Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Identification of Nutrient-Responsive Arabidopsis and Rapeseed MicroRNAs by Comprehensive Real-Time Polymerase Chain Reaction Profiling and Small RNA Sequencing   

Identification of Nutrient-Responsive Arabidopsis and Rapeseed MicroRNAs by Comprehensive... Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Identification of Nutrient-Responsive Arabidopsis and Rapeseed MicroRNAs by Comprehensive Real-Time Polymerase Chain Reaction Profiling and Small 1[C][W][OA] RNA Sequencing 2 2 2 2 Bikram Datt Pant , Magdalena Musialak-Lange ,Przemyslaw Nuc ,Patrick May , Anja Buhtz, Julia Kehr, Dirk Walther, and Wolf-Ru¨ diger Scheible* Max-Planck Institute of Molecular Plant Physiology, 14476 Potsdam, Germany Comprehensive expression profiles of Arabidopsis (Arabidopsis thaliana) MIRNA genes and mature microRNAs (miRs) are currently not available. We established a quantitative real-time polymerase chain reaction platform that allows rapid and sensitive quantification of 177 Arabidopsis primary miR transcripts (pri-miRs). The platform was used to detect phosphorus (P) or nitrogen (N) status-responsive pri-miR species. Several pri-miR169 species as well as pri-miR398a were found to be repressed during N limitation, whereas during P limitation, pri-miR778, pri-miR827, and pri-miR399 species were induced and pri-miR398a was repressed. The corresponding responses of the biologically active, mature miRs were confirmed using specific stem-loop reverse transcription primer quantitative polymerase chain reaction assays and small RNA sequencing. Interest- ingly, the latter approach also revealed high abundance of some miR star strands. Bioinformatic analysis of small RNA sequences with a modified miRDeep algorithm led to the identification of the novel P limitation-induced miR2111, which is encoded by two loci in the Arabidopsis genome. Furthermore, miR2111, miR169, a miR827-like sequence, and the abundances of several miR star strands were found to be strongly dependent on P or N status in rapeseed (Brassica napus) phloem sap, flagging them as candidate systemic signals. Taken together, these results reveal the existence of complex small RNA-based regulatory networks mediating plant adaptation to mineral nutrient availability. In recent years, approximately 21-nucleotide-long et al., 2006; Sunkar et al., 2007). MiR395 and miR399 microRNAs (miRs) have been recognized as important have been shown to be specifically induced during regulators of gene expression in animals and plants sulfur and phosphorus (P) limitation, respectively (Bartel, 2004; Dugas and Bartel, 2004; Kidner and (Jones-Rhoades and Bartel, 2004; Fujii et al., 2005; Martienssen, 2005; Chuck et al., 2009). In plants, miRs Bari et al., 2006; Chiou et al., 2006; Kawashima et al., were shown to posttranscriptionally regulate diverse 2008). MiR399 targets the transcript of an E2-conjugase aspects of development like leaf polarity (Emery et al., that is mutated in phosphate (Pi)-accumulating Arabi- 2003), leaf shape (Palatnik et al., 2003), the transition dopsis (Arabidopsis thaliana) pho2 mutants, and this from the juvenile to the mature growth phase (Wu phenotype is recapitulated in miR399-overexpressing and Poethig, 2006), flowering time (Aukerman and plants (Aung et al., 2006; Bari et al., 2006). MiR399 Sakai, 2003), stomatal development (Kutter et al., presumably acts in a dualistic manner to inhibit PHO2. 2007), and nodule development (Combier et al., 2006). It promotes transcript decay (Allen et al., 2005; Bari MiRs also regulate the adaptation of plants to abiotic et al., 2006) but also appears to inhibit PHO2 expres- stresses, including macronutrient limitations (Sunkar sion by repressing translation (Bari et al., 2006), a and Zhu, 2004; Fujii et al., 2005; Bari et al., 2006; Chiou mechanism that is probably widespread in plants (Aukerman and Sakai, 2003; Brodersen et al., 2008). Recently, miR167 has been associated with lateral root This work was supported by the Max-Planck Society. These authors contributed equally to the article. outgrowth in response to nitrogen (N) limitation * Corresponding author; e-mail scheible@mpimp-golm.mpg.de. (Gifford et al., 2008). Primary miR transcript 167a The author responsible for distribution of materials integral to the (pri-miR167a) was shown to be approximately 5-fold findings presented in this article in accordance with the policy repressed by N in root pericycle cells, permitting the described in the Instructions for Authors (www.plantphysiol.org) is: target ARF8 transcript to accumulate and initiate lateral Wolf-Ru¨diger Scheible (scheible@mpimp-golm.mpg.de). root outgrowth. Furthermore, miR398 was reported to [C] Some figures in this article are displayed in color online but in be down-regulated by oxidative stresses and copper black and white in the print edition. [W] limitation but induced by Suc, thereby repressing two The online version of this article contains Web-only data. [OA] copper/zinc (Cu/Zn) superoxide dismutases and a Open Access articles can be viewed online without a sub- scription. cytochrome c oxidase subunit (Sunkar et al., 2006; www.plantphysiol.org/cgi/doi/10.1104/pp.109.139139 Yamasaki et al., 2007; Dugas and Bartel, 2008). Plant Physiology , July 2009, Vol. 150, pp. 1541–1555, www.plantphysiol.org  2009 American Society of Plant Biologists 1541 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Despite these examples, little information about qRT-PCR. Although pri-miRs are not the biologically stress- or nutrient-responsive plant miRs is available. active molecules, several previous studies have shown This is due to their often low expression levels and the that the response of a pri-miR can reflect that of the absence of miR or MIRNA gene probes on widely used encoded mature miR (Jones-Rhoades and Bartel, 2004; transcriptomics platforms such as Affymetrix Gene- Bari et al., 2006; Pant et al., 2008) and thus can serve as Chips. Custom-made microarrays can be designed to a valid indicator. Therefore, qRT-PCR profiling of pri- include probes for miRs and MIRNA genes for a miRs can serve as a useful tool to discover responses to broader response analysis, but these are not very particular stimuli, which can then be confirmed by sensitive (Axtell and Bartel, 2005; Liu et al., 2008). analysis of the mature species. Reverse transcription followed by quantitative PCR In this work, we first established the pri-miR plat- analysis (qRT-PCR) with nonspecific double-stranded form and then used it to discover previously unknown DNA-binding fluorophores, such as SYBR Green, is a nutrient-responsive pri-miRs. The corresponding ma- powerful alternative for highly sensitive, rapid, multi- ture miRs were investigated by targeted assays and parallel, and cost-effective expression analysis (Udvardi further confirmed by small RNA sequencing, which et al., 2008). Shi and Chiang (2005) and Chen et al. also revealed novel insights. The results indicate that (2005) reported two qRT-PCR-based methods to mea- small RNAs play a much more important role in sure the levels of mature miRs. The first approach nutrient signaling than previously thought. relies on in vitro polyadenylation of mature miRs followed by RT with an oligo(dT) adapter primer and RESULTS amplification using SYBR Green with a miR-specific forward primer and a compatible reverse primer. In A qRT-PCR Platform for Arabidopsis pri-miRs the second approach, each specific miR is reverse transcribed from total RNA using a specific stem-loop Sequences of 184 annotated Arabidopsis miR stem primer, followed by TaqMan PCR amplification. Al- loops were obtained from the miRBase database though it is desirable to quantify the biologically active (www.microrna.sanger.ac.uk). These sequences are not mature miR species, a limitation of both qRT-PCR strictly pre-miRs but may include flanking sequence methods is that they are unable to differentiate the from the presumed pri-miR. Pri-miR sequences of expression strengths of MIRNA genes that yield members from the same family can be almost identi- (nearly) identical mature miR molecules. Hence, the cal, complicating the design of specific PCR primers. interpretation of the results will be dominated by This occurs in the miR169 family, where miR169i strongly expressed member(s) of a given MIRNA gene through miR169n are located in three highly homolo- family, while the contribution of lowly expressed gous, tandem-arrayed stretches (Supplemental Fig. members will go unnoticed. This might be especially S1), and in the miR854 family, where the pre-miR important if these genes are expressed in an organ- or sequences of the four annotated members are 97% to cell type-specific manner. 100% identical. However, it is questionable whether Deep sequencing using new technologies (e.g. miR854s are true miRs, as the sequences are located Illumina-Solexa chemistry) is another approach being close to (or in) the centromere of chromosome 5 and adopted for the analysis of small RNA/miR abun- are annotated as transposable elements in The Arabi- dance (German et al., 2008; Moxon et al., 2008a; Szittya dopsis Information Resource (TAIR) database (www. et al., 2008). The high number of sequence reads arabidopsis.org). Therefore, we treated miR854 as a promises sensitivity, yet the necessary expertise re- single miR, leaving a total of 181 sequences for which quired and the labor and cost involved are consider- primers were designed. able. Data from small RNA sequencing together with To ensure maximum specificity and efficiency dur- miR prediction algorithms like miRDeep (Friedla¨nder ing PCR amplification of pri-miR cDNA under a stan- et al., 2008) also provide the basis for the discovery of dard set of reaction conditions (Fig. 1A), a stringent set novel miRs, even in organisms that have been specif- of criteria was used for primer design (see “Materials ically surveyed to identify miRs. and Methods”). PCR primers were tested on cDNA We have developed a qRT-PCR platform for parallel from Arabidopsis wild-type ecotype Columbia (Col-0) analysis of 177 currently known Arabidopsis MIRNA seedlings, which was free of genomic DNA contami- gene pri-miRs. This platform provides a sensitive yet nation. Using this cDNA as template, 150 primer pairs inexpensive tool for Arabidopsis researchers to carry gave unique PCR products of the expected size, while out miR expression analysis. A comparable approach 27 primer pairs yielded no product and four gave has previously been described for monitoring the unspecific products. The 27 primer pairs were retested expression of 23 human miR precursors (Schmittgen using Arabidopsis Col-0 genomic DNA as template. et al., 2004). As pri-miRs are generated by RNA Fifteen primer pairs resulted in the expected genomic polymerase II in plants and animals (Bracht et al., product, showing that the primers anneal to the 2004; Cai et al., 2004; Kurihara and Watanabe, 2004; correct sequence and suggesting that the targeted Lee et al., 2004), they contain 5# caps and 3# poly(A) pri-miRs (e.g. 159c, 166c, 395a to 395f, and 404; Sup- tails. The latter make the transcripts amenable to oligo plemental File S2) were below the detection limit in the (dT)-primed RT and thus multiparallel analysis by cDNA samples or that amplification from the cDNA 1542 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs was inhibited. Evidence in support of the former hypothesis comes from the observation that some of the primers (e.g. pri-miR395a to -395f) did amplify the expected products from a cDNA sample derived from sulfur-limited seedlings (Supplemental Fig. S2). The 16 primer pairs that did not yield any product from cDNA or genomic DNA templates, or that amplified unex- pected/unspecific products, were redesigned, finally resulting in 175 validated primer pairs and only six MIRNA genes (highlighted in red in Supplemental File S2) for which no working pairs could be established. Specificity of PCR primers was assessed by melting curve analysis of PCR products (Supplemental Fig. S3), by separating the PCR products via electrophore- sis on high-resolution agarose gels (Fig. 1B), and by double-stranded sequencing of a subset of the pri-miR PCR products (Supplemental Fig. S4). In all cases, the sequences of the PCR products were identical to those of the intended pri-miR targets. The average amplifi- cation efficiency (E) of the primers, as determined by LinRegPCR (Ramakers et al., 2003), was high; for 102 primer pairs, E was greater than 90%, and for another 39 pairs, E was 81% to 90% (Fig. 1C; Supplemental File S2). Twenty-one primer pairs did amplify a correct product from genomic DNA, but no product was obtained with cDNA from the samples investigated (see below); hence, no E value could be established for these primers. The fraction of pri-miRs detected (i.e. expressed in at least two biological replicates with a threshold fluo- rescence cycle number [C ] of less than 40) was just below 80%, irrespective of the growth conditions tested (Supplemental File S2). This percentage is com- parable to the percentage of transcription factor (TF) genes detected at this threshold (approximately 83%; Czechowski et al., 2004). However, the average DC value of pri-miRs was higher than for the TF gene transcripts (Fig. 1D). Hence, pri-miRs are often less abundant than TF transcripts. Low abundance (C . 