Overexpression of the mutated EMB2279/SOT5 and EMB2654 CDS in WT leads to RDR6-dependent gene silencing
We previously reported that a weak allele of emb2279-2/sot5 mutant exhibits a virescent phenotype, which is caused by a point mutation that significantly reduces splicing efficiency of the seventh intron of SOT5 and generates two additional mRNA variants [8]. The smallest transcript that lacks 22-base pairs (bp) at the 3' end of the seventh exon is predicted to produce a truncated SOT5 protein with only 6 PPR motifs, named SOT5-m1 (Fig.1), whereas the largest transcript that contains the seventh intron is predicted to produce a mutated protein with 10 PPR motifs, named SOT5-m2 (Fig.1). While the wild type SOT5 encodes a protein with 11 PPR motifs (Fig.1 and Table 1). To test whether the two predicted proteins are functional or not in plants, we cloned SOT5-m1 and SOT5-m2 CDS and transformed them under the control of the cauliflower mosaic virus 35S promoter into WT plants. We obtained 30 and 42 positive T1 transformants for 35S:SOT5-m1 and 35S:SOT5-m2 constructs, respectively. Our results showed that 80% of the 35S:SOT5-m1 transgenic lines exhibited severe leaf chlorosis at the early growth stage, and these leaves were gradually turned into pale green later; and 48% of 35S:SOT5-m2 transgenic lines displayed relatively mild leaf chlorosis and virescence (Table 1 and Fig. 2a). This chlorosis and virescent phenotype was similar to that of the sot5 mutant. Thus, we suspected that the transgenic lines with chlorosis were the cosuppression lines, in which expression of the endogenous SOT5 was silenced. Then, we analyzed expression levels of SOT5 in T1 transformants with chlorosis. Indeed, RT-PCR and quantitative PCR (qPCR) analyses showed that the level of the endogenous SOT5 transcripts (En-SOT5), detected by the specific primer pair spanning 5’UTR (the red arrows highlighted on EMB2279/SOT5 cDNA in Fig.1), was significantly decreased in 35S:SOT5-m1/Col-0 transgenic line2 and line3 (less than 40% of the WT expression level), whereas the level of total SOT5 transcripts (including endogenous and exogenous SOT5 in transgenic lines), detected by the primer pair spanning intron 7 (Fig. 2b) or by the primer pair in CDS (black arrows highlighted on EMB2279/SOT5 cDNA in Fig.1) was significantly increased (Fig. 2c), compared with that in non-transgenic WT plants. Consistently, in 35S:SOT5-m2/Col-0 transformants, the expression level of endogenous SOT5 was also significantly decreased (less than 60% of the WT expression level, Fig. 2b and 2c), while the content of total SOT5 transcripts was significantly increased (Fig. 2d). Meanwhile, splicing efficiency of the plastid rpl2 gene, a SOT5 target, was dramatically reduced in these transformants (Fig. 2b and 2e). In contrast, splicing efficiency of the plastid atpF gene, a non-target gene of SOT5, was not altered in all transgenic plants (Fig. 2b and 2e). These results indicate that overexpression of the mutated SOT5 CDS in WT leads to suppression of the endogenous gene expression, probably through the post transcriptional gene silencing (PTGS, also known as sense cosuppression) pathway.
To confirm this hypothesis, we transformed 35S:SOT5-m1 and 35S:SOT5-m2 constructs into the rdr6-11 mutant. RDR6 is a key component in the PTGS pathway and the rdr6-11 mutant presented elongated and curled downward leaves (Fig. 3a) [21, 22]. It has been demonstrated that overexpression of homologous genes in this mutant was not able to trigger the PTGS pathway [21-23]. We obtained twenty-five 35S: SOT5-m1/rdr6 and eighteen 35S: SOT5-m2/rdr6 T1 transformants, and found that none of the transformants exhibited the chlorosis phenotype (Fig. 3a). RT-PCR analysis showed that the endogenous SOT5 transcript level was not decreased, although the total SOT5 mRNA level was significantly increased in the transformants (Fig.3b-3d). Consequently, no splicing defect of plastid rpl2 was detected in these transformants (Fig. 3e). Taken together, our results indicate that overexpression of the mutated SOT5 CDS in the WT background leads to PTGS in a RDR6-dependent manner.
