Glucosinolates are amino acid-derived natural plant products characteristic of the order Brassicales, including the agriculturally important oilseed rape, the Brassica vegetables and the model plant Arabidopsis 
. Glucosinolates are major defensive metabolites against both herbivores and pathogens and function as attractants for specialized insects 
. In addition, glucosinolates have gained increasing importance due to their function as flavour compounds and cancer-preventive agents in the human diet, as well as their potential for use as biopesticides 
Glucosinolates are divided into aliphatic, aromatic and indole glucosinolates dependent on the nature of the amino acid from which they are derived. The three subgroups have a pathway for the synthesis of the core structure, where distinct, but homologous genes catalyze the conversion of amino acids to the corresponding glucosinolate. Methionine-derived glucosinolates are modified via elongation of the side chain prior to entering the pathway for the synthesis of the core glucosinolate structure. Short-chained, aliphatic glucosinolates have been through one to four cycles of the chain-elongation machinery, whereas the long-chained have been through five or six cycles (). Furthermore, glucosinolate structures are often secondarily modified on their side chains with e.g. hydroxylations and alkenylations. The completion of the Arabidopsis genomic sequence has resulted in identification of most of the biosynthetic genes involved in the formation of the core structure, the chain-elongation machinery and secondary modifications ().
Biosynthetic pathway of aliphatic glucosinolates in Arabidopsis Col-0.
Our knowledge about the genes controlling glucosinolate regulation is limited to a few regulatory genes. This is in contrast to the complex regulation known for glucosinolate content which includes herbivore, pathogen, biotic and abiotic stress responses 
. Two genes are known to specifically regulate indole glucosinolates. ATR1
was identified as a gain-of-function mutant, atr1D,
in a screen for mutants with a
. The transcript levels of CYP79B2, CYP79B3
which encodes for enzymes in the biosynthesis of indole glucosinolates, are up- and down-regulated in atr1D
and a knockout mutant, respectively 
, a close homologue of ATR1
, was identified as an activation tagged line with high indole glucosinolates 
. Complementation studies suggest that MYB34
have overlapping, but distinct functions in regulating biosynthetic genes for indole glucosinolates.
Other regulators have been shown to regulate both indole and aliphatic glucosinolates. TFL2
, a plant homologue of the animal heterochromation protein (HP1) was shown to affect both aliphatic and indole glucosinolates, probably through indirect developmental effects 
. A screen for high-glucosinolate mutants identified IQD1
, a calmodulin-binding nuclear protein, as a potential regulator 
. The iqd1
mutant has elevated levels of both aliphatic and indole glucosinolates, although transcript levels of genes in the indole glucosinolate pathway are up-regulated and transcripts in the aliphatic pathway are repressed 
. Over-expression of a third transcription factor, AtDof1.1
, results in an approximately two-fold increase in the levels of both indole and short-chained aliphatic glucosinolates. Accordingly, these 35S:Dof1.1
lines up-regulate transcript levels for ATR1/MYB34
and the genes involved in indole glucosinolate biosynthesis, CYP79B2, CYP79B3
as well as of MAM1
involved in chain-elongation machinery for aliphatic glucosinolates 
. Although these three regulatory genes, TFL2, IQD1
, affect both aliphatic and indole glucosinolates, many studies indicate that aliphatic and indole glucosinolates are typically not co-ordinately regulated in response to pathogens, hormones or insect herbivores 
When we initiated our study, the aim was to identify regulators specific for biosynthesis of the aliphatic glucosinolates, which had not yet been identified. Our approach was to exploit the natural variation in production of glucosinolates among Arabidopsis ecotypes using QTL mapping to identify candidate regulators. We have used a combination of metabolite QTL and expression QTL (eQTL) analysis within the Bay-0×Sha Recombinant Inbred Lines (RILs) to show that a region of the genome controls variation in both the accumulation of aliphatic glucosinolates and transcript levels for the aliphatic biosynthetic genes. Combining this data with information from co-expression databases and phylogenetic analysis identified three candidate transcription factors for aliphatic glucosinolates, MYB28, MYB29, and MYB76, belonging to an Arabidopsis specific MYB group. Glucosinolate and transcript profiling of over-expression lines indicated that MYB28, MYB29 and MYB76 positively regulate accumulation of aliphatic glucosinolates and the corresponding transcripts. This was further supported by a decrease in levels of aliphatic glucosinolates in T-DNA insertion mutants of MYB28, MYB29 and MYB76, where the mutants reveal differential specificity for chain lengths. Further, a double knockout mutant in MYB28 and MYB29 was devoid of aliphatic glucosinolates and displayed very low levels of transcripts not predicted from the single knockout mutants. During our study, Hirai et al. (2007) provided evidence that MYB28 is a regulator of aliphatic glucosinolates using an omics-based approach and Gigolashvili et al. (2007) has used activation tag approach to identify MYB28 as regulator of aliphatic glucosinolates. Their data is discussed in relation to our data showing that all three MYBs are regulators of aliphatic glucosinolates independent of biotic stresses.