33) also resulted in higher variability between replicate measurements, making detection of small expression changes less reliable. Figure 1. qRT-PCR platform for pri-miR transcript quantification. A, Identification of N- and P-Responsive Typical real-time RT-PCR amplification plots of 175 miR amplicons Arabidopsis pri-miRs showing increase in SYBR Green fluorescence (DRn; log scale) with PCR cycle number. Note the similar slope of most curves as they cross The qRT-PCR platform was used to identify pri-miRs the fluorescence threshold of 0.18 (green line), which reflects similar that are induced or repressed in 9-d-old Arabidopsis amplification efficiency. B, Separation of PCR products on 4% (w/v) seedlings during N or P limitation (Supplemental File high-resolution agarose gels reveals single products of the expected S2). The expected physiological status of the seedlings size (black numbers). A selection of 10 amplicons is shown. Size was confirmed by evaluation of marker gene expression standards in bp are indicated to the left. C, Distribution of efficiencies for 154 primer pairs (gray bars). For 21 primer pairs (white bar) no (Supplemental Fig. S5). NRT2.5 (At1g12940)and primer efficiencies were obtained in the conditions investigated. D, AMT1.5 (At3g24290) were both strongly induced in Expression strength distribution of MIRNA genes (white bars) and of N-limited seedlings, and PHT1.4 (At2g38940) was in- transcription factor genes (black bars). Genes were grouped according duced by P limitation, as found previously (Scheible to their DC values, calculated with UBQ10 as reference. [See online et al., 2004; Morcuende et al., 2007). article for color version of this figure.] Twenty pri-miRs exhibited differential expression in N- or P-limited conditions (Fig. 2), based on an aver- age change in normalized cycle number of at least three (|DDC | $ 3) between nutrient limitation and Plant Physiol. Vol. 150, 2009 1543 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. unknown: pri-miR447c, -778, and -827 all increased (DDC = 4.2–7.6) during P starvation, whereas pri- miR398a strongly decreased (DDC = 26.9; Fig. 2; Supplemental File S2). Also, pri-miRs 169m and 169n displayed induction (DDC = 3.7–4.2) during P limitation, and the same two pri-miR169s plus five additional ones (pri-miR169h through -169l) were decreased (DDC = 23.1 to 24.9) in N-limited seed- lings (Fig. 2). Pri-miR398a and pri-miR447c were not only respon- sive to P limitation but also showed similar responses in N limitation, albeit not as strong, with pri-miR398a being slightly repressed (DDC = 22.45; Supplemental File S2) and pri-miR447c induced (DDC = 3.5; Fig. 2). Furthermore, pri-miRs156e, -156g, and -157d were found to be induced (DDC = 3.1–4.4) in N-limited seedlings. Pri-miR167a, which was reported to be more highly expressed in N-limited root pericycle cells (Gifford et al., 2008), showed no clear response in our experiments with N-limited seedlings. However, pri-miR167d was less expressed (DDC = 22.5) in N-limited seedlings (Supplemental File S2). Nutrient-Responsive Mature miR Species To examine if the mature miRs derived from the nutrient-responsive pri-miRs also showed a nutrient response, we used a qRT-PCR approach similar to the one described by Chen et al. (2005; Supplemental Table S1). In addition to nutrient-replete (FN), N-limited Figure 2. Identification of N/P limitation-responsive pri-miRs in Arabi- (2N), and P-limited (2P) Arabidopsis seedlings, we dopsis. For each pri-miR species, the average DDC value 6 SE from also included carbohydrate-limited (2C) seedlings three biological replicates (with two technical replicates for each) is (Supplemental Fig. S5) in the analysis. Mature depicted. DDC = DC 2 DC , and DC =C 2 C . T TFN T -nutrient T T pri-miR T UBQ10 miR399s were not analyzed, since their high abun- Pri-miRs induced in nutrient limitation thus have a positive DDC value dance during P limitation is already well documented and vice versa. The fold induction can be inferred from the equation (see introduction). DDC (1 + E) , where E is PCR efficiency (Supplemental File S2). Only pri- MiR778 and miR827 were both strongly induced by miRs with average DDC values of .3or ,23 (as indicated by the light P limitation, thus confirming the response of their pri- gray shading) are shown. Results for P and N limitation are shown as miRs. However, they did not respond to either N or C dark gray and white bars, respectively. limitation (Fig. 3) and remained almost undetectable under these conditions, suggesting that both of these full nutrition (FN). The most prominent change was miRs are involved in P-specific regulation events. The the known induction of pri-miR399s during P limita- response of miR398a was also similar to that of the tion, with DDC ranging from 6.5 to 20 (Fig. 2; Bari corresponding pri-miR, being decreased by P and N et al., 2006). Pri-miR399d was undetectable in FN limitation and also by C limitation (Fig. 3). conditions and rose to levels comparable to some of Given the previous report of Gifford et al. (2008) and the most strongly expressed genes in Arabidopsis, the moderate repression of pri-miR167d we observed such as UBQ10 (Czechowski et al., 2005). Very high in N-limited seedlings, we also investigated miR167. expression of pri-miR399s and mature miR399 during The assay for miR167 showed a relatively high ex- P limitation is necessary to fully suppress the activity pression level (40 2 DC = 35), but this might be due to of its target PHO2 transcript (Bari et al., 2006). Mech- low specificity in the assay leading to detection of anistically, this seems to involve translational repres- miR167 derived from several primary transcripts. Our sion, as suggested by previous results (Bari et al., 2006) result obtained with N-limited seedlings (Fig. 3) does and our finding that miR399d-overexpressing seed- not confirm the reported down-regulation of miR167 lings have substantially lower but still detectable in root pericycle cells by organic N (Gifford et al., levels of PHO2 transcript (10%–20% of control plants), 2008). Possible explanations for this discrepancy are while their molecular phenotypes are identical to pho2 the nonspecificity of the assay for miR167 and/or the seedlings (Supplemental Fig. S6). lack of spatial resolution in our analysis, which could qRT-PCR profiling revealed several pri-miR species mask any cell type-specific response of a particular for which nutrient responsiveness was previously MIRNA gene and its derived miR. 1544 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs quencing (SRS) with Illumina-Solexa technology. Three cDNA libraries prepared from nutrient-replete (FN) seedlings, P-limited (2P) seedlings, and 2P seedlings that were resupplied with 3 mM phosphate for 3 h were sequenced (Supplemental Table S2). Sequence reads with 100% identity to Arabidopsis pre-miRs were extracted, and identical reads were totaled (Supplemental Table S2; Supplemental Fig. S7; Supplemental File S3) and normalized for each library. A plot of the distribution of read lengths for pre-miR- matching sequences (Supplemental Fig. S7A) illus- trates that these consist almost exclusively of 20- and 21-mers, with the latter being the most abundant, whereas a plot of all genome-matching sequences reveals a substantial number of 24-mers and other lengths (Supplemental Fig. S7B). Comparison also shows a significantly (approximately 30%) higher number of 21-mers in the group of genome-matching sequences, possibly indicating the presence of un- known miRs. The normalized read numbers for miR399s were Figure 3. Nutrient responsiveness of mature miRs. qRT-PCR results are high during P limitation and very low in nutrient- shown for nutrient-replete (white bars), P-limited (black bars), replete (FN) conditions (Fig. 4A), as expected (Bari N-limited (gray bars), and C-limited (hatched bars) Arabidopsis seedlings. et al., 2006; Pant et al., 2008). There was no appreciable Expression levels are given on a logarithmic scale expressed as 40 2 decrease of miR399s at 3 h after Pi readdition, sup- DC , where DC is the difference in qRT-PCR threshold cycle number of T T porting earlier results that suggested that miR399s are the respective miR and the reference gene UBQ10 (At4g05320); rather stable (Bari et al., 2006). SRS also confirmed high therefore, 40 equals the expression level of UBQ10 (the number 40 expression of miR778 and miR827 during P limitation was chosen because the PCR run stops after 40 cycles). The results are averages 6 SE of three biological replicates. Significance of the changes (Fig. 4A). Very few (one or two) or no reads were found during P, N, or C limitation was checked with Student’s t test. P , obtained for miR398a, -447c, -845, or -169h to -n, 0.05 is indicated by circles, and P , 0.01 is indicated by plus signs. whereas these miRs were detectable by qRT-PCR, suggesting that our qRT-PCR analysis was more sen- sitive than the SRS approach. SRS showed three addi- Nucleolytic cleavage of pri-miR169h to -n (Supple- tional miRs (i.e. miR408, miR829, and miR863) to be mental Fig. S1) by the DICER endoribonuclease results 4- to 10-fold more abundant during P limitation. Anal- in identical miR169 molecules, precluding a specific ysis by qRT-PCR supported the increase in levels of assay. Nonetheless, since all seven pri-miR sequences miR408 but was unable to confirm the higher abun- were less abundant in N limitation (Fig. 2), we were able dance of miR829 or miR863 (Supplemental Fig. S8). to confirm lower miR169h to -n levels in N limitation (Fig. Surprisingly, sequence reads representing star (*) 3). The moderate induction found for two of these pri- strands of some of the nutrient-responsive miRs (i.e. miR169 species during P limitation was not supported miR398a*, miR399a*, miR399c*, miR399d*, miR399f*, by the miR169h to -n assay; on the contrary, mature and miR778*) were found to be present in numbers miR169h to -n showed a significant (approximately similar to, or exceeding (up to 200-fold in the case of 3-fold) decrease in P limitation. A second assay designed miR398a*), those for the corresponding miRs (Fig. 4A; for miR169a to -g also indicated lower abundance during Supplemental File S3), whereas others (e.g. miR399b*, N and P limitation (Fig. 3), although no clear change was miR827*, and miR863*) were absent. The presence of detected for the corresponding pri-miRs. star strands and their specific induction by Pi limita- We were also able to confirm a slight induction tion was confirmed by qRT-PCR (Supplemental Fig. (DDC w 2) of miR447 during P but not N limitation, S8). These results raise the question of the biological whereas the induction of miR156 during N limitation, function of these abundant miR*s. Another interesting as suggested by pri-miR156e and -156g (Fig. 2), was observation from the SRS was that the sequence reads not confirmed (Fig. 3). Again, a nonspecific assay and corresponding to miR778 and miR863 do not perfectly high expression of other pri-miRs from the same agree with the annotated miR sequences determined family (Supplemental File S2) probably account for from flower samples (Fahlgren et al., 2007) but look to this discrepancy. be shifted by one or several nucleotides (Supplemental