We then asked whether the above result could be repeated with other EMB PPR genes. To address this question, we overexpressed a truncated CDS of EMB2654 (EMB2654-11M) which encodes a truncated protein with only 11 PPR motifs into the WT background (Fig. 1 and Table 1). While the wild type EMB2654 has 18 PPR motifs. It has been reported that EMB2654, a P family PPR protein, was required for trans-splicing of the plastid gene rps12 intron 1 [10]. Interestingly, all the 47 T1 transgenic lines exhibited leaf chlorosis (Table1 and Fig. 4a), indicating the silence of endogenous EMB2654. RT-PCR analysis showed that expression of the endogenous EMB2654 was down-regulated (about 20% of the WT expression level) while the total mRNA of EMB2654 was significantly increased in these transformants (Fig. 4b-d). Indeed, the splicing efficiency of rps12 intron 1 was significantly decreased in the cosuppression lines, while the splicing efficiency of the non-target gene, clpP1 intron2 was not significantly decreased (Fig. 4e). However, when the truncated EMB2654-11M CDS was overexpressed in rdr6-11, no leaf chlorosis was observed among 30 transformants (Fig. 4f). RT-PCR results showed that the endogenous EMB2654 mRNA was not decreased, and the total EMB2654 mRNA was significantly increased in these transformants (Fig.4g-4i). Consistently, splicing efficiency of the rps12 intron 1 was not decreased in these transformants (Fig. 4j). Thus, these results confirm that overexpression of the truncated EMB2654 CDS in the WT background leads to PTGS in a RDR6-dependent manner.
Functional analysis of EMB976 via its cosuppression lines
EMB976 is a functionally unknown PPR protein, which belongs to P subfamily containing 22 PPR motifs and is predicted to be localized in chloroplasts. Its knockout mutant has been demonstrated to be embryonically lethal [2]. To study its physiological role in plant growth, we constructed two plasmids named EMB976-7M and EMB976-14M, which encode the truncated protein with 7 and 14 PPR motifs, respectively (Fig.1 and Table 1), and transformed them into the WT background. Our results showed that four of eight 35S:EMB976-14M/Col-0 T1 transformants displayed virescent leaves (Fig. 5a and Table 1), whereas all of the thirteen 35S:EMB976-7M//Col-0 transformants had the same phenotype as WT (Table 1). To confirm whether gene silencing occurred in 35S:EMB976-14M/ Col-0, we checked expression levels of the endogenous EMB976 in the transformants with the virescent phenotype. As expected, the endogenous EMB976 was significantly down-regulated (about 40% of the WT expression level) while the total EMB976 mRNA was significantly increased in these transformants (Fig.5b and 5c), indicating that the phenotype of the transgenic lines is caused by the silencing of EMB976. Since the P family PPR proteins were often involved in organelle RNA stability and splicing, we further examined the intron splicing of chloroplast genes in these 35S:EMB976-14M/ Col-0 cosuppression lines. Indeed, RT-PCR analysis showed that the precursors of ndhA, clpP1 intron 2 and ycf3 intron 1 were significantly and specifically accumulated in these lines (Fig. 5d). qPCR analysis showed that the splicing efficiency of clpP1 intron 2 and ycf3 intron 1 was significantly and specifically decreased in these lines, compared with that of ycf3 intron 2 (Fig. 5e), suggesting chloroplast clpP1 intron 2 and ycf3 intron 1 were the possible targets of EMB976. However, further experiments are required for verification of the results. Again, no yellow young leaves were observed among the transformants when EMB976-14M was overexpressed in rdr6-11 (Fig. 5f). This result was consistent with that of RT-PCR analysis. In these transformants, the endogenous EMB976 transcripts was not significantly decreased while the total EMB976 transcripts was strongly increased (Fig. 5g-5i). Consistently, the splicing efficiency of clpP1 intron 2 and ycf3 intron 1 was not decreased, compared with that of ycf3 intron 2 (Fig. 5j). Thus, we conclude that gene silencing triggered by overexpession of truncated EMB genes is dependent on RDR6.
Table 1 The cosuppression phenotype and frequency of transgenic plants expressing various constructs in WT
Plamid construct a
|
Length of encoding protein (aa)
|
Number of PPR motifs
|
Phenotype of cosuppression
|
Number of total T1 transformants
|
Number of T1 transformants with visible chlorosis
|
Frequency of cosuppressionb
|
Severity of chlorosis in cosuppression lines
|
SOT5
|
978
|
11
|
yellow inflorescence and cauline leaves
|
35
|
20
|
57%
|
very mild
|
SOT5-m1
|
712
|
6
|
albino young leaves at seedling stage
|
30
|
24
|
80%
|
strong
|
SOT5-m2
|
1006
|
10
|
partial albino leaves at seedling stage
|
42
|
20
|
48%
|
mild
|
EMB2654
|
822
|
18
|
ND
|
ND
|
ND
|
ND
|
ND
|
EMB2654-11M
|
550
|
11
|
chlorosis leaves at seedling stage
|
47
|
47
|
100%
|
mild
|
EMB976
|
1038
|
22
|
ND
|
ND
|
ND
|
ND
|
ND
|
EMB976-7M
|
443
|
7
|
WT-like
|
13
|
0
|
0%
|
ND
|
EMB976-14M
|
747
|
14
|
yellow young leaves at seedling stage
|
8
|
4
|
50%
|
mild
|
aThe pGWB2 plasmid containing the mutated or truncated CDS indicated by the name.
b cosuppressed T1 tranfromants/total T1 transformants. ND: not determined.