Evolutionary History and Activity of RNase H1-Like Proteins in Arabidopsis thaliana

Abstract

RNase H1 is an endonuclease specific toward the RNA strand of RNA:DNA hybrids. Members of this protein family are present in most living organisms and are essential for removing RNA that base pairs with DNA. It prevents detrimental effects of RNA:DNA hybrids and is involved in several biological processes. Arabidopsis thaliana has been previously shown to contain three genes encoding RNase H1 proteins that localize to three distinct cellular compartments. We show that these genes originate from two gene duplication events. One occurred in the common ancestor of dicots and produced nuclear and organellar RNase H1 paralogs. Second duplication occurred in the common ancestor of Brassicaceae and produced mitochondrial- and plastid-localized proteins. These proteins have the canonical RNase H1 activity, which requires at least four ribonucleotides for endonucleolytic digestion. Analysis of mutants in the RNase H1 genes revealed that the nuclear RNH1A and mitochondrial RNH1B are dispensable for development under normal growth conditions. However, the presence of at least one organellar RNase H1 (RNH1B or RNH1C) is required for embryonic development. The plastid-localized RNH1C affects plastid DNA copy number and sensitivity to replicative stress. Our results present the evolutionary history of RNH1 proteins in A. thaliana, demonstrate their canonical RNase H1 activity and indicate their role in early embryonic development.

Keywords: Arabidopsis thaliana; RNase H; organellar genomes.

Fig. 1

Phylogeny of plant RNase H1 proteins. Phylogenetic tree of all full-length predicted RNase H1 proteins. Support values at tree branches are posterior probability scores, which integrate over the uncertainty in both the alignment and the phylogeny. Predicted protein localizations obtained using TargetP are marked with colors. Mitochondria—red, chloroplast—green, other—blue. A version of this phylogenetic tree with all species names, sequence IDs and support values is shown in Supplementary Fig. S1.

Fig. 2

Evolutionary history of four RNase H1 proteins in Arabidopsis thaliana. (A) Predicted localization of RNase H1 proteins from distinct clades identified in Fig 1. Predictions were performed using TargetP. (B) Detailed phylogenetic tree of all RNase H1 proteins in Brassicaceae. Support values and color-coded predicted localization are used as described for Fig 1. (C) Splice variants of RNH1B predicted in Araport11 (AT5G51080.1) and identified by 5′ RACE (AT5G51080.2 and AT5G51080.3). (D) 5′ RACE of RNH1B. Asterisk indicates a band, which is likely to be nonspecific. (E) Domain composition of four RNase H1 proteins in A. thaliana. MTS, mitochondrial targeting sequence (presequence); CTS, chloroplast targeting sequence (transit peptide), RNH1, RNase H1 domain. (F) Subcellular localization of GFP-tagged RNH1A in transiently transformed Arabidopsis protoplasts. GFP, Hoechst and chlorophyll are represented in green, blue and red, respectively. (G) Subcellular localization of GFP-tagged RNH1B.3 in transiently transformed Arabidopsis protoplasts. GFP, MitoTracker and chlorophyll are represented in green, magenta and red, respectively. Scale bars indicate 10 μm. Additional cells transformed with RNH1B.3-GFP are shown in Supplementary Fig. S2B.

Fig. 3

In vitro activity assays for RNH1A and RNH1B proteins from Arabidopsis thaliana. Wild-type RNH1A (A), catalytic mutant of RNH1A (B), wild-type RNH1B (C) and wild-type human RNase H1 (D) (positive control) were incubated with fluorescently labeled oligonucleotide substrates containing different combinations of deoxyribonucleotides (black) and ribonucleotides (red). The triangles indicate increasing protein:substrate molar ratio. The reaction products were analyzed on TBE-urea PAGE and scanned for fluorescence. The experiment was repeated three times.

Fig. 4

Effects of mutations in genes encoding RNase H1 proteins. Approximately 3-week-old plants of Col-0 wild type (A), atrnh1a (B), atrnh1b (C), atrnh1c (D), atrnh1a, atrnh1b double mutant (E), atrnh1a, atrnh1c double mutant (F), atrnh1a, atrnh1b double mutant heterozygous for atrnh1c (G) and atrnh1a, atrnh1b, atrnh1c triple mutant expressing RNH1C::RNH1C-GFP (H). (I) Immunoprecipitation and detection of GFP-tagged RNH1C expression in triple atrnh1a, atrnh1b, atrnh1c mutant background. The bottom panel shows nonspecific staining with Ponceau S. (J) Relative chlorophyll content of Col-0 wild type and mutants in genes encoding RNase H1 proteins. Error bars indicate the standard deviation from three biological replicates. Asterisks indicate P < 0.05 obtained from t-test. (K) Schematic representation of RNH1 genes in Arabidopsis thaliana. Arrows indicate START codons, and triangles indicate T-DNA insertion positions. Red boxes indicate positions of the conserved RNase H domain. (L, M) Expression of genes encoding RNases H1 measured (L) upstream of T-DNA insertion and (M) downstream of T-DNA insertion. Primer pairs are marked in (K). Error bars indicate 95% confidence intervals. Asterisks indicate P < 0.05 in comparison to Col-0 wild type.

Fig. 5

At least one organellar RNase H1 is required for embryo development. (A) Mature silique of Col-0 wild type. (B) atrnh1a, atrnh1b double mutant heterozygous for atrnh1c (atrnh1a/b/c^+/−). (C) atrnh1a, atrnh1b, atrnh1c triple mutant expressing RNH1C::RNH1C-GFP. Scale bars correspond to 1 mm. (D) Percentage of properly developing seeds in siliques. Error bars indicate standard deviation from at least six independent siliques. (E) Development of embryos in normally developing seeds obtained from atrnh1b^−/− atrnh1c^+/− parents. (F) Development of embryos in abnormally developing seeds obtained from atrnh1b^−/− atrnh1c^+/− parents. Embryos are marked with dotted lines.

Fig. 6

Loss of RNH1C leads to increased chloroplast DNA content and hypersensitivity to replication stress. Relative levels of nuclear (A), mitochondrial (B) and chloroplast (C) DNA in Col-0 wild type and atrnh1 mutants. Error bars indicate 95% confidence intervals. Asterisks indicate P < 0.05 in comparison to Col-0 wild type. Phenotype of plants subjected to HU (D), Aph (E) and mock (F) treatments. (G, H) Relative root length of Col-0 wild type and atrnh1 mutants upon 2 mM HU (F) and 40 µM Aph (G) treatments. Asterisks indicate P < 0.0001determined using two-way Analysis of Variance (ANOVA).