File S3). Furthermore, the miR778/miR778* duplex P Status-Responsive miRs Detected by Small appears to be unusual in having a 5# overhang that is RNA Sequencing seven nucleotides long (Supplemental File S3). These results from Arabidopsis seedlings could indicate that As an independent approach to verify the P respon- siveness of mature miRs, we used small RNA se- the cleavage site of some pre-miRs by DICER proteins Plant Physiol. Vol. 150, 2009 1545 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Figure 4. P-responsive miRs as detected by small RNA sequencing. A, The number of normalized sequence reads (ppm) is shown for all miRs and their corresponding star (*) strands that show at least a 5-fold change between P limitation (black bars) and P-replete (white bars) conditions and for which at least 10 total reads were scored in at least one condition (Supplemental File S3). Normalized read numbers are also shown for a time point of 3 h after resupply of 3 mM potassium phosphate to previously P-limited seedlings (3hP; gray bars). Missing bars indicate zero reads. B, Putative miR2111a and -b precursor sequences and structures. The sequences of mature miR2111 and the presumed star strands are shown in boldface on the top and bottom strands, respectively. The number of absolute reads for the three conditions tested (see A) is indicated. C, qRT-PCR expression of the 82- and 103-nucleotide-long pri-miR2111a and -b amplicons in various conditions (FN, full nutrients; 2P, P limitation; 3hP, 3 h of Pi readdition; 2N, N limitation; 2C, carbohydrate limitation). Expression levels are plotted on a logarithmic scale as described in the legend to Figure 3. The results are averages 6 SE of three biological replicates with two technical replicates for each. is not totally fixed and may change with developmen- dence on P status (Fig. 4A). It is highly abundant tal stage. during P limitation but almost absent in nutrient- replete seedlings and so closely resembles the behav- ior of known P-responsive miRs (miR399s, miR778, Discovery of a Novel P Status-Responsive miR and miR827) in these conditions. When Pi is resup- The SRS data were analyzed further to look for novel plied to Pi-limited seedlings, the abundance of the P status-dependent miRs. Candidate miRs were pre- novel miR2111 fell by approximately 2-fold within 3 h dicted using a miRDeep algorithm (Friedlander et al., of Pi readdition (Fig. 4A). This response differs from 2008) that was adapted for plant miR precursor se- those of miR399s and miR827, which do not fall ¨ rapidly after Pi readdition, but is similar to that of quences (P. May and M. Friedlander, unpublished miR778, suggesting that the biological half-lives of data). Among the predicted miR candidates was one, named miR2111, that showed a very strong depen- miRs can vary considerably. 1546 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs The DNA sequence of miR2111 is present twice in was previously reported by Giavalisco et al. (2006) and the Arabidopsis genome: upstream of At5g02040, and Buhtz et al. (2008). between At3g09280 and At3g09290. For these loca- Further analysis revealed that in addition to Bna- tions, the algorithms miRDeep and miRCat (http:// miR399 (Fig. 5B; Pant et al., 2008), a Bna-miR399-like srna-tools.cmp.uea.ac.uk) predicted two precursor se- sequence, Bna-miR2111, a Bna-miR2111-like sequence, quences/structures, named pre-miR2111a and -b (Fig. and an Ath-miR827-like sequence are highly abundant 4B). The observation that the read numbers of the two in rapeseed phloem sap during P limitation, while no star strands (miR2111a* and miR2111b*) are similar sequences homologous to the P-responsive Ath- and that they decrease after Pi readdition or are almost miR778 were found. We were also able to determine absent in P-replete conditions reveals that both loci that (1) two sequences with homology to Ath- have P-responsive expression and contribute to a miR399d* and Ath-miR399f*, (2) Bna-miR2111a* and similar extent. The strong P limitation response of Bna-miR2111b*, and (3) five miR2111*-like sequences miR2111 was confirmed by qRT-PCR (Supplemental were clearly present in phloem sap during P limitation Fig. S8). In addition, PCR products designed to en- but absent or much less abundant in phloem sap of compass larger stretches (82 and 103 nucleotides) of nutrient-replete or N-limited rapeseed (Fig. 5B). Fur- pre-miR2111a and -b could be amplified from oligo thermore, we found Bna-miR169m to be present in (dT)-primed cDNA pools, showing that miR2111 is phloem sap of nutrient-replete plants, and the relative derived from poly(A)-tailed primary transcripts, and abundance of this species strongly decreased in these also displayed strong P responsiveness (Fig. 4C). phloem sap of N-limited or P-limited rapeseed plants Furthermore, both the mature miR2111 and its pri- (Fig. 5B). These results pinpoint miR2111s, miR2111*s, miRs displayed specificity for P, as N or C limitation miR399*s, miR827, and miR169 as novel candidate did not affect the expression levels (Fig. 4C; Supple- long-distance signals for reporting P or N status in the mental Fig. S8). We were also able to find potential plant system and suggest an expansion of the miR2111 orthologs of pre-miR2111s in rapeseed (Brassica napus; family in rapeseed. Supplemental Fig. S9). The rapeseed 2111a and 2111b precursors share 85% and 83% identity at the DNA Target Predictions for Nutrient-Regulated miRs level with their Arabidopsis counterparts and are predicted to fold into stable extended hairpin struc- To identify candidate miR target genes, we used tures by the miRCat algorithm. PCR primer pairs for four prediction algorithms: miRU (Zhang, 2005), target Arabidopsis pri-miR2111a and -2111b were added to search in WMD2 (Ossowski et al., 2008), the prediction the qRT-PCR platform, thus bringing the number tool in the UEA plant sRNA toolkit (Moxon et al., of MIRNA genes represented to 177 (Supplemental 2008b), and PITA (Kertesz et al., 2007). In addition, we File S2). mined experimental data from degradome studies (Addo-Quaye et al., 2008; German et al., 2008). The first three algorithms score the complementarity of MiRs with P or N Status-Dependent Abundance in miR and target RNA sequences based on established Rapeseed Phloem Sap rules (e.g. the “seed rule”; Allen et al., 2005), thereby MiR399 was previously found to be highly abun- also exploring the observation that plant miRs seem to dant in phloem sap from rapeseed and pumpkin bind almost perfectly to their cognate mRNAs (Cucurbita maxima) during P limitation and to consti- (Rhoades et al., 2002; Lai, 2004). In contrast, PITA tute a shoot-derived long-distance signal for the reg- assesses site accessibility in miR target recognition. ulation of plant Pi homeostasis (Buhtz et al., 2008; Lin This includes prediction of target RNA secondary et al., 2008; Pant et al., 2008). To test whether additional structure and calculation of the free energy gained miRs show nutrient status-dependent abundance in from the formation of the miR-target duplex and the phloem sap, we collected sap from nutrient-replete, energetic cost of unpairing the target to make it acces- P-limited, and N-limited rapeseed plants and pre- sible to the miR. All candidate miR targets predicted pared small RNA libraries for sequencing. with miRU, WMD2, or the UEA plant sRNA toolkit, Analysis of the resulting small RNA reads (Supple- and up to 20 best targets from the PITA analysis, are mental Table S3) showed that sequences homologous shown in Supplemental File S4, with a selection of the to miRs known to be present in the phloem (e.g. predicted targets shown in Table I. miR156, miR159, and miR167) were present in the In addition to the confirmed miR399 target PHO2,a libraries, whereas sequences homologous to miR171, potential target of miR399b/c predicted by all algo- which is abundant in leaf or stem tissue but undetect- rithms is the receptor kinase gene ACR4, which re- able in phloem (Yoo et al., 2004; Buhtz et al., 2008), stricts formative cell divisions in the Arabidopsis root were completely absent (Fig. 5A). This indicates that (De Smet et al., 2008). Another repeatedly predicted the phloem sap samples were not noticeably contam- miR399b/c target is At4g00170, encoding a vesicle- inated with small RNAs from stem tissue. Small RNA associated membrane protein. A DEAD box helicase species found in the libraries should thus represent gene (At4g09730) appears as a candidate miR399a/d authentic phloem constituents. The high purity of target from analysis with conventional prediction al- rapeseed phloem sap obtained using the same method gorithms, whereas PITA analysis identifies IPS1,a Plant Physiol. Vol. 150, 2009 1547 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Figure 5. miRs with P or N status-dependent abundance in rapeseed phloem sap. A, Abundance of miRs known to be present (miR156, miR159, and miR167) or absent (miR171) in rapeseed phloem sap. B, MiR and miR* sequences with P or N status- dependent abundance. Depicted are normalized read numbers as detected by Solexa sequencing in small RNA libraries prepared from phloem sap samples of P-starved (black bars), nutrient-replete (white bars), and N-starved (gray bars) rapeseed plants. The significance of the changes between P-starved (or N-starved) and nutrient-replete conditions was analyzed with a x test and Benjamini-Hochberg P value correction; P , 0.0003 is marked with circles. Sequences of the species shown are as follows: 169m, 5#-TGAGCCAAAGATGACTTGCCG-3#; 399, 5#-TGCCAAAGGAGATTTGCCCGG-3#; 399-like, 5#-TGC- CAAAGGAGATTTGTCCGG-3#; 399*-like 1, 5#-GGGCGAATACTCTTATGGCAGA-3#; 399*-like 2, 5#-GGGCAAGATCTC- TATTGGCAGA-3#; 2111, 5#-TAATCTGCATCCTGAGGTTTA-3#; 2111-like, 5#-TAATCGGCATCCTGAGGTTTA-3#; 2111a*, 5#-ATCCTCGGGATACAGATTACC-3#; 2111b*, 5#-GTCCTCGGGATGCGGATTACC-3#; 2111*-like 1, 5#-ATCCTCGGGATG- CGGATTACC-3#; 2111*-like 2, 5#-ATCCTCGGGATACGGATTACC-3#; 2111*-like 3, 5#-ATCCTCGGGACACAGATTACC-3#; 2111*-like 4, 5#-GACCTCAGGATGCGGATTACC-3#; 2111*-like 5, 5#-TCCTCGGGATACAGATTACC-3#; 156, 5#-TGACAG- AAGAGAGTGAGCAC-3#; 159, 5#-TTTGGATTGAAGGGAGCTCTA-3#; 167, 5#-TGAAGCTGCCAGCATGATCTA-3#; 171, 5#-TGATTGAGCCGCGCCAATATC-3#. miR399 target-like interactor (Franco-Zorrilla et al., the CSD2 miR398a duplex (Brodersen and Voinnet, 2007). There are also clear target predictions for 2009). miR399*s, with At3g11130, encoding a clathrin heavy chain and thus another protein involved in vesicle transport in the secretory pathway, being a potential DISCUSSION target of miR399d*. One of the best potential targets of miR399f* is the CLAVATA3-related At3g25905/CLE27, qRT-PCR of pri-miRs Is Suitable for Discovery of a gene encoding a small peptide that is highly ex- Stimulus-Dependent miRs pressed in shoot apices (Sharma et al., 2003). Obvious targets of miR827 are the E3 ligase gene The role of miRs during the adaptation of plants to NLA (At1g02860) and its homolog At1g63010 (Table I; abiotic and nutritional stresses is a field that attracts Fahlgren et al., 2007). Also, miR2111 is predicted to increasing interest (Chiou, 2007; Sunkar et al., 2007), target an E3 ligase gene (At3g27150) and a calcineurin- but information on stress-dependent expression of like phosphoesterase gene (At1g07010). A likely target MIRNA genes is limited. To overcome this situation, of miR778 is the histone methyltransferase gene we developed a qRT-PCR platform for quantification SUVH6 (At2g22740), with SUVH5 (At2g35160) being of almost all known Arabidopsis pri-miRs. The results a possible target as well. Interestingly, miR2111a* and obtained with the platform and from targeted assays miR2111b* also appear to target genes required for for mature miRs show that the responses of pri-miRs chromatin remodeling/modification (i.e. At2g23380/ frequently match the responses of the biologically CURLY LEAF and At2g28290/SPLAYED). active mature species, thus validating pri-miR profil- Confirmed targets of miR398a include two Cu/Zn ing as a useful discovery tool. It is also rapid, sensitive, superoxide dismutase genes (CSD1 and CSD2) and and inexpensive and should represent an attractive COX5b.1 (see introduction). The prediction algorithms initial approach for the investigation of MIRNA gene detected CSD1, COX5b.1,and At1g12520/CCS1, a chap- expression. The usefulness