Comparative analysis of differential gene expression in the culms of sorghum
Despite numerous attempts involving a variety of target genes, the identity of the key regulatory genes of sucrose metabolism in sugarcane is still illusive. To date, genomic research into sucrose accumulation in sugarcane has focused on genes that are expressed in association with stalk development/maturation, with the aim of identifying key regulatory steps in sucrose metabolism. The identification of possible controlling points, however, is complicated by the polyploid nature of sugarcane. Although these studies have yielded extensive annotated gene lists and correlative data, the identity of key regulatory genes remains elusive. A close relative of sugarcane, Sorghum bicolor, is diploid, has a small genome size and accumulates sucrose in the stalk parenchyma. The main aim of the work presented in this thesis was to use S. bicolor as a model to identify genes that are differentially expressed during sucrose accumulation in the stalk of low and high sucrose genotypes. In the first part of the study, a macroarray protocol for identification of differentially expressed genes during sorghum development was established. Firstly, the macroarray sensitivity of probe-target hybridisation was optimised with increasing amounts of target DNA i.e. 0.005-0.075 pmol. The hybridisation signal intensity increased as expected with increasing amounts of probe until the hybridisation signals reached maximum levels at 0.05 pmol. As a result, to ensure quantitative cDNA detection, probes were arrayed at 0.05 pmol when 1 μg target cDNA was used. Secondly, intra-array and inter-array membrane reproducibility was found to be high. In addition, the protocol was able to detect species of mRNA at the lowest detection limit tested (0.06%) and permits the detection of an eight-fold variation in transcript levels. The conclusion was therefore that the protocol was reproducible, robust and can reliably detect changes in mRNA levels. In the second part of the study, sugar accumulation levels in the immature and maturing internodal tissues of sorghum GH1 and SH2 genotypes were compared during the boot and softdough stages. Sugars (i.e. fructose, glucose and sucrose) accumulated differently in the immature and maturing internodes in both sorghum genotypes during the boot and softdough stages, with sucrose being the dominant sugar in both stages. Based on these differences in sugar accumulation patterns, immature and maturing internodal tissues of sorghum genotypes were compared for differentially expressed genes. A number of genes were found to be significantly differentially expressed during both stages. In order to validate the reliability of the macroarray analysis, fourteen genes were arbitrarily selected for semi-quantitative RT-PCR. Seven genes (50%) revealed a similar pattern of transcript expression, confirming the macroarray results. The other seven genes, however, showed a different expression trend compared with the macroarrays. In this study, ESTs from rice and sugarcane were used for probing sorghum. The probability of cross-hybridisation between the probes and various isoforms of the homologous sorghum sequences is thus high, potentially leading to the identification of false positives. In addition, variation in expression patterns could have been introduced by technical and biological variation. Lastly, to verify that changes in the levels of a transcript are also reflected in changes in enzyme activity, seven candidates were tested for enzyme activity. Only three i.e. soluble acid invertase (SAI), sucrose synthase (SuSy) and alcohol dehydrogenase (ADH), out of these seven genes showed enzyme activity levels reflective of the relative transcript expression. We concluded that changes in transcript levels may or may not immediately lead to similar changes in enzyme activity. In addition, enzyme activity may be controlled at transcriptional and at posttranscriptional levels. In conclusion, sugar accumulation in low (GH1) and high (SH2) sucrose sorghum genotypes is influenced by differences in gene expression. In addition, the power of macroarrays and confirmation with semi-quantitative RT-PCR for identification of differentially expressed genes in sorghum genotypes was demonstrated. Moreover, the transcript and enzyme activity patterns of SAI, SuSy and ADH genes showed expression patterns similar to those of sugarcane during sucrose accumulation. Therefore, using sorghum as a model promises to enhance and refine our understanding of sucrose accumulation in sugarcane.