Genetic Analysis 275Toward the 1950s, with the improvement of the resolution power of genetic re-
combination experiments (see section 4), the internal linkage relationships of the
genes themselves started to unfold. Altenburg and Muller [1920] called attention
to a complex gene of Drosophila,Truncate, being a variable character, both somat-
ically and genetically [Carlson, 1959]. Its complex nature, exhibiting combinations
of three presumably independent phenotypes, was not resolved at that time. E.
B. Lewis conceived of closely linked but recombinable genes of Drosophila, such
asStubble-stubloid, Star-asteroidorbithorax-Ultrabithorax-bithoraxoid,as “nests”
of genes of similar function, possibly due to ancient duplication of one gene fol-
lowed by its functional diversification with evolutionary time. He tried to extend
this notion, when recombinants were detected in apparently non-complex loci such
aswhiteandlozengeDrosophila eye-color genes. Rarely recombination occurred
between two mutants that did not complement each other —hence by definition,
being of the “same gene” — when juxtaposed in a compound heterozygote (in
trans:a 1 +/ +a 2 ). The phenotype of the recombinant in which the two muta-
tions were now on the same homologue chromosome (incis: a 1 a 2 /++,the
other homologue chromosome being non-mutant), was wild type. Lewis called
these mutants “pseudo-alleles.”. He suggested that pseudo-alleles belonged to dis-
tinct adjacent genes, one producing the substrate for the function of the other. If
the products were induced in close proximity to the genes and could defuse only
to the adjacent gene on thesamechromosome, but not as far as that gene on
thehomologouschromosome, then thecisconstellation, but not that of thetrans
constellation, would result in complementarity. This interpretation rescued the
notion of discrete and distinct genes [Lewis, 1951].
Similar studies, in more “classical” Drosophila genes [Chovnick, 1961; Green,
1963] suggested that they too were divisible into two (or three) distinct subunits,
rather than being single genes with mutation-site distributed all over. This no-
tion was finally overcome only with Chovnicket al.’s study of therosy gene of
Drosophila, in the footsteps Benzer’s genetic analysis of therIIlocus of the T4
bacteriophage of Escherichia coli (see below). In these studiesrosycould be
mapped as a continuous linear entity with recombinable mutation sites all along
[Chovnick, 1989; Chovnicket al., 1964].
Although the scale of recombination experiments with Drosophila was upgraded
considerably, achieving resolution between mutants of 0.01 cM apart and less (cM
- centi-Morgan –1% recombination), the introduction of experimental organisms
in which more efficient methods for screening for the relevant recombinant progeny
was imperative. Pontecorvo utilized auxotrophic mutants (that could not grow on
a minimally-defined medium, on which the prototrophic wild-type could grow) of
the moldAspergillus nidulans. He hybridized independently obtained mutants of
the same gene and screened for the products of rare recombinations by plating the
progeny spores on the minimal medium. Inter-allelic (or, inter-pseudo-allelic) re-
combinants could be discerned from (equally rare) back-mutations by checking for
recombination of outside markers — those to the left and right of the experimental
target markers. It soon became clear that auxotrophic mutations in given genes