of the pri-miR platform is erone that activates CSD, as potential targets of also demonstrated by comparative analysis of the hyl1 miR398a (Table I). CSD2 was not found, most likely mutant and wild-type plants, revealing accumulation due to a bulge and GU wobble in the seed region of of specific pri-miRs in hyl1 and thus an important role 1548 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Table I. Selected targets of P- and N-responsive miRs/miR*s miR/miR* Stimulus Predicted Target Gene(s) Gene Product(s) abcdef 169a +N (?) Several NF-YA Several nuclear factor Y A subunits abcdef 169b/c +N (?) Several NF-YA Several nuclear factor Y A subunits abce At5g42120 Lectin protein kinase family protein abcdef 169d to -g +N (?) Several NF-YA Several nuclear factor Y A subunits ae be At1g70700 and At1g48500 Jasmonate-ZIM domain proteins 4 and 9 abcdef 169h to -n +N, +P Several NF-YA Several nuclear factor Y A subunits ab At5g42120 Lectin protein kinase family protein bdef 398a +P, +N, +C At1g08830 Cu/Zn superoxide dismutase CSD1 be At1g12520 Superoxide dismutase chaperone CCS1 cd At3g15640 Cytochrome c oxidase subunit COX5b.1 abcdef 399a 2P At2g33770 E2 conjugase PHO2 ac At4g09730 ATP-dependent RNA helicase At4g27850 Pro-rich protein abcdef 399b/c 2P At2g33770 E2 conjugase PHO2 abce At3g59420 ACR4 kinase abc At3g54700 Phosphate transporter bce At4g00170 VAMP family protein At4g27850 Pro-rich protein At3g09922 IPS1 abcdef 399d 2P At2g33770 E2 conjugase PHO2 ac At4g09730 ATP-dependent RNA helicase abcdef 399f 2P At2g33770 E2 conjugase PHO2 At3g09922 IPS1 abcde 778 2P At2g22740 Histone methyltransferase SUVH6 At2g35160 Histone methyltransferase SUVH5 At5g51980 WD40-repeat protein abcde 827 2P At1g02860 SPX E3 ligase NLA abce At1g63010 SPX E3 ligase abce 2111 2P At3g27150 F box protein abcde At2g23370 Unknown protein abcde At1g07010 Calcineurin-like phosphoesterase ae 398a* +P, +N, +C At4g00950 MEE47 (maternal effect embryo arrest 47) abce At5g06120 Ran-binding protein abe 399c* 2P At5g64470 DUF231 protein ae 399d* 2P At3g11130 Clathrin heavy chain abe 399f* 2P At3g25905 CLAVATA3/ESR-RELATED27 abe 778* 2P Ag1g69610 Ribosomal protein 2111a* 2P At2g28290 SWI2/SNF2-like protein SPLAYED abce 2111b* 2P At2g23380 SET domain protein CURLY LEAF abe At1g60380 NAC domain transcription factor 24 a b c d Predicted by miRU. Predicted by WMD2. Predicted by UEA toolbox. Predicted by e f degradome data. Predicted by PITA. Experimentally confirmed. for the HYL1 double-stranded RNA-binding protein platform as new MIRNA genes are discovered. In this in selective pri-miR maturation (Szarzynska et al., regard, we invite readers to send us relevant sequence 2009). We found only a few examples where the P or information. N response of pri-miRs was not confirmed for the corresponding mature miRs. Although this could have Evidence for miR2111 Being a True miR been due to technical reasons, regulation or attenua- tion at the level of miR maturation could also account The novel Pi-responsive miR2111 was revealed by a for these differences. version of miRDeep optimized for analysis of plant It is likely that the number of recognized Arabi- sequences (P. May, unpublished data), but this alone is dopsis MIRNA loci will further increase (Lindow and not conclusive proof that it is a true miR. A number of Krogh, 2005; Lindow et al., 2007), especially with revised criteria exist for annotation of plant miRs deeper analysis of SRS data, possibly reaching num- (Meyers et al., 2008). The primary and only criterion bers similar to those currently known for rice (Oryza that is necessary and sufficient for annotation as a miR sativa; 269), zebra fish (337), mouse (472), and human is conclusive evidence of precise biogenesis from a (678; http://microrna.sanger.ac.uk). The easily scal- qualifying stem loop, and this criterion is met by able qRT-PCR approach allows regular updates of the miR2111. First, miR2111 was shown to be derived from Plant Physiol. Vol. 150, 2009 1549 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. longer poly(A)-tailed precursor transcripts, as it was two or more complementary algorithms, giving possible to obtain precursor PCR amplicons from greater confidence in the predictions. Target analysis cDNA pools generated by RT with an oligo(dT) primer with PITA also indicated that inclusion of thermody- (Fig. 4C). Second, the two miR2111 precursors fold into namics of RNA-RNA interactions can change the characteristic hairpin structures with stabilities typical results greatly (Hofacker, 2007). Only PITA correctly for known miR precursors. Third, there is precise identified IPS1, a noncoding RNA that qualifies as excision of 21-nucleotide miR2111/miR2111* duplexes a bona fide miR399 target-like interactor (Franco- from the two stem-loop precursors (Supplemental File Zorrilla et al., 2007), emphasizing the complementarity S3). Fourth, miR2111 and miR2111a/b* are derived and predictive power of PITA. The number of poten- from opposite stem arms and form duplexes with tial targets identified with PITA (Supplemental File S4) typical 3# overhangs of two-nucleotide length. Fifth, further suggests that an individual plant miR could base pairing between miR2111 and the star strands is also have a larger range of action, for example, by extensive (Fig. 4B). Finally, the observed small RNA inducing widespread changes in protein synthesis, as abundance corresponds entirely to the duplexes. An- recently reported for several human miRs (Selbach cillary criteria for plant miR annotation include (1) the et al., 2008). existence of target genes, (2) conservation between Three P starvation-inducible miRs (miR399, species, and (3) biogenesis that is dependent on miR2111, and miR827) have confirmed or likely target DICER-like (DCL) proteins. There are obvious poten- genes involved in protein degradation via the 26S tial target genes for miR2111 (Table I), although the proteasome. MiR827 targets the E3 ligase gene NLA biological significance of these being targeted by (At1g02860). NLA transcript also drops 2- to 3-fold miR2111 is as yet unknown. We also detected pre- during P limitation when miR827 is highly expressed miR2111 homologs and found mature miR2111 in (Morcuende et al., 2007). NLA is crucial for anthocy- rapeseed (Fig. 5; Supplemental Fig. S9). Although the anin synthesis, and the nla mutant displays severely dependence of miR2111 biogenesis on DCL1 or DCL4 reduced anthocyanin content and early leaf senescence still needs to be tested, there is already overwhelming during N limitation (Peng et al., 2007a, 2007b, 2008). evidence that miR2111 is a true miR. However, during P limitation or during simultaneous P and N limitation, the mutant displays wild-type-like anthocyanin levels and no leaf senescence (Peng et al., More Widespread Regulation by Small RNAs during 2008), showing that the signal derived from P limita- P Limitation tion is sufficient to induce anthocyanin production in So far, miR399s have been the only small RNA species known to strongly increase during P limitation (see introduction). Five MIRNA399 genes that encode the slightly different mature miR399s exist in Arabi- dopsis. Still the only confirmed target of miR399s is PHO2, while IPS1 is a miR399 interactor (Franco- Zorrilla et al., 2007). This situation and expansion of the MIRNA399 gene family in other plant species such as Medicago truncatula (F. Krajinski, personal commu- nication) and rice (Lindow et al., 2007), however, indicate a larger miR399 regulatory network. The new finding that at least four additional miRs and several miR*s with strong P status-dependent expres- sion exist now suggests that regulation/signaling by small RNAs during P limitation is even more wide- spread. Regulatory activity of miR* species and their presence in argonaute complexes have been demon- strated previously (Mi et al., 2008; Okamura et al., 2008). The fact that the SRS read numbers for some miR* sequences clearly exceed the numbers for the corresponding miRs indicates that they are not merely by-products that are slowly degraded. Target Predictions and Biological Processes Potentially Figure 6. Hypothetical models for miR827 and miR169 functions. A, Affected by P-Regulated Small RNAs Model showing cross talk between P limitation and N limitation signaling pathways that affect anthocyanin synthesis. B, Model inte- We used several prediction algorithms and mined grating published features of the systemic regulation of nodulation by degradome data (Table I) to determine likely targets of CLE (Okamoto et al., 2009), SUNN/HAR1 (Krusell et al., 2002), and P-regulated small RNAs. This combinatorial approach HAP2 and miR169 (Combier et al., 2006), with novel results concern- revealed several miR targets that were predicted by ing miR169 described in this work. 1550 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs the nla mutant. The link between P limitation and NLA expressing miR169 show enhanced leaf water loss and provided by miR827 (Fig. 6A) suggests that NLA are more sensitive to drought stress, whereas NFYA5 activity is actively down-regulated during P limita- overexpressers show the opposite phenotypes (Li tion. This could indicate that plants select one or the et al., 2008). In addition to the effects of the nitrate other input signal depending on nutrient conditions transporter CHL1 (Guo et al., 2003) or nitrate reductase- and, therefore, the existence of hierarchies in the mediated nitric oxide generation (Desikan et al., 2002) interplay of macronutrient regulatory networks. on stomatal opening, low expression of miR169 during MiR2111 is predicted to target the E3 ligase gene N limitation could thus contribute to drought tolerance At3g27150 (Table I). At3g27150 displays strictly root- of N-limited plants (Lodeiro et al., 2000; Castaings et al., specific expression in large-scale transcriptome data 2008). sets like AtGenExpress (Schmid et al., 2005), suggest- In legume species, nodule development is depen- ing that it functions in the root. This is interesting in dent on the presence of previously established nodules the context of the high abundance of miR2111 in and N/nitrate availability, creating a root-to-shoot phloem sap during Pi limitation (Fig. 5), suggesting signal that activates the CLAVATA1-like receptor ki- another systemic regulatory circuitry, analogous to the nase SUNN in M. truncatula or HAR1 in Lotus japoni- miR399-PHO2 paradigm (Pant et al., 2008). cus. A recent report suggests that a nitrate-induced Regulation of chromatin status appears to be an- CLAVATA3/ESR-related (CLE) peptide is this root-to- other biological process influenced by P limitation- shoot signal (Fig. 6B; Okamoto et al., 2009). HAR1 induced miRs, as suggested by the best predicted exerts negative shoot control of root nodulation target genes of miR778, miR2111b*, and miR2111a*, (Krusell et al., 2002; Nishimura et al., 2002) through a namely SUVH6 (At2g22740), encoding a SET domain- shoot-to-root signal that might include auxin transport containing histone methyltransferase; At2g23380/ (van Noorden et al., 2006). It is also known that the CURLY LEAF, a SET domain gene required for his- miR169 target gene HAP2-1 in M. truncatula is a key tone methylation and genetic imprinting (Schubert regulator for the differentiation of nodule primordia et al., 2006); and At2g28290/SPLAYED, encoding a (Combier et al., 2006; Fig. 6B). MiR169 overexpression chromatin-remodeling complex subunit required for or knockdown of HAP2-1 leads to a developmental maintenance and identity of the shoot apical meristem block of nodule formation (Combier et al., 2006). (Kwon et al., 2005). Repression of miR169s by N limitation, as detected MiR398a is strongly reduced in P, N, and C limita- in our experiments, points toward a potential mecha- tion (Fig. 3), indicating a more general response to nistic link between low N status and nodule develop- nutrient stress. Repression of miR398a by C limitation ment in legumes. High abundance of miR169 in phloem also correlates with its induction by Suc (Dugas and sap during N-replete growth and the sharp decrease Bartel, 2008). Targets of miR398a include CSD1 and during N and P limitation (Fig. 5) also flags miR169 as a CSD2 (see introduction). CSD is required for detoxifi- potential long-distance signal (Fig. 6B) that is able to cation