Difco), and grown overnight at 37°C. S. sonnei 53G, obtained from the Walter Reed Army Institute of Research collection, was streaked onto brain heart infusion agar plates. After overnight incubation at 37°C, form I colonies were selected from these plates and used to inoculate brain heart infusion medium for overnight growth at 370C. Plasmid preparations and DNA hybridizations. Small-scale preparations of large plasmid DNA from S. sonnei 53G and the vaccine strain were obtained by a modification of the procedure of Cassie et al. (5). Following the removal of the bacterial chromosomal DNA by salt precipitation, the plasmid DNA was extracted with phenol, extracted twice with chloroform, and ethanol precipitated. Restriction enzyme digestions of large plasmid DNA, transfer to nitrocellulose of the DNA fragments following agarose gel electrophoresis on 0.6% agarose gels, and hybridization conditions have been described previously (12). The ipaH probe used in these studies was obtained from plasmid pWR391 and contained a 2.9-kb EcoRI fragment that encodes the ipaH structural gene cloned into pBR322 (12). We also used (i) an ipaBCD probe excised from plasmid pHC17 and containing a 4.7-kb HindIII fragment cloned into pUC18 (29) and (ii) a virG probe containing a 6.1-kb EcoRI-SalI fragment cloned into pUC18 (pHS3192, obtained from L. Hale). For the virG probe, we excised a 1.2-kb HindIII-EcoRV fragment containing a significant portion of the structural gene (17). The lacZ probe was prepared by excising the E. coli lacZ gene from pUC19 with BglI and EcoRI. The S. sonnei form I antigen gene was obtained from pHS2-14 (obtained from D. Kopecko). This clone synthesizes the form I antigen, as determined by a serological assay, and contains three HindlIl fragments (11 kb, 2.1 kb, and 900 bp) cloned into pHC79; the 11-kb fragment is responsible for form I antigen synthesis (15a). The 11-kb insert as well as the entire pHS2-14 plasmid was used to probe large plasmid DNA from the three clonal variants. As a control for the whole plasmid probe, pBR322 DNA was also hybridized to vaccine plasmid DNA. Serological studies. Qualitative determinations for the presence of the form I somatic antigen were made by slide agglutination assays with serotype-specific rabbit antisera as described previously (6, 26). RESULTS
Three clonal types found in vaccine lots. As reported
previously (26), three clonal types were isolated from lyophilized vials of vaccine strain 5076-1C, lots 2, 5, and 8. These variants differed in both restriction fragment patterns of enzyme-digested large plasmid DNA (Fig. 1) and expression of the form I antigen and the Lac phenotype (Table 1). Examination of the large plasmid restriction profiles of each clonal type indicated that DNA rearrangements had occurred in the variants (Fig. 1). Comparisons of phenotypic traits for each type (Table 1) suggested that these rearrangements involved changes in expression of the Lac phenotype and the form I antigen. Correlation of variant type with form I antigen production was also demonstrated when analyses of the predominant variant type and form I antigen production in different vaccine lots were done. Large plasmid DNA from 10 to 15 randomly selected lactose-positive colonies from lots 2, 5, 8, and 10 was analyzed by restriction enzyme digestion; the colonies were also scored for form I antigen
1
-
2kb
!
9.4 -6.6
2
4.4 --
2.3
--
2.0
-
2
]
1_
i
FIG. 1. Restriction enzyme patterns of plasmid DNA isolated from individual colonies of three clonal types of vaccine strain 5076-1C (described in Table 1). Plasmid DNA samples were digested with BglII, electrophoresed on 0.6% agarose gels, and stained with ethidium bromide. Lanes: 1, type 3 plasmid DNA; 2, type 2 plasmid DNA; 3, type 1 plasmid DNA. Lambda HindIII DNA molecular size standards (unmarked lane) are indicated at the left (in kilobases).
