BERGAMO, ITALY
Istituto Sperimentale per la Cerealicoltura

Cloning and characterization of the glossy1 gene of maize

--Sturaro, M, Hartings, H, Motto, M

The glossy-1 (gl1) locus of maize is required for the formation of the epicuticular wax layer of young plants. gl1 mutant seedlings can be visually identified because of their glossy leaf surface which is different from the dull surface of wild-type seedlings. The gl1 locus was isolated by transposon tagging. Seven unstable mutations, gl1-m1 to gl2-m9 were induced in a parental strain carrying an active transposable Activator (Ac) element in the unstable wx-m7 allele. Genetic tests on the gl1-m5 allele indicated that it was not caused by the Ac element but by the insertion of the transposable element Enhancer/Suppressor-Mutator (En/Spm). A HindIII restriction fragment segregating with the mutant phenotype was identified, by Southern analysis, using sequences from the En/Spm element as a probe. Part of the fragment was cloned and was shown to carry part of the unstable gl2-m5 allele. These gl1 sequences were used to identify a genomic fragment carrying the wild-type allele and to isolate its corresponding cDNA sequence.

Computer-aided analysis of the genomic sequence permitted us to identify the putative exons encompassing the gl1 transcript. A database search for proteins homologous to the deduced gl1 polypeptide bolstered the postulated mRNA sequence. On the basis of these data two primers were designed to isolate the full coding sequence of gl1 by RT-PCR. A single 2,056 bp fragment including a 1,866 nucleotide long ORF was amplified from RNA extracted from wild-type seedling leaves. An in-frame stop codon is present 87 bp upstream of the ATG start codon of the main ORF. No alternative translation start sites can be found, indicating that the amplified fragment includes the complete coding region. Putative CAAT- and TATA-box motifs can be found in the promoter sequence 200 bp and 146 bp upstream of the ATG start codon, respectively, while a putative polyadenylation site is present 312 bp downstream of the translation stop codon.

Translation of the 1,866 nucleotide long open reading frame present in the gl1 cDNA sequence gives rise to a putative polypeptide of 621 amino acids with an apparent molecular weight of 69.6 kDa and pI of 9.89. Hydropathy analysis predicts the presence of several transmembrane domains in the N-terminal region of the Gl1 polypeptide, as well as a hydrophilic C-terminal domain. Furthermore, a histidine-rich motif characteristic of a family of membrane-bound desaturases/hydroxylases is present in the N-terminal part.

A database search for proteins homologous to Gl1 with the TBLASTX algorithm revealed several sequences exhibiting high levels of similarity with the query sequence used. In particular, a putative polypeptide of 619 amino acids encoded by a cDNA from O. sativa (AK060786) showed 84% identity over its entire coding sequence. Furthermore, significant homologies, with a 67% identity score, were found with the products of two other rice cDNAs (AK066569 and AK070469), with the WAX2 locus of Arabidopsis thaliana encoding a protein involved in cuticle synthesis (62% identity), and a partial polypeptide (L33792) derived from Senecio odora (55% identity). From this analysis it was argued that the highest degree of homology consistently regards the C-terminal part of the deduced proteins.

A comparison of the deduced Gl1 protein sequence and the product of the A. thaliana CER1 locus, a putative aldehyde decarbonylase active in the epicuticular wax biosynthesis pathway, reveals an overall identity of 35%. This similarity score is significantly lower than the degree of similarity encountered between the putative Gl1 and A. thaliana WAX2 proteins (62%). Since previous results had attributed to the gl1 locus a role as a CER1-orthologue from maize (Hansen et al., Plant Physiol. 113:1091–1100, 1997), we further investigated amino acid sequence similarities among a restricted group of Gl1-homologous sequences by means of phylogenetic analysis. These analyses suggest the presence of two groups of protein sequences, the former containing the CER1 protein as a founder sequence, the latter including the WAX2 sequence. Interestingly, the Gl1 sequence shows a high level of homology with the members of the WAX2 group, while a second maize sequence (GenBank AY104752) is located within the CER1 group, with which it shares 55% amino acid identity. Hence, phylogenetic analysis indicates that the gl1-related sequences can be divided into two subgroups, each comprising genes from at least three species: maize, rice, and Arabidopsis.

In conclusion, the genomic and cDNA sequences we have isolated differ from the putative gl1 gene and transcript previously identified (Hansen et al., 1997). The protein encoded by gl1 shows significant homology with the entire sequence of the WAX2 gene product of Arabidopsis, involved in both cutin synthesis and cuticular wax production.