of reactive oxygen species that increase during report shoot N and P status to the roots, similar to the nutrient limitations and other environmental stresses role of miR399 (Pant et al., 2008). It will be interesting to (e.g. heat or drought; Apel and Hirt, 2004; Shin et al., test whether miR169 abundance is increased by nitrate/ 2005; Sunkar et al., 2006). Therefore, down-regulation N in legumes and whether miR169 expression and/or of miR398a leading to higher CSD activity would be an phloem abundance is dependent on SUNN/HAR1 to appropriate response to nutrient stress. establish if this is a novel shoot-to-root signal for the control of nodule differentiation. Potential Biological Impact of miR169 Regulation by CONCLUSION N Availability MiRs are emerging as increasingly interesting (sys- The targets of miR169s are several HAP2 trans- temic) regulators during mineral nutrient stress in cription factors (i.e. nuclear factor YA subunits plants. The discovery of new nutrient-dependent miRs [NF-YA]; Table I; Combier et al., 2006; Fahlgren et al., opens up the possibility of testing their roles and those 2007; Li et al., 2008). Transcripts of several of these of their predicted targets during adaptation of plants genes, including At3g05690/NF-YA2, At1g54160/ to nutrient deficiency. The qRT-PCR platform de- NF-YA5, At3g14020/NF-YA6, At1g72830/NF-YA8,and scribed here serves as a useful initial approach to test At5g06510/NF-YA10, increase during N and P limita- the response of annotated miRs in a given biological tion (Supplemental Fig. S10; Scheible et al., 2004; qRT- scenario, providing opportunities to discover new PCR confirmation not shown), thereby showing the signaling and regulatory networks. opposite response compared with miR169. In Arabidopsis, miR169 was reported to influence drought resistance via inhibition of the A5 subunit of MATERIALS AND METHODS NF-Y, a ubiquitous transcription factor that is highly Plant Materials expressed in guard cells and crucial for the expression of a number of drought stress-responsive genes (Li Nine-day-old nutrient-replete and N-, P-, or C-limited wild-type Arabi- et al., 2008). Nfya5 knockout mutants and plants over- dopsis (Arabidopsis thaliana Col-0) seedlings were grown in sterile liquid Plant Physiol. Vol. 150, 2009 1551 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. cultures as described previously (Scheible et al., 2004; Morcuende et al., 2007; 300 mL of EBR buffer (50 mM Mg acetate, 0.5 M ammonium acetate, 1 mM Osuna et al., 2007). The physiological status of the plant materials was EDTA, and 0.1% SDS) for 10 to 16 h at 20Cto25C (300 rpm). After phenol/ confirmed by expression analysis of marker genes (Supplemental Fig. S5) chloroform and chloroform extraction, the aqueous phase was mixed with prior to qRT-PCR analysis. Rapeseed (Brassica napus ‘Drakkar’) was germi- 1 mL of glycogen and 900 mL of 96% (v/v) ethanol, then cooled to 220C for nated and grown hydroponically (Buhtz et al., 2008) in a full-nutrient solution 2 h and centrifuged (25 min, 16,000g,4C). The RNA pellet was washed twice containing 4 mM N and 0.5 mM P. After 47 d, plants were divided into three with 75% (v/v) ethanol and dissolved in 6 mL of water. sets, and one set was supplied with lower N (2 mM KNO ) nutrient solution, 5# and 3# RNA adaptor ligations with RNA primers, RT, and PCR were one with low P (0.1 M KPi) nutrient solution, and one with full-nutrient performed according to Lu et al. (2007), except for a 3# RNA adaptor 3# end solution. From day 61 onward (i.e. approximately 1 week before flowering modification consisting of a C3 hydrocarbon spacer (Biomers.net). The PCR started), low-N, low-P, and FN plants were supplied with nutrient solutions (25 mL) was terminated with 75 mL of stop buffer (10 mM Tris-HCl, pH 8.0, containing no N, no P, or full-nutrient solution, respectively. N- or P-limited 1mM EDTA, and 0.4 M ammonium acetate). After phenol (pH = 8.0)/ plants developed clear signs of N or P starvation (e.g. reduced leaf biomass, chloroform extraction, 1 mL of glycogen (Roche; 20 mg mL ) and 300 mLof earlier flowering, reduced chlorophyll in 2N leaves; data not shown). Phloem ethanol were added to precipitate cDNA. The cDNA was denatured in loading sap was sampled between days 72 and 82 as described previously and was not buffer II (Ambion) and separated on an 8% polyacrylamide/7 M urea gel. The significantly contaminated by cell sap from other tissues (Giavalisco et al., cDNA band was eluted and precipitated as above. The pellet was washed 2006; Buhtz et al., 2008). with 70% (v/v) ethanol, air dried, and dissolved in 14 mL of water. The cDNA concentration was measured using a NanoDrop ND-1000 and checked by 15% polyacrylamide/7 M urea gel electrophoresis with oligonucleotides of known Primer Design and qRT-PCR Analysis concentration. Quality control was performed by TOPO cloning and Sanger sequencing of several plasmid clones (Lu et al., 2007). Illumina-Solexa A first set of pri-miR primers (pri-miR156 through -404) was designed by sequencing was performed at GATC Biotech for Arabidopsis and at FASTERIS Eurogentec. Primers for pri-miR405 through -870, and primers that replaced for rapeseed libraries. malfunctioning primers from the first set were designed using Primer Express 2.0.0 (Applied Biosystems) and Oligo 6.71 (Molecular Biology Insights). To Analysis of Deep Sequencing Results ensure maximum specificity and efficiency during PCR amplification of pri- miR cDNA under a standard set of reaction conditions (Fig. 1A), a stringent set Sequencing reads of lengths between 15 and 32 nucleotides were used after of criteria was used for primer design. This included predicted melting trimming sequence adapters and low-complexity regions [e.g. poly(A)] and temperatures of 61C 6 2C, limited self-complementarity, and PCR amplicon after removing reads of low quality (containing n runs, where n . 12). The lengths of 50 to 150 bp. Secondary hits were minimized by aligning primer read sets from the different conditions were subsequently mapped onto the candidates to all known Arabidopsis transcript sequences via BLAST searches Arabidopsis genome (TAIR8 assembly) using RazerS software. RazerS is an and eliminating primer pairs with more than the specific hit. Stem-loop efficient and generic read-mapping tool allowing the user to align reads of sequences for which no satisfactory primers could be found were elongated by arbitrary length using either the Hamming distance or the edit distance. 100 bp of flanking genomic sequence on each side before primer design was RazerS is part of the generic sequence analysis library Seqan (Doring et al., reinitiated. Annealing sites of the primers on the pri-miR sequence are 2008). Only perfect matches to the genome (i.e. full-length alignments with highlighted in Supplemental File S1. Sequences of the qRT-PCR primers are 100% identity) were retained. To investigate the read distributions for avail- given in Supplemental File S2. Cartridge-purified primers were purchased able TAIR8 annotations of genes and transposable elements (available from from Eurogentec, mixed with the corresponding forward or reverse primer ftp://ftp.arabidopsis.org/home/tair/Genes/TAIR8_genome_release) and to upon arrival to a final concentration of 50 mM each, arrayed on 96-deep-well find statistically significant changes in read distributions, we used the x test plates, and frozen at 280C for long-term storage. Working stocks (0.5 mM)of together with Benjamini-Hochberg P value correction. The x test is known to each primer pair were prepared from the storage stocks in two serial 10-fold have a good predictive power and robustness for gene expression analysis dilution steps and kept at 220C for short-term storage and used within 2 (Man et al., 2000). Normalization of small RNA data was performed by weeks. dividing the read number of each individual small RNA sequence by the RNA isolation, cDNA synthesis, and qRT-PCR analysis were carried out as number of redundant reads (15–32 nucleotides) in each library (Supplemental described previously by Czechowski et al. (2004, 2005) and Udvardi et al. Tables S2 and S3). (2008). Mature miR expression was analyzed using the method of Chen et al. (2005), as described (Pant et al., 2008). MiR-specific RT stem-loop primers are given in Supplemental Table S1. Primer sequences for marker genes are given MiR Prediction in the legend to Supplemental Figure S5. Potential miRs together with their precursor sequences were predicted using the miRDeep software tool (Friedla¨nder et al., 2008). The miRDeep Isolation of Small RNAs, Library Preparation, and algorithm was adjusted to plant precursor structures to take account of the Deep Sequencing following features of some plant miRs: (1) longer pre-miRs; (2) pre-miRs that contain more than one miR sequence; (3) the more diverse read distribution of Total RNA was isolated with Trizol reagent (Invitrogen) supplemented sequenced small RNAs on plant pre-miR sequences; and (4) nonhairpin pre- with 0.5% (w/v) N-lauroylsarcosine sodium salt, 3 mM b-mercaptoethanol, miR structures (P. May, unpublished data). and 5 mM EDTA. After phase separation, one phenol/chloroform and two chloroform extractions were performed. The aqueous phase (500 mL) was Target Gene Predictions mixed with 3 mL of glycogen (Roche; 20 mg mL ) before RNA was precip- itated with 625 mL of ethanol and 250 mLof0.8 M sodium citrate/1.2 M sodium Candidate miR target genes were determined using publicly available chloride. Samples were incubated for 30 min at room temperature and then prediction algorithms, including miRU (Zhang, 2005), the target search in centrifuged (25 min, 16,000g,4C). The precipitate was washed with 80% (v/v) WMD2 (Ossowski et al., 2008), the prediction tool in the UEA plant sRNA ethanol, air dried, and dissolved in 2 mM Tris-HCl (pH 7.5). Efficiency of small toolkit (Moxon et al., 2008b), and PITA (Kertesz et al., 2007). The programs RNA extraction and total RNA quality was checked by northern-blot hybrid- were used with their default settings. ization with a P-labeled oligonucleotide complementary to miR399 (Bari et al., 2006). RNA concentration was measured with a NanoDrop ND-1000 MiRBase accession numbers for all annotated Arabidopsis miRs (NanoDrop Technologies), and integrity was measured with an Agilent-2100 are available at http://microrna.sanger.ac.uk/cgi-bin/sequences/mirna_ Bioanalyzer (Agilent Technologies; RNA 6000 NanoChips). summary.pl?org=ath. GenBank accession numbers for the novel miR2111 Total RNA from three independent biological replicates (3 3 20 mg) was sequences described in this work are FN391952 (Ath-miR2111b), FN391950 mixed with 23 loading buffer II (Ambion), denatured for 2 min at 90C, and (Ath-miR2111a), FN391951 (Bna-miR2111b), and FN391953 (Bna-miR2111a). separated on a 15% polyacrylamide/7 M urea/13 TBE gel at 300 V. Synthetic, phosphorylated 18-mer and 24-mer RNA markers (Biomers.net) and a 10-bp DNA ladder (Invitrogen) were used to localize small RNAs (18–30 nucleo- Supplemental Data tides) as well as ligation and PCR products on gels stained with SYBR Gold (Invitrogen). RNA and PCR products were eluted from polyacrylamide gels in The following materials are available in the online version of this article. 1552 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Supplemental Figure S1. Genome arrangement and sequence similarity of organidentitybyamicroRNA and its APETALA2-liketargetgenes. miR169i to -n precursors. Plant Cell 15: 2730–2741 Aung K, Lin SI, Wu CC, Huang YT, Su CL, Chiou TJ (2006) Pho2, a Supplemental Figure S2. Strong induction of mir395 primary transcripts phosphate overaccumulator, is caused by a nonsense mutation in a during sulfur limitation. microRNA399 target gene. Plant Physiol 141: 1000–1011 Axtell MJ, Bartel DP (2005) Antiquity of microRNAs and their targets in Supplemental Figure S3. Melting curves of pri-miR amplicons land plants. Plant Cell 17: 1658–1673 Supplemental Figure S4. Sequencing results of four pri-miR amplicons. Bari R, Pant BD, Stitt M, Scheible WR (2006) PHO2, microRNA399, and PHR1 define a phosphate-signaling pathway in plants. Plant Physiol Supplemental Figure S5. Marker gene expression in nutrient-limited 141: 988–999 Arabidopsis seedlings. Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and Supplemental Figure S6. Strong overexpression of miR399d mimics function. Cell 116: 281–297 molecular phenotypes of pho2 mutants. Bracht J, Hunter S, Eachus R, Weeks P, Pasquinelli AE (2004) Trans- splicing and polyadenylation of let-7 microRNA primary transcripts. Supplemental Figure S7. Number and length distribution of small RNA RNA 10: 1586–1594 sequences. Brodersen P, Sakvarelidze-Achard L, Bruun-Rasmussen M, Dunoyer P, Supplemental Figure S8. qRT-PCR verification of P limitation-induced Yamamoto YY, Sieburth L, Voinnet O (2008) Widespread translational small RNA species. inhibition by plant miRNAs and siRNAs. Science 320: 1185–1190 Brodersen P, Voinnet O (2009) Revisiting the principles of microRNA Supplemental Figure S9. miR2111 precursors from rapeseed. target recognition and mode of action. Nat Rev Mol Cell Biol 10: 141–148 Supplemental Figure S10. Nutrient-dependent expression of HAP2 genes Buhtz A, Springer F, Chappell L, Baulcombe DC, Kehr J (2008) Identifi- in Arabidopsis. cation and characterization of small RNAs from the phloem of Brassica napus. Plant J 53: 739–749 Supplemental Table S1. Primers used for RT of mature miR and qPCR Cai X, Hagedorn CH, Cullen BR (2004) Human microRNAs are processed quantification. from capped, polyadenylated transcripts that can also function as Supplemental Table S2. Read numbers from small RNA sequencing of mRNAs. RNA 10: 1957–1966 Arabidopsis libraries. Castaings L, Camargo A, Pocholle D, Gaudon V, Texier Y, Boutet-Mercey S, Taconnat L, Renou JP, Daniel-Vedele F, Fernandez E, et al (2008) The Supplemental Table S3. Read numbers from small RNA sequencing of nodule inception-like protein 7 modulates nitrate sensing and metab- rapeseed phloem sap. olism in Arabidopsis. Plant J 57: 426–435 Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT, Barbisin Supplemental File S1. Arabidopsis miR precursor sequences and primer M, Xu NL, Mahuvakar VR, Andersen MR, et al (2005) Real-time annealing sites. quantification of microRNAs by stem-loop RT-PCR. Nucleic Acids Res Supplemental File S2. Quantitative real-time PCR results for all investi- 33: e179 gated pri-miR species. Chiou TJ (2007) The role of microRNAs in sensing nutrient stress. Plant Cell Environ 30: 323–332 Supplemental File S3. Structures and small RNA reads of P-responsive Chiou TJ, Aung K, Lin SI, Wu CC, Chiang SF, Su CL (2006) Regulation of miRs. phosphate homeostasis by microRNA in Arabidopsis. Plant Cell 18: Supplemental File S4. Target gene predictions. 412–421 Chuck G, Candela H, Hake S (2009) Big impacts by small RNAs in plant development. Curr Opin Plant Biol 12: 81–86 Note Added in Proof Combier JP, Frugier F, de Billy F, Boualem A, El-Yahyaoui F, Moreau S, Vernie´ T, Ott T, Gamas P, Crespi M, et al (2006) MtHAP2-1 is a key The novel miR described in this work was independently reported by transcriptional regulator of symbiotic nodule development regulated by Fahlgren et al. (Fahlgren N, Sullivan CM, Kasschau KD, Chapman EJ, microRNA169 in Medicago truncatula. Genes Dev 20: 3084–3088 Cumbie JS, Montgomery TA, Gilbert SD, Dasenko M, Backman TW, Givan Czechowski T, Bari RP, Stitt M, Scheible WR, Udvardi MK (2004) Real- SA, et al [2009] Computational and analytical framework for small RNA time RT-PCR profiling of over 1400 Arabidopsis transcription factors: profiling by high-throughput sequencing. RNA 15: 992–1002). To unify the unprecedented sensitivity reveals novel root- and shoot-specific genes. naming, this miR is referred to as miR2111 in the final published version. Plant J 38: 366–379 Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible WR (2005) Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol 139: 5–17 Desikan R, Griffiths R, Hancock J, Neill S (2002) A new role for an old enzyme: nitrate reductase-mediated nitric oxide generation is required ACKNOWLEDGMENTS for abscisic acid-induced stomatal closure in Arabidopsis thaliana. Proc We thank Dr. John Lunn (Max-Planck Institute of Molecular Plant Phys- Natl Acad Sci USA 99: 16314–16318 iology) and Dr. Peter Do¨rner (University of Edinburgh) for proofreading and De Smet I, Vassileva V, De Rybel B, Levesque MP, Grunewald W, Van suggestions on the manuscript. Damme D, van Noorden G, Naudts M, Van Isterdael G, De Clercq R, et al (2008) Receptor-like kinase ACR4 restricts formative cell divisions Received March 27, 2009; accepted May 18, 2009; published May 22, 2009. in the Arabidopsis root. Science 322: 594–597 Doring A, Weese D, Rausch T, Reinert K (2008) SeqAn an efficient, generic C++ library for sequence analysis. BMC Bioinformatics 9: 11 Dugas DV, Bartel B (2004) MicroRNA regulation of gene expression in LITERATURE CITED plants. Curr Opin Plant Biol 7: 512–520 Dugas DV, Bartel B (2008) Sucrose induction of Arabidopsis miR398 Addo-Quaye C, Eshoo TW, Bartel DP, Axtell MJ (2008) Endogenous represses two Cu/Zn superoxide dismutases. Plant Mol Biol 67: 403–417 siRNA and miRNA targets identified by sequencing of the Arabidopsis Emery JF, Floyd SK, Alvarez J, Eshed Y, Hawker NP, Izhaki A, Baum SF, degradome. Curr Biol 18: 758–762 Bowman JL (2003) Radial patterning of Arabidopsis shoots by class III Allen E, Xie Z, Gustafson AM, Carrington JC (2005) MicroRNA-directed HD-ZIP and KANADI genes. Curr Biol 13: 1768–1774 phasing during trans-acting siRNA biogenesis in plants. Cell 121: Fahlgren N, Howell MD, Kasschau KD, Chapman EJ, Sullivan CM, 207–221 Cumbie JS,Givan SA,Law TF,Grant SR,Dangl JL,etal (2007) High- Apel K, Hirt H (2004) Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu Rev Plant Biol 55: 373–399 throughput sequencing of Arabidopsis microRNAs: evidence for fre- Aukerman MJ, Sakai H (2003) Regulation of flowering time and floral quent birth and death of MIRNA genes. PLoS ONE 2: e219 Plant Physiol. Vol. 150, 2009 1553 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Franco-Zorrilla JM, Valli A, Todesco M, Mateos I, Puga MI, Rubio- Man MZ, Wang X, Wang Y (2000) POWER_SAGE: comparing statistical Somoza I, Leyva A, Weigel D, Garcı´a JA, Paz-Ares J (2007) Target tests for SAGE experiments. Bioinformatics 16: 953–959 mimicry provides a new mechanism for regulation of microRNA activ- Meyers BC, Axtell MJ, Bartel B, Bartel DP, Baulcombe D, Bowman JL, Cao ity. Nat Genet 39: 1033–1037 X, Carrington JC,ChenX,Green PJ,etal (2008) Criteria for annotation Friedla¨nder MR, Chen W, Adamidi C, Maaskola J, Einspanier R, Knespel of plant microRNAs. Plant Cell 20: 3186–3190 Mi S, CaiT, Hu Y,ChenY,HodgesE, NiF,Wu L, LiS,Zhou H, LongC, S, Rajewsky N (2008) Discovering microRNAs from deep sequencing data using miRDeep. Nat Biotechnol 26: 407–415 et al (2008) Sorting of small RNAs into Arabidopsis argonaute com- FujiiH,Chiou TJ,Lin SI,AungK, Zhu JK (2005) A miRNA involved in plexes is directed by the 5# terminal nucleotide. Cell 133: 116–127 phosphate-starvation response in Arabidopsis. Curr Biol 15: 2038–2042 Morcuende R,BariR,Gibon Y, ZhengW,Pant BD, Bla¨sing O, Usadel B, German MA, Pillay M, Jeong DH, Hetawal A, Luo S, Janardhanan P, Czechowski T, UdvardiMK, StittM,etal (2007) Genome-wide Kannan V, Rymarquis LA, Nobuta K, German R, et al (2008) Global reprogramming of metabolism and regulatory networks of Arabidopsis identification of microRNA-target RNA pairs by parallel analysis of in response to phosphorus. Plant Cell Environ 30: 85–112 Moxon S, Jing R, Szittya G, Schwach F, Rusholme P, Pilcher RL, Moulton RNA ends. Nat Biotechnol 26: 941–946 Giavalisco P, Kapitza K, Kolasa A, Buhtz A, Kehr J (2006) Towards the V, Dalmay T (2008a) Deep sequencing of tomato short RNAs identifies proteome of Brassica napus phloem sap. Proteomics 6: 896–909 microRNAs targeting genes involved in fruit ripening. Genome Res 18: Gifford ML, Dean A, Gutierrez RA, Coruzzi GM, Birnbaum KD (2008) 1602–1609 Cell-specific nitrogen responses mediate developmental plasticity. Proc Moxon S, Schwach F, Dalmay T, Maclean D, Studholme DJ, Moulton V Natl Acad Sci USA 105: 803–808 (2008b) A toolkit for analysing large-scale plant small RNA datasets. Guo FQ, Young J, Crawford NM (2003) The nitrate transporter AtNRT1.1 Bioinformatics 24: 2252–2253 (CHL1) functions in stomatal opening and contributes to drought Nishimura R, Hayashi M, Wu GJ, Kouchi H, Imaizumi-Anraku H, Murakami Y, Kawasaki S, Akao S, Ohmori M, Nagasawa M, et al susceptibility in Arabidopsis.Plant Cell 15: 107–117 Hofacker IL (2007) How microRNAs choose their targets. Nat Genet 39: (2002) HAR1 mediates systemic regulation of symbiotic organ develop- 1191–1192 ment. Nature 420: 426–429 Jones-Rhoades MW, Bartel DP (2004) Computational identification of Okamoto S, Ohnishi E, Sato S, Takahashi H, Nakazono M, Tabata S, plant microRNAs and their targets, including a stress-induced miRNA. Kawaguchi M (2009) Nod factor/nitrate-induced CLE genes that drive Mol Cell 14: 787–799 HAR1-mediated systemic regulation of nodulation. Plant Cell Physiol Kawashima CG, Yoshimoto N, Maruyama-Nakashita A, Tsuchiya YN, 50: 67–77 Saito K, Takahashi H, Dalmay T (2008) Sulphur starvation induces the Okamura K, Phillips MD, Tyler DM, Duan H, Chou YT, Lai EC (2008) The expression of microRNA-395 and one of its target genes but in different regulatory activity of microRNA* species has substantial influence on cell types. Plant J 57: 313–321 microRNA and 3# UTR evolution. Nat Struct Mol Biol 15: 354–363 Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E (2007) The role of site Ossowski S, Schwab R, Weigel D (2008) Gene silencing in plants using accessibility in microRNA target recognition. Nat Genet 39: 1278–1284 artificial microRNAs and other small RNAs. Plant J 53: 674–690 Kidner CA, Martienssen RA (2005) The developmental role of microRNA Osuna D, Usadel B, Morcuende R, Gibon Y, Bla¨sing OE, Ho¨hne M, in plants. Curr Opin Plant Biol 8: 38–44 Gu¨ nter M, Kamlage B, Trethewey R, Scheible WR, et al (2007) Tem- Krusell L, Madsen LH, Sato S, Aubert G, Genua A, Szczyglowski K, Duc poral responses of transcripts, enzyme activities and metabolites after G, Kaneko T, Tabata S, de Bruijn F, et al (2002) Shoot control of root adding sucrose to carbon-deprived Arabidopsis seedlings. Plant J 49: development and nodulation is mediated by a receptor-like kinase. 463–491 Nature 420: 422–426 Palatnik JF, Allen E, Wu X, Schommer C, Schwab R, Carrington JC, Kurihara Y, Watanabe Y (2004) Arabidopsis micro-RNA biogenesis Weigel D (2003) Control of leaf morphogenesis by microRNAs. Nature through Dicer-like 1 protein functions. Proc Natl Acad Sci USA 101: 425: 257–263 12753–12758 Pant BD, Buhtz A, Kehr J, Scheible WR (2008) MicroRNA399 is a long- Kutter C, Scho¨b H, Stadler M, Meins F Jr, Si-Ammour A (2007) Micro- distance signal for the regulation of plant phosphate homeostasis. Plant RNA-mediated regulation of stomatal development in Arabidopsis.Plant J 53: 731–738 Cell 19: 2417–2429 Peng M, Bi YM, Zhu T, Rothstein SJ (2007a) Genome-wide analysis of Kwon CS,Chen C,WagnerD (2005) WUSCHEL is a primary target for Arabidopsis responsive transcriptome to nitrogen limitation and its transcriptional regulation by SPLAYED in dynamic control of stem cell regulation by the ubiquitin ligase gene NLA. Plant Mol Biol 65: fate in Arabidopsis. Genes Dev 19: 992–1003 775–797 Lai EC (2004) Predicting and validating microRNA targets. Genome Biol Peng M, Hannam C, Gu H, Bi YM, Rothstein SJ (2007b) A mutation in 5: 115 NLA, which encodes a RING-type ubiquitin ligase, disrupts the adapt- LeeY,Kim M, HanJ,Yeom KH, LeeS,Baek SH,Kim VN (2004) ability of Arabidopsis to nitrogen limitation. Plant J 50: 320–337 MicroRNA genes are transcribedbyRNA polymerase II.EMBOJ 23: Peng M, Hudson D, Schofield A, Tsao R, Yang R, Gu H, Bi YM, Rothstein 4051–4060 SJ (2008) Adaptation of Arabidopsis to nitrogen limitation involves Li WX,OonoY,Zhu J, He XJ, Wu JM, Iida K, Lu XY,Cui X, Jin H,Zhu JK induction of anthocyanin synthesis which is controlled by the NLA (2008) The Arabidopsis NFYA5 transcription factor is regulated tran- gene. J Exp Bot 59: 2933–2944 scriptionally and posttranscriptionally to promote drought resistance. Ramakers C, Ruijter JM, Deprez RH, Moorman AF (2003) Assumption- Plant Cell 20: 2238–2251 free analysis of quantitative real-time polymerase chain reaction (PCR) Lin SI, Chiang SF, Lin WY, Chen JW, Tseng CY, Wu PC, Chiou TJ (2008) data. Neurosci Lett 339: 62–66 Rhoades MW, Reinhart BJ, Lim LP, Burge CB, Bartel B, Bartel DP (2002) Regulatory