production. In addition, form I antigen production of different vaccine lots was assayed by analysis of lyophilized vials of each lot. Results from screening of randomly selected individual colonies from lots 2, 5, and 8 indicated that these lots contained >90% type 1 organisms; agglutination tests performed on lyophilized vials of these vaccine lots produced a strong reaction against form I antisera. In contrast, lot 10, which contained >70% type 2 organisms, did not produce a strong agglutination reaction against form I antisera and was not released for human trials (unpublished results). ipaH hybridization patterns of clonal types 1 and 2. Probing of invasion plasmid DNA isolated from Shigella strains with the ipaH gene segment produced multiple hybridizing bands; variations in ipaH hybridization patterns often accompanied plasmid DNA rearrangements (4a, 12). Since restriction fragment analysis had indicated that DNA rearrangements might be occurring in the vaccine plasmid, large plasmid DNA isolated from the clonal variants was digested with a variety of restriction enzymes, subjected to agarose gel electrophoresis, transferred to nitrocellulose, and probed with ipaH DNA to determine whether plasmid rearrangements could be detected. Only two of the clonal variants showed hybridization to the ipaH probe; type 3 variants apparently had deletions of ipaH-hybridizing DNA. ipaH TABLE 1. Characteristics of clonal types of vaccine strain
5076-1C Clonaltypea
Lac
phenotype
1 2
+ +
3
-
ipaH
Form I antigen production
hybridization
+
+ +
-
_
a As determined by restriction enzyme analysis of plasmid DNA from each colony with BglII.
PLASMID VARIATIONS IN VACCINE STRAIN 5076-1C
VOL. 29 1991
kb
1
2
3
4
5
6
7
8
29
10 1 112 13 14
9
23 -9.4 -
miag
-iiiSg~~~~~
6.6 4.4 r.(
_
U
4
"
_b
_
2.3 2.0 -
FIG. 2. Southern blot analysis of plasmid DNA isolated from clonal types 1 and 2 hybridized with the ipaH probe. Lanes 1, 3, 5, 7, 9, 11, and 13 represent plasmid DNA from type 2; lanes 2, 4, 6, 8, 10, 12, and 14 represent plasmid DNA from type 1. Plasmid DNA digested with BglII (lanes 1 and 2), BglII and Sacl (lanes 3 and 4), Sacl (lanes 5 and 6), KpnI and Sacl (lanes 7 and 8), KpnI (lanes 9 and 10), Sacl and XbaI (lanes 11 and 12), and XbaI (lanes 13 and 14) was electrophoresed on 0.6% agarose gels, transferred to nylon filters, and hybridized to ipaH DNA. Lambda HindIll molecular size markers are shown at the left (in kilobases).
hybridization patterns obtained when plasmid DNA from type 1 and 2 variants was digested with BglII, BgIII and Sacl, Sacl, and XbaI indicated that DNA rearrangements had occurred within these restriction sites (Fig. 2). On the other hand, DNA rearrangements had occurred outside the KpnI and HindIII restriction sites, since no restriction fragment size changes were observed in the ipaH-hybridizing bands when plasmid DNA from type 1 and 2 variants was digested with these enzymes (Fig. 2; see Fig. SA). Hybridization of lacZ and form I antigen probes to large plasmid DNA from clonal variants. ipaH hybridization patterns from the three clonal types indicated that plasmid rearrangements and, in the case of clonal variant 3, deletions of plasmid material had occurred in the vaccine plasmid. To determine whether changes in the expression of the Lac phenotype and the form I somatic antigen were due to DNA rearrangements or to deletion of the locus responsible for the expression of the phenotype, we used the E. coli lacZ gene segment and DNA expressing the form I somatic antigen of S. sonnei to probe large plasmid DNA from the clonal variants. The lacZ gene probe only hybridized to plasmid DNA from clonal types 1 and 2 (Fig. 3), indicating that the lac locus had been deleted from the large plasmid of type 3 variants. When the 11-kb fragment from recombinant clone pHS2-14, which expresses the form I antigen, was used to probe HindIII-digested S. sonnei 53G invasion plasmid DNA, a strongly hybridizing band at 11 kb and a more weakly hybridizing band at 6.5 kb were observed (Fig. 4). When HindIII-digested large plasmid DNA from the three clonal variants was hybridized to this probe, type 1 DNA displayed the S. sonnei hybridization pattern, type 2 DNA only hybridized to the weaker 6.5-kb band, and type 3 DNA did not hybridize at all (Fig. 4A). Hindlll-digested type 1 DNA also hybridized to 2.1-kb and 900-bp fragments when
2
3
4
23 9.4 6.64.4-
2.0 -
-
