Reprint from Genetics, Vol. 46, No. 11, November, 1961. Printed in U.S.A. ANOMALOUS HOMOLOGY OF FLAGELLAR PHASES IN SALMONELLA* TETSUO IINO National Institute of Genetics, Misima, Japan Received May 30, 1961 P HASE-1 and phase-2 antigen types of a diphasic salmonella strain are deter- mined by distinct genes, H, and H,, respectively (LeperBeRG and Epwarps 1953; LepERBERG and Irno 1956). They are nonallelic and are transduced inde- pendently. Therefore, an H-antigen type recombinant obtained by a transduc- tion usually expresses the antigen type of the donor in one phase, but retains that of the recipient in the other. The present paper deals with a genetic analysis of an abnormal recombinant type which expresses phase-1 types of both donor and recipient. MATERIALS AND METHODS Salmonella ty phimurium SW1061, the original strain of the phase-1 duplicate type reported in the present paper, is an H,-inactive mutant derived from TM2 (Stocker, ZINDER and LepERBERG 1953). This wild-type strain expresses 7 anti- gen in phase-1 and 7.2 in phase-2 (designated by 7:7.2). SW1061 does not produce flagella in phase-1, though it carries the i-antigen type determinant (designated by 0(i):1.2). The genotype of SW1061 is symbolized by Fla,+- Ah,H,'H,?* (Inno 1961a). Genetic markers of the remaining salmonella strains, S. abony CDC-103 b-enx and S. heidelberg SL28 O(r:1.2), used in the present experiment are Fla,+ Ah,+H,°H,°"* and Fla, Ah,+H,"H,'”, respectively (LepER- BERG and Itwo 1956). Fla,, Ah, and H, are closely linked with each other and are transduced simultaneously in high frequency. The general procedures of cultivation, transduction and selection of serotypic recombinants were conducted according to the methods of LepErBerc and I1no (1956). Phage PLT22 was used exclusively as the mediator in transductions. Anti-H sera were prepared by the method of Enwarps and Ewrne (1955). Antigen type was determined qualitatively by slide agglutination test with anti- sera of titer 100, and quantitatively by tube agglutination test. RESULTS AND CONCLUSIONS In a transduction experiment from CDC-103 to SW1061 (Inno 1961a), a transductional clone which alternates phases between b and @ (flagellaless) was 1 Contribution from the National Institute of Genetics, Japan. No. 389. This work was sup- ported by a research grant from the National Institute of Allergy and Infectious Diseases (E-2872), Public Health Service, U.S.A. Genetics 46: 1471-1474 November 1961. 1472 TETSUO IINO obtained (Table 1). The mixture of a broth culture of the 0:b clone with the lysate of a phase-2 culture of TM2 was brushed on an NGA plate supplemented with anti-b serum as a selective agent. Twenty swarms which developed as a result of H-transductions were isolated and typed for their antigen and alter- native phase. Seven clones among them were agglutinated by anti-i serum, and after anti-i serum selection produced the alternative phase which is agglutinated by anti-b serum. These i:b type clones do not dissociate into motile and non- motile type any more, The remaining 13 clones were all agglutinated by anti-7.2 serum. An alternative H phase has not been obtained from these clones by anti-7.2 serum selection, but the dissociation of 0 type was observed in NGA plates. These results are formulated as follows: CDC-103 b:enx —x SW1061 0(7):1.2 > O(1):b TM2 i:1.2 —x O(i):b > i:b & O(i):1.2. Thus the abnormal type of recombinant obtained in the first transduction has lost the phase-2 antigen of the recipient and expresses the phase-1 antigen type of both donor and recipient alternatively. The phase-1 antigen acquired from the donor was not replaced by phase-1 antigen but phase-2 antigen in the second step transduction. The identity of b antigens of the 7:b type and the donor strain, CDC-103, was examined by cross agglutination and absorption experiments. Antisera were pre- pared against 0.5 percent formalin-saline suspensions of CDC-103 and the i:b clone. Anti-CDC-103 serum and anti-i:b serum were absorbed by i:enz cells obtained from S. typhimurium TM2 i:1.2 —x CDC-103 b:enx, and anti-b (CDC-103) and anti-b (i:b) sera were obtained, respectively. The aggluti- nation titers of anti-b (CDC-103) serum against cultures of both CDC-103 and i:b types were identical (24,000). After complete absorption of anti-b (CDC- 103) serum by the i:b cultures, the serum did not agglutinate the cells of CDC-103 any more. In parallel, anti-b (i:b) serum showed the same titer (26,000) against both CDC-103 and the 7: type cells, and it was completely absorbed by the cells of CDC-103. Thus, as far as the agglutination reaction is concerned, the 7: b type has identical b antigen as CDC-103. In order to test the linkage of i and b determinants of the 7:b type, a phage PLT22 lysate of an i:b clone culture was mixed with a broth culture of SL28 and brushed on NGA plates. The swarms grown as a result of Fla,+ transduc- TABLE 1 Transduction from a diphasic strain of S. abony, CDC-103 b:enx to SW1061 0(4):1.2. Trans- ductional clones were screened on NGA plates supplemented with anti-12 serum Number of Number of Antigen type transductional clones in reversional clones: phase-1 : phase-2 Expt. 1 Expt. 2 control for expt. 1 b:1.2 95 23 0 O:enx 84 13 0 1:1.2 4 1 1 O:b 0 1 0 HOMOLOGY OF FLAGELLAR PHASES 1473 tion were isolated and their antigen types were examined. Twenty-nine among 39 of those transductional clones were r:1.2. The remaining ten clones were i:1.2 type, which was expected from the simultaneous transduction of Fla, and H,. These two were the sole recombinant types, and no antigen type which con- tains b was obtained (LepEenserc and Iino 1956, 4 3.5). From