network of microRNA399 and PHO2 by systemic signaling. Plant Physiol 147: 732–746 Prediction of plant microRNA targets. Cell 110: 513–520 LindowM,JacobsenA,Nygaard S,Mang Y,Krogh A (2007) Intragenomic Scheible WR, Morcuende R, Czechowski T, Fritz C, Osuna D, Palacios- matching reveals a huge potential for miRNA-mediated regulation in Rojas N, Schindelasch D, Thimm O, Udvardi MK, Stitt M (2004) plants. PLoS Comput Biol 3: e238 Genome-wide reprogramming of primary and secondary metabolism, Lindow M, Krogh A (2005) Computational evidence for hundreds of non- protein synthesis, cellular growth processes, and the regulatory infra- conserved plant microRNAs. BMC Genomics 13: 119 structure of Arabidopsis in response to nitrogen. Plant Physiol 136: Liu HH, Tian X, Li YJ, Wu CA, Zheng CC (2008) Microarray-based analysis 2483–2499 of stress-regulated microRNAs in Arabidopsis thaliana. RNA 14: Schmid M, Davison TS, Henz SR, Pape UJ, Demar M, Vingron M, 836–843 Scho¨lkopf B, Weigel D, Lohmann JU (2005) A gene expression map of ´ ´ ´ Lodeiro AR, Gonzalez P, Hernandez A, Balague LJ, Favelukes G (2000) Arabidopsis thaliana development. Nat Genet 37: 501–506 Comparison of drought tolerance in nitrogen-fixing and inorganic Schmittgen TD, Jiang J, Liu Q, Yang L (2004) A high-throughput method nitrogen-grown common beans. Plant Sci 154: 31–41 to monitor the expression of microRNA precursors. Nucleic Acids Res Lu C, Meyers BC, Green PJ (2007) Construction of small RNA cDNA 32: e43 libraries for deep sequencing. Methods 43: 110–117 Schubert D, Primavesi L, Bishopp A, Roberts G, Doonan J, Jenuwein T, 1554 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Goodrich J (2006) Silencing by plant Polycomb-group genes requires Mueller-Roeber B, Jarmolowski A, Szweykowska-Kulinska Z (2009) dispersed trimethylation of histone H3 at lysine 27. EMBO J 25: 4638–4649 Gene structures and processing of Arabidopsis thaliana HYL1- Selbach M, Schwanha¨user B, Thierfelder N, Fang Z, Khanin R, Rajewsky dependent pri-miRNAs. Nucleic Acids Res 9: 3083–3093 N (2008) Widespread changes in protein synthesis by microRNAs. Szittya G, Moxon S, Santos DM, Jing R, Fevereiro MP, Moulton V, Nature 455: 58–63 Dalmay T (2008) High-throughput sequencing of Medicago truncatula Sharma VK, Ramirez J, Fletcher JC (2003) The Arabidopsis CLV3-like short RNAs identifies eight new miRNA families. BMC Genomics (CLE) genes are expressed in diverse tissues and encode secreted 9: 593 proteins. Plant Mol Biol 51: 415–425 Udvardi MK, Czechowski T, Scheible WR (2008) Eleven golden rules of Shi R, Chiang VL (2005) Facile means for quantifying microRNA expres- quantitative RT-PCR. Plant Cell 20: 1736–1737 sion by real-time PCR. Biotechniques 39: 519–525 van Noorden GE, Ross JJ, Reid JB, Rolfe BG, Mathesius U (2006) Shin R, Berg RH, Schachtman DP (2005) Reactive oxygen species and root Defective long-distance auxin transport regulation in the Medicago hairs in Arabidopsis root response to nitrogen, phosphorus and potas- truncatula super numeric nodules mutant. Plant Physiol 140: 1494–1506 sium deficiency. Plant Cell Physiol 46: 1350–1357 Wu G, Poethig RS (2006) Temporal regulation of shoot development in Sunkar R, Chinnusamy V, Zhu J, Zhu JK (2007) Small RNAs as big players Arabidopsis thaliana by miR156 and its target SPL3. Development 133: in plant abiotic stress responses and nutrient deprivation. Trends Plant 3539–3547 Sci 12: 301–309 Yamasaki H, Abdel-Ghany SE, Cohu CM, Kobayashi Y, Shikanai T, Pilon Sunkar R, Kapoor A, Zhu JK (2006) Posttranscriptional induction of two M (2007) Regulation of copper homeostasis by micro-RNA in Arabi- Cu/Zn superoxide dismutase genes in Arabidopsis is mediated by dopsis. J Biol Chem 282: 16369–16378 downregulation of miR398 and important for oxidative stress tolerance. Yoo BC, Kragler F, Varkonyi-Gasic E, Haywood V, Archer-Evans S, Lee Plant Cell 18: 2051–2065 YM,Lough TJ,Lucas WJ (2004) A systemic small RNA signaling system Sunkar R, Zhu JK (2004) Novel and stress-regulated microRNAs and other in plants. Plant Cell 16: 1979–2000 small RNAs from Arabidopsis. Plant Cell 16: 2001–2019 Zhang Y (2005) miRU: an automated plant miRNA target prediction server. Szarzynska B, Sobkowiak L, Pant BD, Balazadeh S, Scheible W-R, Nucleic Acids Res 33: W701–W714 Plant Physiol. Vol. 150, 2009 1555 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLANT PHYSIOLOGY Oxford University Press

Identification of Nutrient-Responsive Arabidopsis and Rapeseed MicroRNAs by Comprehensive Real-Time Polymerase Chain Reaction Profiling and Small RNA Sequencing   

Loading next page...
 
/lp/oxford-university-press/identification-of-nutrient-responsive-arabidopsis-and-rapeseed-50bMovsuPA

References (95)

Publisher
Oxford University Press
Copyright
© 2009 American Society of Plant Biologists
ISSN
0032-0889
eISSN
1532-2548
DOI
10.1104/pp.109.139139
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Identification of Nutrient-Responsive Arabidopsis and Rapeseed MicroRNAs by Comprehensive Real-Time Polymerase Chain Reaction Profiling and Small 1[C][W][OA] RNA Sequencing 2 2 2 2 Bikram Datt Pant , Magdalena Musialak-Lange ,Przemyslaw Nuc ,Patrick May , Anja Buhtz, Julia Kehr, Dirk Walther, and Wolf-Ru¨ diger Scheible* Max-Planck Institute of Molecular Plant Physiology, 14476 Potsdam, Germany Comprehensive expression profiles of Arabidopsis (Arabidopsis thaliana) MIRNA genes and mature microRNAs (miRs) are currently not available. We established a quantitative real-time polymerase chain reaction platform that allows rapid and sensitive quantification of 177 Arabidopsis primary miR transcripts (pri-miRs). The platform was used to detect phosphorus (P) or nitrogen (N) status-responsive pri-miR species. Several pri-miR169 species as well as pri-miR398a were found to be repressed during N limitation, whereas during P limitation, pri-miR778, pri-miR827, and pri-miR399 species were induced and pri-miR398a was repressed. The corresponding responses of the biologically active, mature miRs were confirmed using specific stem-loop reverse transcription primer quantitative polymerase chain reaction assays and small RNA sequencing. Interest- ingly, the latter approach also revealed high abundance of some miR star strands. Bioinformatic analysis of small RNA sequences with a modified miRDeep algorithm led to the identification of the novel P limitation-induced miR2111, which is encoded by two loci in the Arabidopsis genome. Furthermore, miR2111, miR169, a miR827-like sequence, and the abundances of several miR star strands were found to be strongly dependent on P or N status in rapeseed (Brassica napus) phloem sap, flagging them as candidate systemic signals. Taken together, these results reveal the existence of complex small RNA-based regulatory networks mediating plant adaptation to mineral nutrient availability. In recent years, approximately 21-nucleotide-long et al., 2006; Sunkar et al., 2007). MiR395 and miR399 microRNAs (miRs) have been recognized as important have been shown to be specifically induced during regulators of gene expression in animals and plants sulfur and phosphorus (P) limitation, respectively (Bartel, 2004; Dugas and Bartel, 2004; Kidner and (Jones-Rhoades and Bartel, 2004; Fujii et al., 2005; Martienssen, 2005; Chuck et al., 2009). In plants, miRs Bari et al., 2006; Chiou et al., 2006; Kawashima et al., were shown to posttranscriptionally regulate diverse 2008). MiR399 targets the transcript of an E2-conjugase aspects of development like leaf polarity (Emery et al., that is mutated in phosphate (Pi)-accumulating Arabi- 2003), leaf shape (Palatnik et al., 2003), the transition dopsis (Arabidopsis thaliana) pho2 mutants, and this from the juvenile to the mature growth phase (Wu phenotype is recapitulated in miR399-overexpressing and Poethig, 2006), flowering time (Aukerman and plants (Aung et al., 2006; Bari et al., 2006). MiR399 Sakai, 2003), stomatal development (Kutter et al., presumably acts in a dualistic manner to inhibit PHO2. 2007), and nodule development (Combier et al., 2006). It promotes transcript decay (Allen et al., 2005; Bari MiRs also regulate the adaptation of plants to abiotic et al., 2006) but also appears to inhibit PHO2 expres- stresses, including macronutrient limitations (Sunkar sion by repressing translation (Bari et al., 2006), a and Zhu, 2004; Fujii et al., 2005; Bari et al., 2006; Chiou mechanism that is probably widespread in plants (Aukerman and Sakai, 2003; Brodersen et al., 2008). Recently, miR167 has been associated with lateral root This work was supported by the Max-Planck Society. These authors contributed equally to the article. outgrowth in response to nitrogen (N) limitation * Corresponding author; e-mail scheible@mpimp-golm.mpg.de. (Gifford et al., 2008). Primary miR transcript 167a The author responsible for distribution of materials integral to the (pri-miR167a) was shown to be approximately 5-fold findings presented in this article in accordance with the policy repressed by N in root pericycle cells, permitting the described in the Instructions for Authors (www.plantphysiol.org) is: target ARF8 transcript to accumulate and initiate lateral Wolf-Ru¨diger Scheible (scheible@mpimp-golm.mpg.de). root outgrowth. Furthermore, miR398 was reported to [C] Some figures in this article are displayed in color online but in be down-regulated by oxidative stresses and copper black and white in the print edition. [W] limitation but induced by Suc, thereby repressing two The online version of this article contains Web-only data. [OA] copper/zinc (Cu/Zn) superoxide dismutases and a Open Access articles can be viewed online without a sub- scription. cytochrome c oxidase subunit (Sunkar et al., 2006; www.plantphysiol.org/cgi/doi/10.1104/pp.109.139139 Yamasaki et al., 2007; Dugas and Bartel, 2008). Plant Physiology , July 2009, Vol. 150, pp. 1541–1555, www.plantphysiol.org  2009 American Society of Plant Biologists 1541 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Despite these examples, little information about qRT-PCR. Although pri-miRs are not the biologically stress- or nutrient-responsive plant miRs is available. active molecules, several previous studies have shown This is due to their often low expression levels and the that the response of a pri-miR can reflect that of the absence of miR or MIRNA gene probes on widely used encoded mature miR (Jones-Rhoades and Bartel, 2004; transcriptomics platforms such as Affymetrix Gene- Bari et al., 2006; Pant et al., 2008) and thus can serve as Chips. Custom-made microarrays can be designed to a valid indicator. Therefore, qRT-PCR profiling of pri- include probes for miRs and MIRNA genes for a miRs can serve as a useful tool to discover responses to broader response analysis, but these are not very particular stimuli, which can then be confirmed by sensitive (Axtell and Bartel, 2005; Liu et al., 2008). analysis of the mature species. Reverse transcription followed by quantitative PCR In this work, we first established the pri-miR plat- analysis (qRT-PCR) with nonspecific double-stranded form and then used it to discover previously unknown DNA-binding fluorophores, such as SYBR Green, is a nutrient-responsive pri-miRs. The corresponding ma- powerful alternative for highly sensitive, rapid, multi- ture miRs were investigated by targeted assays and parallel, and cost-effective expression analysis (Udvardi further confirmed by small RNA sequencing, which et al., 2008). Shi and Chiang (2005) and Chen et al. also revealed novel insights. The results indicate that (2005) reported two qRT-PCR-based methods to mea- small RNAs play a much more important role in sure the levels of mature miRs. The first approach nutrient signaling than previously thought. relies on in vitro polyadenylation of mature miRs followed by RT with an oligo(dT) adapter primer and RESULTS amplification using SYBR Green with a miR-specific forward primer and a compatible reverse primer. In A qRT-PCR Platform for Arabidopsis pri-miRs the second approach, each specific miR is reverse transcribed from total RNA using a specific stem-loop Sequences of 184 annotated Arabidopsis miR stem primer, followed by TaqMan PCR amplification. Al- loops were obtained from the miRBase database though it is desirable to quantify the biologically active (www.microrna.sanger.ac.uk). These sequences are not mature miR species, a limitation of both qRT-PCR strictly pre-miRs but may include flanking sequence methods is that they are unable to differentiate the from the presumed pri-miR. Pri-miR sequences of expression strengths of MIRNA genes that yield members from the same family can be almost identi- (nearly) identical mature miR molecules. Hence, the cal, complicating the design of specific PCR primers. interpretation of the results will be dominated by This occurs in the miR169 family, where miR169i strongly expressed member(s) of a given MIRNA gene through miR169n are located in three highly homolo- family, while the contribution of lowly expressed gous, tandem-arrayed stretches (Supplemental Fig. members will go unnoticed. This might be especially S1), and in the miR854 family, where the pre-miR important if these genes are expressed in an organ- or sequences of the four annotated members are 97% to cell type-specific manner. 