A microRNA superfamily regulates nucleotide binding site-leucine-rich repeats and other mRNAs

Abstract

Analysis of tomato (Solanum lycopersicum) small RNA data sets revealed the presence of a regulatory cascade affecting disease resistance. The initiators of the cascade are microRNA members of an unusually diverse superfamily in which miR482 and miR2118 are prominent members. Members of this superfamily are variable in sequence and abundance in different species, but all variants target the coding sequence for the P-loop motif in the mRNA sequences for disease resistance proteins with nucleotide binding site (NBS) and leucine-rich repeat (LRR) motifs. We confirm, using transient expression in Nicotiana benthamiana, that miR482 targets mRNAs for NBS-LRR disease resistance proteins with coiled-coil domains at their N terminus. The targeting causes mRNA decay and production of secondary siRNAs in a manner that depends on RNA-dependent RNA polymerase 6. At least one of these secondary siRNAs targets other mRNAs of a defense-related protein. The miR482-mediated silencing cascade is suppressed in plants infected with viruses or bacteria so that expression of mRNAs with miR482 or secondary siRNA target sequences is increased. We propose that this process allows pathogen-inducible expression of NBS-LRR proteins and that it contributes to a novel layer of defense against pathogen attack.

Figure 1.

Sequence Diversity of miR482. (A) Multiple alignment of diverse miR482 family members. Alignment of six unique miR482 sequences identified from tomato small RNA data sets. (B) Alignment of a tomato miR482 isoform with atmiR472. (C) Alignment of all known miR482 isoforms. Alignments were made with ClustalX2, and miR482 sequences were taken from miRBase (http://www.mirbase.org/). Sequences shown in red are not conserved among miRNAs. sl, S. lycopersicum; ath, Arabidopsis; gh, G. hirsutum; vv, V. vinifera; pab, Picea abies; ag, Aquilegia caerulea; cs, Citrus sinensis; gma, G. max; gsi, G. soja; pt, Populus trichocarpa; pta, Pinus taeda; stu, S. tuberosum; pv, Phaseolus vulgaris; mdm, Malus domestica; gr, G. raimondii.

Figure 2.

Abundance of miR2118/482 Superfamily Members in Different Plants. (A) Small RNA data sets were accessed through GEO, and miRNA abundance was analyzed through miRProf (Moxon et al., 2008) and expressed as counts per million reads. miR482 and 2118 sequences are defined in the text. Samples were from leaf/shoot tissue (green), floral tissue (orange), panicle/tassel (red), and other (black). Asterisk indicates those plant species where miRNA family members were not cloned. PHP, Physcomitrella patens; SEL, Selaginella; MAR, Marsilea; CYC, Cycas; GIB, Gingko biloba; PAB, Picea abies; ARF, Aristolochia fimbriata; NAD, Nuphar advena; PEA, Persea americana; ATH, Arabidopsis; SIL, Silene latifolia; VVI, V. vinifera; CAA, Capsicum annum; NIT, Nicotiana tabacum; NBE, N. benthamiana; PET, Petunia hybrida; SLY, S. lycopersicum; SPE, S. pennellii; SPI, S. pimpinellifolium; STU, S. tuberosum; LAS, Lactuca sativus; MIM, Mimulus; MTR, Medicago truncatula; GMA, G. max; PVU, Phaseolus vulgaris; CMA , Cucurbita maxima; PTR, Populus trichocarpa; CPA, Carica papaya; CSI, Citrus sinensis; GAR, Gossypium arboreum; HVU, Hordeum vulgare; TAE, Triticum aestivum; MIS, Miscanthes; SBI, Sorghum bicolor; ZMA, Zea mays; OSA, O. sativa; PVI, Panicum virgatum; SIT, Setaria italica; MAC, Musa acuminata; ZOM, Zostera marina. (B) RNA gel blot analysis of tomato miR482 isoforms and cross-hybridizing homologs in different species. Fifteen micrograms of total RNA from young seedlings was used for analysis of each sample. Total RNA was electrophoresed in a 15% polyacrylamide gel, transferred to membrane, and probed with corresponding DNA oligonucleotides (see Supplemental Table 1 online) labeled with [γ-³²P]ATP. U6 serves as loading control. M, decade (Ambion) size marker.

Figure 3.

Predicted Targets of miR482 Isoforms in Tomato. (A) Predicted miR482 targets among the 186 NBS-LRR sequences in tomato. (B) Representation of miR482 targets in the tomato genome as either unique or repeated sequence motifs. (C) Coding sequence of the miR482 targets in the mRNAs of NBS-LRR mRNAs. (D) Variable residues in the miR482 family correspond to wobble or variable sequences in NBS-LRR mRNAs. Target sequences were predicted using the TAPIR algorithm.

Figure 4.

CNL-Type NBS-LRR mRNAs Are Preferred Targets of miR482 Family Members. (A) The targets for tomato miR482 were identified among the Arabidopsis NBS-LRR mRNA sequences and classified as either CNL or TIR type. (B) The CNL-type NBS-LRRs were classified as CNL types A to D according to Meyers et al. (2003). Each type is represented similarly in the total and miR482 targets all except that the CNL-D types are underrepresented in the targets (left panel).

Figure 5.

