454 BOHUN B. KINLOCH, JR. 



NEW APPROACHES TO ANALYSIS OF GENETIC INTERACTIONS IN 

 WILD HOST -PATHOGEN POPULATIONS 



Development of the gene-for-gene hypothesis has provided both a 

 penetrating insight into the genetic structure and evolutionary mechanisms 

 of host-parasite relationships and a powerful tool for applied research. 

 Person (1959, 1966, 1967) discussed the origin of complementary genes as 

 an effective strategy for enabling both host and parasite to survive under 

 the mutually antagonistic selection pressures imposed by each. Mode 

 (1958) showed by a mathematical model how these genes would soon reach 

 equilibrium at intermediate frequencies in wild, outcrossing populations 

 as a necessary condition for their coevolution. Further elaboration of 

 the theory and evidence for it are discussed elsewhere (Flor, 1955; Mode, 

 1958; Person, 1959) . The usefulness of the theory derives from the kinds 

 of predictable properties intrinsic to complementary genie relationships. 

 These properties were described by Person (1959) who showed how their 

 recognition in any specific host-pathogen system -ipso facto implied the 

 existence of such a relationship, and also could form the basis of a 

 complete genetic analysis of the system. 



An elegant example of the efficacy of Person's analytical approach 

 was demonstrated by Noronha-Wagner and Bettencourt (1967) , who studied 

 leaf rust of coffee, caused by Eemileia vastatrix Berk. § Br. This is a 

 difficult system and in several ways analagous to the autoecious Peridermia ■ 

 on pines. Both hosts are long-lived trees with high degrees of hetero- 

 zygosity. The life cycle and sexual behavior of H. vastatrix is as uncon- 

 ventional and obscure as some of the peridermia. Nevertheless, in only 

 one generation of breeding, they were able to identify four dominant genes 

 for resistance in the host population, infer their complements for virulence 

 in the pathogen, and predict the likely existence of four other pathogenic 

 races not yet isolated in nature. 



The method depended on characterizing the reaction spectrum elicited 

 by each of 12 physiological races of the rust on each of a set of 8 host 

 clones that differentiated them. Progenies obtained by selfing the various 

 clones were usually either identical to their parents in reaction to the 

 12 races, which indicated that parents were homozygous at loci for rust 

 reaction, or segregated into parent-type reaction and susceptibility to 

 all races in a 3:1 ratio, implicating a dominant gene in the heterozygous 

 condition in the parent. Occasionally, selfed offspring segregated into 

 three reaction categories which included, in addition to parental type 

 and susceptibility to all races, a third reaction type characteristic of 

 another host differential. This second kind of segregation pattern 

 suggested heterozygosity at more than one locus in the parent. Subse- 

 quently, various Fj crosses among selected differentials produced an 

 array of genotypes with all possible combinations of the four genes (Table 1 

 and confirmed their identity and dominance relationships. 



Elements for the genetic analysis of this host-parasite combination 

 thus consisted only of a variable rust population, a variable host popula- 

 tion (both subject to cloning), and the selfed and Fj generation of the « 

 latter. Recognition of the properties that inhere in a gene-for-gene 

 system enabled elucidation of the relationship and number of complementary 

 genes involved. The key property is the complementary geometrical series 

 of host-pathogen interactions, in which, for each additional gene for 

 resistance in the host, the number of races capable of attacking it is 

 reduced by half (Table 1) . With host genotypes falling into this kind of 

 array, corresponding and complementary genotypes of the pathogen were 

 deduced as a logical and necessary consequence of the theory. 



