Reciprocal regulation of glycine-rich RNA-binding proteins via an interlocked feedback loop coupling alternative splicing to nonsense-mediated decay in Arabidopsis

Abstract

The Arabidopsis RNA-binding protein AtGRP8 undergoes negative autoregulation at the post-transcriptional level. An elevated AtGRP8 protein level promotes the use of a cryptic 5' splice site to generate an alternatively spliced transcript, as_AtGRP8, retaining the 5' half of the intron with a premature termination codon. In mutants defective in nonsense-mediated decay (NMD) abundance of as_AtGRP8 but not its pre-mRNA is elevated, indicating that as_AtGRP8 is a direct NMD target, thus limiting the production of functional AtGRP8 protein. In addition to its own pre-mRNA, AtGRP8 negatively regulates the AtGRP7 transcript through promoting the formation of the equivalent alternatively spliced as_AtGRP7 transcript, leading to a decrease in AtGRP7 abundance. Recombinant AtGRP8 binds to its own and the AtGRP7 pre-mRNA, suggesting that this interaction is relevant for the splicing decision in vivo. AtGRP7 itself is part of a negative autoregulatory circuit that influences circadian oscillations of its own and the AtGRP8 transcript through alternative splicing linked to NMD. Thus, we identify an interlocked feedback loop through which two RNA-binding proteins autoregulate and reciprocally crossregulate by coupling unproductive splicing to NMD. A high degree of evolutionary sequence conservation in the introns retained in as_AtGRP8 or as_AtGRP7 points to an important function of these sequences.

Figure 1.

Influence of ectopic AtGRP8 overexpression on the endogenous AtGRP7 and AtGRP8. (A) The immunoblot with total protein of WT plants, the AtGRP8-ox lines 5 and 21 and an AtGRP7-ox line, harvested at zt11, was probed with the AtGRP8 antibody (top) and an antibody against LHCP (light harvesting chlorophyll-binding protein) as loading control (bottom). The absence of crossreaction with overexpressed AtGRP7 protein in AtGRP7-ox plants demonstrates the specificity of the antibody. (B) A duplicate blot was probed with the AtGRP7 antibody (top) and the LHCP antibody (bottom). (C) WT and AtGRP8-ox plants were harvested at zt3 and zt11. The RNA gel blot was hybridized with the AtGRP8 cDNA to determine the total AtGRP8 transcript level. The stripped blot was rehybridized with the gene-specific probe to monitor the endogenous AtGRP8 transcript and subsequently with the gene-specific AtGRP7 probe. The position of the pre-mRNA, as_AtGRP8 and as_AtGRP7 retaining the first half of the intron and the mature mRNA are indicated. Boxes represent exons, lines represent the first and second half of the intron, respectively. The ethidium-bromide stained gel shows equal loading.

Figure 2.

Binding of recombinant AtGRP8 to the AtGRP8 transcript. (A) GST-AtGRP8 protein was incubated with labelled 8-intron_WT ORN in the presence of 5 µg of tRNA and 2, 20, 100 and 250 pmol of unlabelled 8-intron_WT ORN (lanes 3–6) or 8-intron_G₂U₂mut (lanes 7-10), respectively. Lane 1, free ORN. (B) GST-AtGRP8 protein was incubated with labelled 8-UTR_WT ORN in the presence of 5 µg of tRNA and 2, 20, 100 and 250 pmol of unlabelled 8-UTR_WT ORN (lanes 3–6) or 8-UTR_G₆mut (lanes 7-10), respectively. Lane 1, free ORN. (C) To compare binding affinities, 50 fmol of labelled 8-UTR_WT ORN or 8-UTR_938 were incubated with 0.01, 0.05, 0.1, 0.25, 0.5, 1, 2, 2.5, 5 and 10 µM or 0.1, 0.25, 0.5, 1, 2, 3, 4, 5 and 10 µM of GST-AtGRP8, respectively. All reactions contained 1 µg tRNA. Bound and free RNA were quantified and K_d values calculated based on the mean of three independent experiments as described (12). (D) GST-AtGRP8 protein was incubated with labelled 8-UTR_938 ORN in the presence of 5 µg of tRNA and 2, 20, 100 and 250 pmol of unlabelled 8-UTR_938 ORN (lanes 2–5) or 7-UTR_ G₄mut (lanes 7–10), respectively. Lane 6, free ORN.

