Since Shull's original description of heterosis, breeders have made wi
de use of this phenomenon. However, while agronomists have been utiliz
ing heterosis as a means of improving crop productivity, the biologica
l basis of heterosis remains unkown. It is generally believed that our
understanding of heterosis will enhance our ability to form new genot
ypes which may be used directly as F-1 hybrids or form the basis for f
uture selection programmes. While the original concept of heterosis re
sulted from studies at the phenotypic morpohological level, they were
soon followed by biochemical data with the advent of electrophoresis a
nd the consequent ease of accumulation of data related to isozyme vari
ability. However, the large number of restriction fragment length poly
morphisms from more recent studies has allowed the development of link
age maps with a high degree of resolution useful in locating and manip
ulating quantitative trait loci (QTL). When substantial numbers of suc
h neutral markers were used to measure genetic distance in large numbe
rs of maize inbreds, very significant correlations were recorded betwe
en parental genetic distance and hybrid performance. Through the same
approach, a relatively small number of QTLs dispersed through the maiz
e genome were identified which show clear overdominance expression con
trolling heterosis. The hypothesis was made that some QTLs could code
for regulatory proteins since these proteins are able to control a vas
t array of other structural genes, the products of which are necessary
for the expression of complicated characters such as yield and hetero
sis for yield. The few such proteins identified thus far are all multi
meric proteins with the heteropolymers exhibiting significantly differ
ent activities in comparison with the homopolymers, that is in complia
nce with the clear overdominance manifestation of the few QTLs analyze
d so far. In addition, parameters derived from the Variability of geno
me expression assessed through studies of polymorphisms in the amounts
of individual proteins or mRNAs show numerous significant correlation
s between these indices and hybrid vigor. These correlations supported
the conclusion that QTLs could be loci controlling the amount of mRNA
s or proteins synthesized from a number of structural genes, and stres
s the significance of both the regulatory proteins (and their encoding
genes) and the structural genes, being regulated, in manifestation of
complicated characters, such as heterosis.