Comparing Spartina alterniflora ESTs to Oryza sativa and Zea mays gene sequences: Implications for cross-species hybridization microarrays

Background/Question/Methods

Genomic tools developed for model species, such as Oryza sativa and Zea mays, can be applied to closely-related, ecologically important, non-model species.  Gene expression profiling through the use of DNA microarrays is a particularly powerful tool to examine the molecular response of plants to environmental change.  Cross-species (heterologous) microarrays hybridize target non-model cDNA with microarray probes from model species.  The effectiveness of cross-species hybridization is largely based on the coding sequence similarity between the model and non-model species.  Spartina alterniflora (Loisel.), a C₄ grass, is the dominant species of north Atlantic and Gulf Coast salt marshes and provides numerous important ecosystem services and functions.  The close-relatedness of S. alterniflora to model grasses allows for cross-species microarray studies that explore the transcriptomic response of this foundation plant species to environmental change.  Oryza sativa microarrays have been demonstrated to hybridize with Spartina spp. leaf RNA; however, the more closely-related Z. mays may offer better hybridization and ultimately be a superior alternative. To compare the cross-species hybridization potential of S. alterniflora RNA to O. sativa and Z. mays microarrays, we used the BlastN algorithm to compare sequence similarity of 100 S. alterniflora expressed sequence tags (ESTs) to O. sativa and Z. mays genes.

Results/Conclusions

Of the 100 S. alterniflora ESTs examined, 63 showed higher similarities to Z. mays (based on E-values), while 32 had more similar sequences to O. sativa, and 5 were similar between species.  Fifty-eight Spartina ESTs would likely demonstrate faultless hybridization (E-value < 10^-30) with Z. mays, compared to 50 with O. sativa.  Eighteen Spartina ESTs would likely demonstrate intermediate hybridization (E-values between 10^-8 and 10^-30) with Z. mays, as opposed to 20 with O. sativa.  A total of 30 Spartina ESTs would likely not hybridize successfully (E-value > 10^-8) with O. sativa; however, that number is reduced to 24 with Z. mays.  These results are similar to published research on hybridization between Spartina spp. RNA and O. sativa microarrays (~70%), but no study has utilized cross-species microarrays between Spartina spp. and Z. mays.  Our results suggest that Z. mays microarrays may improve hybridization success between target and probe in cross-species microarray studies using Spartina spp.  Improving hybridization success is important when trying to detect significant changes in gene expression under variable field conditions.  We provide a conceptual framework for the design and analysis of microarray experiments, which often lack biological replication and suffer from the misuse of multivariate statistical techniques.