Interpretations of some large non-linear complementation maps.

general theory of interallelic complementation has been proposed by CRICK A and ORGEL (1964), and this has been used by ISHIKAWA (1965) to interpret the relationship between the complementation map and the recombination map of the ad-8 locus in Neurospora crassa and has also been used to interpret a more complex complementation map at the leu-2 locus in Neurospora crassa (GROSS 1962; GILLIE 1966). It seems that most complementation maps appear linear because they are constructed using rather small numbers of complementing mutants, and that random reduction of numbers of mutants in a large non-linear complementation map may result in the map becoming linear (GILLIE 1966). Since most loci may have complex complementation maps if sufficient data were available, it appears important that any theory of complementation should be able to account for such maps. Full confirmation of the CRICK-ORGEL theory may only be possible when chemical, physical and genetical methods are used on one protein. However, the general usefulness of the theory may be tested by seeing whether it is applicable to the more complicated maps in a consistent way. This paper shows how the theory may be used to interpret four more complementation maps. A previous analysis of the leu-2 map showed that complementation groups (or complementing mutants) could be considered to belong to one of two clusters, located at one or other pole of the circular complementation map (GILLIE 1966). Mutants located in groups at any one pole of the map showed no complementation amongst themselves but complemented with mutants in certain groups at the other pole of the map (Figure 1) . Using these observations it was found to be possible to interpret the leu-2 map as a topologically circular interface between two protein monomers in a dimer with a single axis of symmetry. GROSS (1962) has interesting evidence to suggest that the leu-2 and leu-3 gene products interact to form a multimer, possibly a tetramer, of the type aapp where 01 is the gene product of leu-2 and ,8 the gene product of leu-3. This view does not conflict with the interpretation of the leu-2 map given here; the leu-2 gene product is referred to here as a dimer, as this is the simplest situation pertinent to our discussion. The poles of the leu-2 map at which complementation groups were found to cluster were identified with the point at which the polypeptide chain intersected the axis of symmetry of the dimer; because, according to the CRICK-ORGEL theory, mutants involving defects at a particular intersection of the polypeptide chain