SPECIFICITY IN PL.ANT DISEASE 33 



gametic pairing of alleles (Category I interactions); e.g. aa, aA, Aa, and 

 AA. By carrying this comparison further, we can say that in the Category 

 III interaction I is recessive and B is dominant. This in no way implies 

 what is recessive or dominant in the Category I interaction in host or 

 pathogen. This peculiarity is one of the reasons for much of the misunder- 

 standing of the gene-for-gene concept. The phenotype for each Category III 

 genotype in a given environment is typical though not necessarily uniaue. 

 Thus in considering wheat stem rust we can write Lpl/Lrl = Litl = 1+; 

 Lpl/Hrl - Eitl = 3+, etc. This integrity of the phenotype is extremely 

 useful for many studies, provided there is variability in the phenotype 

 of Lit for different corresponding gene pairs. 



In Figure 1, the Category III genetic interaction is illustrated by 

 use of organisms of n + n genetic constitution such as in flax rust. If 

 in this model we dealt with a disease such as mildew of barley we would 

 have a host of n + n constitution and a pathogen of n constitution. A 

 model of this interaction would have only 6 of the classical genotypes of 

 the aegricorpus instead of 9, but this would in no way change the gene- 

 for-gene genotypes. Thus the nuclear condition of host and pathogen need 

 not be considered in learning to understand the gene-for-gene concept. 

 Of course the nuclear condition must be known when applying the concept 

 in experiments. 



CATEGORY IV INTERACTION 



Most host rpathogen relationships involve many corresponding gene pairs 

 We know this because many crop plants are known to carry several to many 

 genes for reaction. For each of these there must be a corresponding gene 

 in the pathogen, according to the gene-for-gene concept. This is called 

 the Category IV interaction. This is easy to visualize by use of the L 

 and E symbolization. For example, a host plant has a genotype LrlLrlErlErA 

 and the pathogen has the genotype LplEplLplEpA (the numerals refer to loci) 

 If we put the pathogen on the host we have: 



Lpl 'dpi Lpl EpA 

 Lrl Lrl Erl ErA 



From our discussion of the Category III interactions we know that Lipl/Lrl 

 = Litl, Hp2/Lr2 = Eitl Lpl/Erl = Eitl, and EpA/ErA = Hit A . What is the 

 result of the system shown? All evidence to date shows that Litl will be 

 expressed or, we might say, be epistatic to the other Category III inter- 

 actions which would give a high infection type. This is why a single gene 

 for "resistance" is effective so long as the pathogen does not have the 

 corresponding genotype for Ep. 



A host rpathogen relationship often has two corresponding gene pairs 

 for Lit. Such a situation can be illustrated as follows: 



Lpl Lpl Lpl EpA 

 Lrl Lrl Erl ErA 



Corresponding gene pairs 1 and 2 are both for Lit and therefore epistatic 

 to Eitl and EitA . Expression of Litl Litl might be in two ways. We will 

 assume that Litl = infection type 1+ and Litl = infection type 2-. Infec- 

 tion type 1+ is lower than 2- and would usually be expressed--and thus be 

 epistatic. Cases are known where the combination of two Lit x s in one 

 system will result in an infection type lower than either of the Category 

 III interactions involved. In the assumed system shown above the infection 

 type of the aegricorpus might be 0. In this case the two Category III 



