Perceived by Charles Darwin in many vegetable plants and rediscovered by George H Shull and Edward M East in maize, heterosis or hybrid vigour is one of the most widely utilized phenomena, not only in agriculture but also in animal breeding. Although, numerous studies have been carried out to understand its genetic and/or molecular basis in the past 100 years, our knowledge of the underlying molecular processes that results in hybrid vigour can best be defined as superficial. Even after century long deliberations, there is no consensus on the relative/individual contribution of the genetic/epigenetic factors in the manifestation of heterosis. However, with the recent advancements in functional genomics, transcriptomics, proteomics, and metabolomics-related technologies, the riddle of heterosis is being reinvestigated by adopting systems-level approaches to understand the underlying molecular mechanisms. A number of intriguing hypotheses are converging towards the idea of a cumulative positive effect of the differential expression of a variety of genes, on one or several yield-affecting metabolic pathways or overall energy-use efficiency, as the underlying mechanism for the manifestation of heterosis. Presented here is a brief account of clues gathered from various investigative approaches targeted towards better scientific understanding of this process.

Illinois long‐term selection strains of maize (Zea mays L.) have been useful for identifying genomic regions controlling kernel oil, protein, and starch concentrations. To identify kernel trait quantitative trait loci (QTL) in a genetic background more relevant to practical breeding, 150 BC1‐derived S1 lines (BC1S1s) were produced from Illinois High Oil and recurrent parent B73. Oil, protein, and starch were measured in BC1S1s and in Mo17‐topcross hybrids (TCs). Kernel mass of BC1S1s and grain yield of TCs were also determined. Starch was positively correlated with mass in BC1S1s (rp = 0.67**, α ≤ 0.01) and with yield in TCs (rp = 0.59**). Oil was negatively correlated with mass in BC1S1s (rp = −0.29**) and with yield in TCs (rp = −0.30**). Oil was negatively correlated with starch in BC1S1s (rp = −0.75**) and TCs (rp = −0.66**). A genetic map with length = 1486 cM was created with 110 markers. Multiple regression models with QTL detected by composite interval mapping (CIM) explained 46.9, 45.2, 44.3, and 17.7% of phenotypic variance for oil, protein, starch, and mass, respectively, in BC1S1s and 17.5, 22.9, 40.1, and 28.7% for oil, protein, starch, and yield, respectively, in TCs. A 22 cM‐interval on chromosome 6 in BC1S1s included oil, protein, and starch QTL, including a QTL explaining 36.7% of the BC1S1 phenotypic variation for oil. No yield QTL were detected in this region. Introgression of this QTL into breeding lines might increase oil while maintaining yield.

Aflatoxin B(1) formed by Aspergillus flavus Fr:Link has been associated with animal disease and liver cancer in humans. We performed genetic studies in progenies derived from maize inbred Tex6, associated with relatively low levels of aflatoxin production, crossed with the historically important inbred B73. (Tex6 x B73) x B73 BC(1)S(1) and Tex6 x B73 F(2:3) mapping populations were produced and evaluated in 1996 and 1997 in Champaign, Ill. Ears were inoculated 20 to 24 days after midsilk using a pinboard method and a mixture of conidia of A. flavus Link:Fr. isolates. Aflatoxin B(1) levels in harvested ears were determined using an indirect competitive ELISA. Molecular markers were assayed on the populations and used to generate maps. Molecular marker - QTL associations for lower levels of aflatoxin production were determined using multiple regression (MR) and composite interval analysis with multiple regression (CIM MR). MR revealed sets of markers associated with lower aflatoxin production in 1996 and 1997, and CIM MR detected a smaller subset of loci significant in 1997. QTLs for lower aflatoxin were attributed to both Tex6 and B73 parental sources. Environment strongly influenced the detection of QTLs for lower aflatoxin production in different years. There were very few chromosome regions associated with QTLs in more than 1 year or population with MR analysis, and none with CIM MR analysis. In 1997, QTLs for lower aflatoxin were detected with CIM MR in bins 5.01-2 and 5.04-5 in the BC(1)S(1) population, and in bins 3.05-6, 4.07-8 and 10.05-10.07 in the F(2:3) population. These QTL associations appear the most promising for further study.

Increased levels of oleic acid may enhance the nutritional and functional value of corn. Corn oil is primarily composed of palmitic, stearic, oleic, linoleic and linolenic fatty acids. Delta-12 desaturase in plants converts oleic acid (18:1) to linoleic acid (18:2) by inserting a double bond at the delta-12 position. Fatty acid desaturase-2 (fad2) encodes delta-12 desaturase that functions in the endoplasmic reticulum while fatty acid desaturase-6 (fad6) encodes delta-12 desaturase that functions in plastids. Complementary DNA (cDNA) clones from putative maize homologs for fad2 and fad6 were identified and the entire clones DNA sequenced. The maize fad2 cDNAs showed an amino-acid identity of 67-77% to fad2 of Glycine, Arabidopsis and Brassica species. Our cDNA sequence comparisons suggested that more than one fad2 gene is transcribed in maize embryos. Two different fad2 cDNAs from an embryo cDNA library map to separate chromosomal positions, providing evidence consistent with two different isoforms of fad2 expressed in the embryo. The fad2 cDNAs from multiple tissue sources clustered into three groups on a phenogram, and map to different positions on chromosomes 4, 5 and 10, which suggests at least three different isoforms of fad2 may be expressed in the maize plant. The two maize fad6 cDNAs share 81% amino-acid identity with the Arabidopsis fad6 and map to chromosome 1. Northern analysis revealed that fad2 is transcribed in embryos at 14, 21, 28 and 35 days after pollination, with the highest level observed at day 14. None of the fad2 or fad6 clones mapped to maize chromosome bins associated with QTLs for the ratio of oleic/linoleic acid, notably bin 6.04 which contains the linoleic1 locus and the largest reported QTL for the oleic/linoleic ratio. This suggests, but does not prove, that some of the QTLs for the oleic/linoleic acid ratio do not involve allelic variants of fad2 or fad6 but rather involve other genes that may influence flux through the enzymes encoded by fad2 or fad6.