Figure 3.

Binding of recombinant AtGRP8 to the AtGRP7 transcript. (A) GST-AtGRP8 protein was incubated with labelled 7-UTR_WT ORN in the presence of 5 μg of tRNA and 2, 20, 100 and 250 pmol of unlabelled 7-UTR_WT ORN (lanes 3–6) or 20, 100, 250 and 500 pmol of unlabelled 7-UTR_G₄mut (lanes 7–10), respectively. Lane 1, free ORN. (B) GST-AtGRP8 protein was incubated with labelled 7-intron_WT ORN in the presence of 5 μg of tRNA and 2, 20, 100 and 250 pmol of unlabelled 7-intron_WT ORN (lanes 3–6) or 20, 100, 250 and 500 pmol of unlabelled 7-intron_G₆mut ORN (lanes 7–10), respectively. Lane 1, free ORN.

Figure 4.

Molecular characterization of transgenic AtGRP8-RQ ox lines. (A) WT, AtGRP8-ox and AtGRP8-RQ-ox plants were harvested at zt3 and zt11. The RNA gel blot was hybridized with the AtGRP8 cDNA to determine the total transcript level (top) and with the gene-specific probe to monitor the endogenous AtGRP8 transcript (middle). The position of the pre-mRNA, as_AtGRP8 retaining the first half of the intron and the mRNA are indicated. Boxes represent exons, lines represent the first and the second half of the intron, respectively. The ethidium-bromide stained gel shows equal loading (bottom). (B) The immunoblot with total protein of WT plants, the AtGRP8-ox lines 5 and 21, and the AtGRP8-RQ-ox lines 1, 4 and 5, harvested at zt11, was probed with the AtGRP8 antibody (top) and an antibody against LHCP as loading control (bottom). (C) The RNA gel blot shown in (A) was stripped and rehybridized with the AtGRP7 probe. (D) The immunoblot with the same protein extracts as shown in (B) was probed with the AtGRP7 antibody (top) and the LHCP antibody (bottom).

Figure 5.

Differential regulation by cold of AtGRP8 and AtGRP7. (A) Col plants grown for 2 weeks in 16 h light/8 h dark cycles at 20°C were transferred to 16-h light/8-h dark cycles at 4°C and harvested on day 0, 1, 2, 4 and 7 at zt2 and zt10, respectively. (B) Immunoblots with total protein extracts from the same plants were probed with the AtGRP8 (A, top) and AtGRP7 (B, top) antibodies and an LHCP antibody (A and B, bottom) as loading control. (C) Semiquantitative RT-PCR of AtGRP8. (D) Semiquantitative RT-PCR of AtGRP7. (E) Semiquantitative RT-PCR of eIF-4A as constitutive control. The exponential range was determined by comparing the signal with increasing number of cycles. The absence of genomic DNA was confirmed with nonretrotranscribed RNA.

Figure 6.

Effect of upf1 and upf3 mutations on steady-state abundance of the AtGRP8 and AtGRP7 pre-mRNAs. RNA from the upf1-5, upf3-1 and upf3-2 mutants and WT harvested at zt10 was reverse-transcribed. PCR amplification of the AtGRP8 (A) and AtGRP7 pre-mRNA (B) was performed using specific primers and 24 cycles. The gel with the PCR products was blotted and hybridized with the AtGRP8 cDNA (A) or AtGRP7 cDNA (B). Amplification with ACTIN primers served as control (C).

Figure 7.

Conservation of the retained part of the GRP introns. The nucleotide sequence corresponding to AtGRP8 exon I, exon II and the intron was aligned with the corresponding sequence of AtGRP7, SaGRP1 (8), BnGRP10 (46), NsGRP1a (47) and PhGRP2 (48).

Figure 8.

Model of the interlocked AtGRP7 and AtGRP8 feedback loops. Increasing AtGRP7 and AtGRP8 protein (depicted as ellipses) levels promote use of the cryptic intronic 5′ splice sites, leading to unproductively spliced as_AtGRP7 and as_AtGRP8 transcripts that are degraded via the NMD pathway.