Phased Secondary siRNAs Initiated by miR482 Target Sites in an NBS-LRR mRNA. Genome view of a CNL-type NBS-LRR locus (LRR1) that aligns with phased secondary siRNAs. Most of the secondary siRNAs align between miR482 target sites. Red bars indicate the start position of sRNAs in phase; blue bars indicate the start position for those out of phase. Rectangles indicate expected phased positions and those in red indicate phased sequences. Orange box indicates position of P-loop. Numbers next to the miR482 target site indicate frequency of 5′ RACE clones matching this prediction out of total clones used for analysis. Arrow color code: red, 21 nucleotides; blue, 24 nucleotides; green, 22 nucleotides. The black triangle indicates cleaved site, and the white triangle indicates proposed cleavage site.

Figure 6.

Transient Expression of Tomato miR482a Targets an NBS-LRR mRNA in N. benthamiana Resulting in RDR6-Dependant Phased siRNA Biogenesis. (A) Targeting of Nicotiana EU713768.1 NBS-LRR by miR482a. The alignment shows the target site of miR482a in a tobacco NBS-LRR mRNA, and the predicted sequences of phased secondary siRNAs are shown at an adjacent site D1. Δ indicates modified sequences underlined in the miR482aΔ construct. (B) RNA gel blot analysis of miR482a and D1 siRNA in Agrobacterium tumefaciens infiltrated zones of N. benthamiana wild-type (WT) and RDR6i lines. Analyses of miR168 and U6 RNAs are included as controls. M, decade marker indicating a 20-nucleotide RNA. (C) U713768.1 mRNA accumulation 3 d after inoculation by quantitative PCR. See Methods for construct design. Error bars indicate sd (n = 3), and asterisks indicate a significant difference from corresponding control samples (t test, P value < 0.05).

Figure 7.

miR482 Silencing of NBS-LRR mRNAs Is Suppressed in Virus-Infected Plants. (A) RNA gel blot analysis of miR482 in TCV-infected tomato leaves. Total RNA from TCV infected and noninfected leaves of N. benthamiana was inoculated to M82 tomato (TCV and Mock). RNAs isolated at the indicated times after inoculation were separated on a 15% polyacrylamide gel. The RNA was transferred to a membrane and probed with radiolabeled DNA oligonucleotides for miR482 with miR168 and U6 probes as loading controls. M, decade marker indicating a 20-nucleotide RNA. (B) RNA gel blot analysis of miR482 in TRV- and CMV-infected tomato leaves. Total sap from TRV infected, CMV infected, and noninfected leaves of N. benthamiana were inoculated to M82 tomato (TRV, CMV, and Mock, respectively). RNAs isolated at the indicated times after inoculation were separated on a 15% polyacrylamide gel. The RNA was transferred to a membrane and probed with radiolabeled DNA oligonucleotides for miR482 with miR168 and U6 probes as loading controls. 4d sy, tissue from systemic uninoculated leaf; M, decade marker (Ambion) indicating a 20-nucleotide RNA. (C) Quantitative PCR analysis for the abundance of two miR482 target mRNAs (LRR1 and LRR2) (Top) and Mi and Hero NBS-LRRs that are not targets of miR482 (Bottom; see [D]). RNA was extracted 4 h after inoculation. Error bars indicate sd (n = 3), and asterisks indicate a significant difference from corresponding control samples (t test, P value < 0.05). (D) The best target sequences of miR482 with Mi and Hero mRNAs.

Figure 8.

miR482 Silencing of NBS-LRR mRNAs Is Suppressed in Plants Infected with P. syringae DC3000. (A) Symptoms produced by the wild type or hrcC mutant P. syringae DC3000 in 4-week-old leaves of tomato at 2 weeks after inoculation. (B) RNA gel blot analysis of miR482 accumulation in mock-, DC3000-, and hrcC-inoculated leaves at various time points. M, decade size marker. (C) Quantitative PCR analysis for the abundance of two miR482 target mRNAs (LRR1 and LRR2) (top panel) and Mi and Hero NBS-LRRs that are not targets of miR482 (bottom panel). RNA was extracted 4 h after inoculation. Error bars indicate sd (n = 4), and asterisks indicate a significant difference from corresponding control samples (t test, P value < 0.05).

Figure 9.

Natural Variation in Secondary siRNAs Derived from LRR1. (A) Genome browser view of secondary siRNAs aligned to LRR1 mRNA. The miR482 target site at the 5′ end of the gene is indicated. The circled region shows the absence of D2+siRNA in S. pennellii. (B) Bar chart showing accumulation of phased secondary siRNAs in two cultivars of tomato (M82 and MicroTom) and two wild species (S. pennellii and S. pimpinellifolium). D1 to D15 indicate the different positions in the phased register that is established by the miR482 targeting event. (C) RNA gel blot analysis of D2+ siRNA in tomato, S. pennellii, their F1 and F2 hybrids, and in introgression lines (ILs). The D2+ probe detected a 24- and 21-nucleotide species (top and bottom signal) of which only the 21-nucleotide species was absent in S. pennellii and the F1 and F2 samples. M, decade size marker. (D) Predicted target for the D2+ siRNA derived from LRR1. (E) Quantitative PCR analysis of abundance of D2+ target gene between M82 and S. pennellii. Error bars indicate sd (n = 6), and asterisks indicate a significant difference from corresponding control samples in S. pennellii (t test, P value < 0.05). (F) Genetic map of IL2 series introgression lines (Eshed and Zamir, 1995). The S. pennellii genomic regions in each IL are marked. (G) RNA gel blot analysis of D2+ siRNA in tomato, S. pennellii, and the IL2 series introgression lines. (H) Accumulation of the D2+ target mRNA in tomato, S. pennellii, and the IL2 series introgression lines. Error bars indicate sd (n = 3), and asterisks indicate a significant difference from corresponding control samples in S. pennellii and IL2-1 (t test, P value < 0.05).