We have reported the results of preliminary screening of genes whose transcriptional levels alter by autotetraploidy in rice using cDNA microarrays (Kishimoto et al. 2005). In the previous study, the criterion for selecting the ESTs was more than a two-fold change with 36 data points chosen for 12 ESTs in 3 japonica lines and 46 data points chosen for 23 ESTs in 2 indica lines. From these, 28 japonica data points (77%) and 38 indica data points (82%) had values that changed by two- to three-fold. No common EST was found between japonica and indica comparisons. The relatively large changes in expression observed in the previous study did not reveal significant differences in expression levels between diploid and their autotetraploid common to japonica and indica.

To re-investigate alteration of transcriptional levels in autotetraploid compared to those in diploid, we analyzed newly isolated RNAs from 8 of the lines used in Kishimoto et al. (2005) and other two lines (with "Dee-geo-woo-gen" background) with the same cDNA microarray system used in Kishimoto et al. (2005) (http://cdna01.dna.affrc.go.jp/RMOS/main_en.html; unique 8987 EST clones of rice are divided onto a pair of slides where each slide has duplicate spots for each EST (one in each of the Left and Right fields).

The seed materials were distributed by NIAS GeneBank. These rice lines were grown in a greenhouse (26-36 degrees C) under natural light condition for 14 days after sowing in bulk for each line from the 5th of June to the 19th of June; these lines were grown in 2001. We used the first slide of the pair of the cDNA microarrays in this study, due to a shortage of RNA resulting from the small number of seeds sown in this study. The first slide carries 4512 EST clones in duplicate (http://cdna01.dna.affrc.go.jp/RMOS/public_data_en.html). The methods on micorarray analysis and data analysis were as described by Yazaki et al. (2003), except for the methods for expression data analysis as described below.

For this series of experiments, we took two following steps in order to analyze quantified expression data: First, using scatter plotting, we compared log-transformed expression data of autotetraploids with those of diploids in the same cultivar background to find ESTs whose transcription levels alter between autotetraploids (4Xs) and diploids (2Xs); Second, we examined whether we would be able to identify ESTs common to different cultivar backgrounds, whose transcription levels alter between 4Xs and 2Xs. Third, using normalized expression data by the method described in Kishimoto et al. (2004), we tabulated the expression data on ESTs common to different cultivar backgrounds, whose transcription levels alter between 4Xs and 2Xs.

By scatter plotting analyses, we found a group of genes whose transcription levels alter between 4Xs and 2Xs. Because each slide in this microarray system carries duplicated EST spots (Left field and Right field) and gives us two data set for 4512 ESTs, we were able to perform scatter plotting analyses of 4512 ESTs in 4 combinations of expression data for each cultivar (4X_Left vs 2X_Left, 4X_Left vs 2X_Right, 4X_Right vs 2X_Left, 4X_Right vs 2X_Right). The scatter plotting analyses detected a group of ESTs (14 ESTs) showing distinct differential up-regulation in 4Xs of "Shinriki", "Dee-geo-woo-gen" and "Surjamkhi" (Fig. 1). We found no EST group showing distinct differential down-regulation common to the autotetraploids.

The EST group found by scatter plotting analyses in the three cultivars also showed up-regulation in another cultivar, "Shensho" (Table 1) though scatter plotting analysis of "Shensho" data did not detect the EST group as a distinct group. The other cultivar "Dular", on the other hand, did not show up-regulation in almost all of the 14 ESTs (Table 1).

Results of Blast search and putative functions for the 14 ESTs are summarized in Table 1. There was no trend in gene function and chromosomal location. For 4 ESTs showing perfect match to organellar DNAs, at least one full length cDNA clone (FL cDNAs) corresponding to each EST has already been isolated, meaning the 4 ESTs would be coded in nuclear genome.

For two ESTs (E_13 and E_14 in Table 1) showing perfect match to 17S and 25S ribosomal RNA genes (rDNAs), their FL cDNAs with poly(A) tails were already found and genomic regions identical to their sequences are located on more than two chromosomes (it is known that ribosomal RNAs do not have poly(A) tails). In addition, although the first slide of this cDNA microarray system contains 4 ESTs and 2 ESTs showing high similarity with more than E-100 (as E-value) to 17S and 25S rice rDNAs, respectively, only two ESTs (E_13 and E_14) showed differential up-regulation in autotetraploids, meaning the other 4 ESTs showing high similarity to 17S and 25S rDNAs did not showed differential up-regulation (All of the 6 ESTs showing high similarity to 17S and 25S rDNAs could have shown differential up-regulation, if the rDNAs had showed up-regulation in autotetrapoids). Getting together, these results suggests the two ESTs (E_13 and E_14) could be coded in genes different from rDNAs and the differential up-regulations of the two ESTs could be due to autotetraploidy. In order to confirm these, a further study should be done using mRNA since total RNAs were used in this study.

In addition, we compared expression data obtained in this study on the ESTs in Tables 1A and 1B by Kishimoto et al. (2005). In short, no EST in Table 1A and 1B in the previous study showed clear changes in expression in this study. Of the ESTs in Tables by Kishimoto et al. (2005), 14 ESTs are spotted on the first slide in this microarray system. In this study, although we examined 70 expression ratio data points [70 data points = 5 cultivars in this study x 14 ESTs (on the first slide) in Table 1A and 1B of Kishimoto et al. (2005)], 56 data points and 66 data points (80% and 94%) showed values less than 2-fold changes and less than 2.5-fold changes, respectively, meaning that all of the ESTs showed expression alterations in the previous study did not showed clear expression changes in this study. It's conceivable that the fluctuations of expression level in the previous study and the discrepancy between the previous study and this study might be caused by differences in the microenvironment for plant growth, sampling stage and latent disease infection (After sampling, some of remaining seedlings in some lines showed disease symptoms similar to those of wilt or blight).

We have to confirm the data of ESTs, which showed up-regulation in 4Xs in this study, by RT-PCR or Northern analysis. It also remains to be solved whether this is due to difference of autotetraploidy or due to limited number of lines used in this study.

References

Kishimoto N., F. Fujii, Y. Sato, K. Takeuchi, K. Toyoshimao, K. Shimbo, Z. Shimatani, Y. Nagata, A. Hashimoto, M. Ishikawa, J. Yazaki, S. Honda, K. Suzuki, K. Kojima, K. Yamamoto, K. Sakata, T. Sasaki and S. Kikuchi, 2005. Preliminary screening of rice genes whose expression levels alter by autotetraploidy using cDNA microarrays. RGN 22: 86-87.

Kishimoto N., J. Yazaki, F. Fujii, K. Nakamura, K. Shimbo, Y. Otsuka, Y. Otake, K. Yamamoto, K. Sakata, T. Sasaki and S. Kikuchi, 2004. Microarray analysis of transcription patterns in rice: Comparison of differential expression among different tissues/organs. RGN 21: 78-80.

Yazaki J., N. Kishimoto, K. Nakamura, F. Fujii, K. Shimbo, Y. Otsuka, J. Wu, K. Yamamoto, K. Sakata, T. Sasaki and S. Kikuchi, 2000. Embarking on rice functional genomics via cDNA microarray: use of 3' UTR probes for specific gene expression analysis. DNA Res. 7: 367-370.