these results it is concluded that the antigen-b determinant of the 7: b clone is not transduced linked to Fla; and H,'. Transduction was next carried out from a b-phase culture of the i:b type clone to an a-phase (phase-1) culture of a diphasic strain SW925 a:enz. Antigen type recombinants were screened on NGA plates supplemented with anti-a and anti-enx sera. Among 44 recombinant clones, 37 expressed i and enzx in phase-1 and phase-2, respectively. The remaining seven clones expressed a and b-antigen types alternatively. No recombinant showing 7.2 type was detected. The fre- quency of the alternative expression of a or b-antigen types is the same as that of phase variation in SW925. Thus, the b determinant was transduced and re- placed with H, of the recipient. In conclusion, the b determinant of the i:b type clone behaves exclusively as an allelic locus to H,. The possibility that a hidden #7,’" is present in the phase-1 antigen duplicate type is excluded by the antiserum selection and the transduc- tion tests. DISCUSSION The possibilities of an H,-duplication or a heterogenote of H, in the phase-1 duplicate strain reported in this paper are excluded for the following reasons: (1) allelism between H, and the b-type determinant and (2) the absence of H,'* in the strain. The phenomena are best explained by the assumption that the duplication of phase-1 antigen occurred by an unequal recombination result- ing in the replacement of H,’* by H,”. The new type now behaves as if it were H,' H,". The occurrence of such an unequal recombination indicates a structural homology between H, and H,. H, and H, might have a certain degree of synaptic affinity. The chance of synapsis of H, and H, in transduction of these genes may be very rare, as the phase-1 duplicate type reported in the present paper is the only occasion of unequal recombination out of over 3,000 serotypic recombinants which have been examined in our laboratory for the various transductional analyses of salmonella H antigens. The behavior of the translocated b-antigen determinant as an allele of H, indicates that the barrier of synapsis between H, and H, are not their own structural differences, but may be the differences in other genes involved in a transductional fragment together with H, or H.. The genes which have been known to be transduced simultaneously with H, are Fla, (Stocker et al. 1953; Leperserc and Iino 1956) and a methylator of lysine in flagellin molecules (Stocker, McDonoucH and AmBier 1961). The genes homologous to these H,-linked genes are not present in the H, region, instead a stability controller of antigenic phases, Vy, is closely linked to H, (Inno 1961b). The genetic homology of H, and H, leads us to a speculation on their phylo- 1474 TETSUO IINO genic relations: one of them might have originated by duplication and transloca- tion of the other; the structural differentiation might have occurred between then thereafter. These postulates trace the process of evolution from the primary monophasic serotype to the diphasic type. The secondary monophasic types might have originated from the latter by inactivation of either H, or H, (Iino 1961a). The occurrence of H, or H, duplication in nature is indicated from the isolation of H,-duplicate strains (LepERBERG 1961) or of so called triphasic strains (re- viewed by KAUFFMANN 1954). The frequency of the alternative expression of two antigen types in the phase-1 duplicate strain is the same as that in the original transductional recipient strain SW1061: the & determinant translocated to the H, locus becomes unstable and changes its activity at the frequency specific to the H, locus of the strain. The stability of the H, state is regulated by VA, which is closely linked to H, (Irno 1961b). Consequently the present results suggest that the presence of an antigen determinant in the vicinity of Vj, is essential for its action on H). SUMMARY An abnormal H-antigen type recombinant which alternatively expresses phase-1 antigens of both donor and recipient was obtained from a transduction between Salmonella typhimurium SW1061 and S. abony CDC-103. The duplicate phase-1 antigen type determinant H,,’ of the recombinant be- haves as an allelic locus of phase-2 antigen type determinant H,. The recombin- ant is presumed to be originated by an unequal recombination: H, locus is replaced by H, of the donor in the transduction. The phylogenic relation between H, and H, was discussed. ACKNOWLEDGMENT The author wishes to express his sincere thanks to Prorsssor J. Leperserc of the Genetics Department, Stanford University, California, for his interest and advice in going through this investigation. LITERATURE CITED Epwarps, P. R., and W. H. Ewre, 1955 Identification of Enterobacteriaceae. Burgess Publ. Co. Minneapolis, Minnesota. Inno, T., 1961a Genetic analysis of O-H variation in Salmonella. Japan. J. Genetics 36: 268— 375. 1961b A stabilizer of antigenic phases in Salrnonella abortus-equi. Genetics 46: 1465-1469. KauFFMann, F., 1954 Enterobacteriaceae. 2nd ed., E. Munksgaard, Copenhagen, Leperserc, J., 1961 A duplication of the H, (flagellar antigen) locus in Salmonella. Genetics 46: 1475-1481. Leperserc, J., and P. R. Epwarps, 1953 Serotypic recombination in Salmonella. J. Immunol. 71: 232-240. LeperBere, J., and T. Inno, 1956 Phase variation in Salmonella. Genetics 41: 743~757. Stocker, B, A. D., M. W. McDonoueu, and R. P. AMpLer, 1961 A gene determining presence or absence of N-methyl-lysine in Salmonella flagellar protein. Nature 189: 556. Stocker, B. A. D., N. D. Zinver, and J. Leprernerc, 1953 Transduction of flagellar characters in Salmonella. J. Gen. Microbiol. 9: 410-433.