Disparity between changes in mRNA abundance and enzyme activity in Corynebacterium glutamicum: implications for DNA microarray analysis.

The relationship between changes in mRNA abundance and enzyme activity was determined for three genes over a span of nearly 3 h during amino acid production in Corynebacterium glutamicum. Gene expression changes during C. glutamicum fermentations were examined by complementary DNA (cDNA) microarrays and by a second method for quantitating RNA levels, competitive reverse transcriptase-PCR (RT-PCR). The results obtained independently by both methods were compared and found to be in agreement, thus validating the quantitative potential of DNA microarrays for gene expression profiling. Evidence of a disparity between mRNA abundance and enzyme activity is presented and supports our belief that it is difficult to generally predict protein activity from quantitative transcriptome data. Homoserine dehydrogenase, threonine dehydratase, and homoserine kinase are enzymes involved in the biosynthesis of l-isoleucine and other aspartate-derived amino acids in C. glutamicum. Our data suggest that different underlying regulatory mechanisms may be connected with the expression of the genes encoding each of these three enzymes. Indeed, whereas in one case the increases in enzyme activity exceeded those in the corresponding mRNA abundance, in another case large increases in the levels of gene expression were not congruent with changes in enzyme activity.

Development and validation of corynebacterium DNA microarrays.

We have developed DNA microarray techniques for studying Corynebacterium glutamicum. A set of 52 C. glutamicum genes encoding enzymes from primary metabolism was amplified by PCR and printed in triplicate onto glass slides. Total RNA was extracted from cells harvested during the exponential-growth and lysine production phases of a C. glutamicum fermentation. Fluorescently labeled cDNAs were prepared by reverse transcription using random hexamer primers and hybridized to the microarrays. To establish a set of benchmark metrics for this technique, we compared the variability between replicate spots on the same slide, between slides hybridized with cDNAs from the same labeling reaction, and between slides hybridized with cDNAs prepared in separate labeling reactions. We found that the results were both robust and statistically reproducible. Spot-to-spot variability was 3.8% between replicate spots on a given slide, 5.0% between spots on separate slides (though hybridized with identical, labeled cDNA), and 8.1% between spots from separate slides hybridized with samples from separate reverse transcription reactions yielding an average spot to spot variability of 7.1% across all conditions. Furthermore, when we examined the changes in gene expression that occurred between the two phases of the fermentation, we found that results for the majority of the genes agreed with observations made using other methods. These procedures will be a valuable addition to the metabolic engineering toolbox for the improvement of C. glutamicum amino acid-producing strains.