100% identical. However, it is questionable whether Deep sequencing using new technologies (e.g. miR854s are true miRs, as the sequences are located Illumina-Solexa chemistry) is another approach being close to (or in) the centromere of chromosome 5 and adopted for the analysis of small RNA/miR abun- are annotated as transposable elements in The Arabi- dance (German et al., 2008; Moxon et al., 2008a; Szittya dopsis Information Resource (TAIR) database (www. et al., 2008). The high number of sequence reads arabidopsis.org). Therefore, we treated miR854 as a promises sensitivity, yet the necessary expertise re- single miR, leaving a total of 181 sequences for which quired and the labor and cost involved are consider- primers were designed. able. Data from small RNA sequencing together with To ensure maximum specificity and efficiency dur- miR prediction algorithms like miRDeep (Friedla¨nder ing PCR amplification of pri-miR cDNA under a stan- et al., 2008) also provide the basis for the discovery of dard set of reaction conditions (Fig. 1A), a stringent set novel miRs, even in organisms that have been specif- of criteria was used for primer design (see “Materials ically surveyed to identify miRs. and Methods”). PCR primers were tested on cDNA We have developed a qRT-PCR platform for parallel from Arabidopsis wild-type ecotype Columbia (Col-0) analysis of 177 currently known Arabidopsis MIRNA seedlings, which was free of genomic DNA contami- gene pri-miRs. This platform provides a sensitive yet nation. Using this cDNA as template, 150 primer pairs inexpensive tool for Arabidopsis researchers to carry gave unique PCR products of the expected size, while out miR expression analysis. A comparable approach 27 primer pairs yielded no product and four gave has previously been described for monitoring the unspecific products. The 27 primer pairs were retested expression of 23 human miR precursors (Schmittgen using Arabidopsis Col-0 genomic DNA as template. et al., 2004). As pri-miRs are generated by RNA Fifteen primer pairs resulted in the expected genomic polymerase II in plants and animals (Bracht et al., product, showing that the primers anneal to the 2004; Cai et al., 2004; Kurihara and Watanabe, 2004; correct sequence and suggesting that the targeted Lee et al., 2004), they contain 5# caps and 3# poly(A) pri-miRs (e.g. 159c, 166c, 395a to 395f, and 404; Sup- tails. The latter make the transcripts amenable to oligo plemental File S2) were below the detection limit in the (dT)-primed RT and thus multiparallel analysis by cDNA samples or that amplification from the cDNA 1542 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs was inhibited. Evidence in support of the former hypothesis comes from the observation that some of the primers (e.g. pri-miR395a to -395f) did amplify the expected products from a cDNA sample derived from sulfur-limited seedlings (Supplemental Fig. S2). The 16 primer pairs that did not yield any product from cDNA or genomic DNA templates, or that amplified unex- pected/unspecific products, were redesigned, finally resulting in 175 validated primer pairs and only six MIRNA genes (highlighted in red in Supplemental File S2) for which no working pairs could be established. Specificity of PCR primers was assessed by melting curve analysis of PCR products (Supplemental Fig. S3), by separating the PCR products via electrophore- sis on high-resolution agarose gels (Fig. 1B), and by double-stranded sequencing of a subset of the pri-miR PCR products (Supplemental Fig. S4). In all cases, the sequences of the PCR products were identical to those of the intended pri-miR targets. The average amplifi- cation efficiency (E) of the primers, as determined by LinRegPCR (Ramakers et al., 2003), was high; for 102 primer pairs, E was greater than 90%, and for another 39 pairs, E was 81% to 90% (Fig. 1C; Supplemental File S2). Twenty-one primer pairs did amplify a correct product from genomic DNA, but no product was obtained with cDNA from the samples investigated (see below); hence, no E value could be established for these primers. The fraction of pri-miRs detected (i.e. expressed in at least two biological replicates with a threshold fluo- rescence cycle number [C ] of less than 40) was just below 80%, irrespective of the growth conditions tested (Supplemental File S2). This percentage is com- parable to the percentage of transcription factor (TF) genes detected at this threshold (approximately 83%; Czechowski et al., 2004). However, the average DC value of pri-miRs was higher than for the TF gene transcripts (Fig. 1D). Hence, pri-miRs are often less abundant than TF transcripts. Low abundance (C . 33) also resulted in higher variability between replicate measurements, making detection of small expression changes less reliable. Figure 1. qRT-PCR platform for pri-miR transcript quantification. A, Identification of N- and P-Responsive Typical real-time RT-PCR amplification plots of 175 miR amplicons Arabidopsis pri-miRs showing increase in SYBR Green fluorescence (DRn; log scale) with PCR cycle number. Note the similar slope of most curves as they cross The qRT-PCR platform was used to identify pri-miRs the fluorescence threshold of 0.18 (green line), which reflects similar that are induced or repressed in 9-d-old Arabidopsis amplification efficiency. B, Separation of PCR products on 4% (w/v) seedlings during N or P limitation (Supplemental File high-resolution agarose gels reveals single products of the expected S2). The expected physiological status of the seedlings size (black numbers). A selection of 10 amplicons is shown. Size was confirmed by evaluation of marker gene expression standards in bp are indicated to the left. C, Distribution of efficiencies for 154 primer pairs (gray bars). For 21 primer pairs (white bar) no (Supplemental Fig. S5). NRT2.5 (At1g12940)and primer efficiencies were obtained in the conditions investigated. D, AMT1.5 (At3g24290) were both strongly induced in Expression strength distribution of MIRNA genes (white bars) and of N-limited seedlings, and PHT1.4 (At2g38940) was in- transcription factor genes (black bars). Genes were grouped according duced by P limitation, as found previously (Scheible to their DC values, calculated with UBQ10 as reference. [See online et al., 2004; Morcuende et al., 2007). article for color version of this figure.] Twenty pri-miRs exhibited differential expression in N- or P-limited conditions (Fig. 2), based on an aver- age change in normalized cycle number of at least three (|DDC | $ 3) between nutrient limitation and Plant Physiol. Vol. 150, 2009 1543 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. unknown: pri-miR447c, -778, and -827 all increased (DDC = 4.2–7.6) during P starvation, whereas pri- miR398a strongly decreased (DDC = 26.9; Fig. 2; Supplemental File S2). Also, pri-miRs 169m and 169n displayed induction (DDC = 3.7–4.2) during P limitation, and the same two pri-miR169s plus five additional ones (pri-miR169h through -169l) were decreased (DDC = 23.1 to 24.9) in N-limited seed- lings (Fig. 2). Pri-miR398a and pri-miR447c were not only respon- sive to P limitation but also showed similar responses in N limitation, albeit not as strong, with pri-miR398a being slightly repressed (DDC = 22.45; Supplemental File S2) and pri-miR447c induced (DDC = 3.5; Fig. 2). Furthermore, pri-miRs156e, -156g, and -157d were found to be induced (DDC = 3.1–4.4) in N-limited seedlings. Pri-miR167a, which was reported to be more highly expressed in N-limited root pericycle cells (Gifford et al., 2008), showed no clear response in our experiments with N-limited seedlings. However, pri-miR167d was less expressed (DDC = 22.5) in N-limited seedlings (Supplemental File S2). Nutrient-Responsive Mature miR Species To examine if the mature miRs derived from the nutrient-responsive pri-miRs also showed a nutrient response, we used a qRT-PCR approach similar to the one described by Chen et al. (2005; Supplemental Table S1). In addition to nutrient-replete (FN), N-limited Figure 2. Identification of N/P limitation-responsive pri-miRs in Arabi- (2N), and P-limited (2P) Arabidopsis seedlings, we dopsis. For each pri-miR species, the average DDC value 6 SE from also included carbohydrate-limited (2C) seedlings three biological replicates (with two technical replicates for each) is (Supplemental Fig. S5) in the analysis. Mature depicted. DDC = DC 2 DC , and DC =C 2 C . T TFN T -nutrient T T pri-miR T UBQ10 miR399s were not analyzed, since their high abun- Pri-miRs induced in nutrient limitation thus have a positive DDC value dance during P limitation is already well documented and vice versa. The fold induction can be inferred from the equation (see introduction). DDC (1 + E) , where E is PCR efficiency (Supplemental File S2). Only pri- MiR778 and miR827 were both strongly induced by miRs with average DDC values of .3or ,23 (as indicated by the light P limitation, thus confirming the response of their pri- gray shading) are shown. Results for P and N limitation are shown as miRs. However, they did not respond to either N or C dark gray and white bars, respectively. limitation (Fig. 3) and remained almost undetectable under these conditions, suggesting that both of these full nutrition (FN). The most prominent change was miRs are involved in P-specific regulation events. The the known induction of pri-miR399s during P limita- response of miR398a was also similar to that of the tion, with DDC ranging from 6.5 to 20 (Fig. 2; Bari corresponding pri-miR, being decreased by P and N et al., 2006). Pri-miR399d was undetectable in FN limitation and also by C limitation (Fig. 3). conditions and rose to levels comparable to some of Given the previous report of Gifford et al. (2008) and the most strongly expressed genes in Arabidopsis, the moderate repression of pri-miR167d we observed such as UBQ10 (Czechowski et al., 2005). Very high in N-limited seedlings, we also investigated miR167. expression of pri-miR399s and mature miR399 during The assay for miR167 showed a relatively high ex- P limitation is necessary to fully suppress the activity pression level (40 2 DC = 35), but this might be due to of its target PHO2 transcript (Bari et al., 2006). Mech- low specificity in the assay leading to detection of anistically, this seems to involve translational repres- miR167 derived from several primary transcripts. Our sion, as suggested by previous results (Bari et al., 2006) result obtained with N-limited seedlings (Fig. 3) does and our finding that miR399d-overexpressing seed- not confirm the reported down-regulation of miR167 lings have substantially lower but still detectable in root pericycle cells by organic N (Gifford et al., levels of PHO2 transcript (10%–20% of control plants), 2008). Possible explanations for this discrepancy are while their molecular phenotypes are identical to pho2 the nonspecificity of the assay for miR167 and/or the seedlings (Supplemental Fig. S6). lack of spatial resolution in our analysis, which could qRT-PCR profiling revealed several pri-miR species mask any cell type-specific response of a particular for which nutrient responsiveness was previously MIRNA gene and its derived miR. 1544 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs quencing (SRS) with Illumina-Solexa technology. Three cDNA libraries prepared from nutrient-replete (FN) seedlings, P-limited (2P) seedlings, and 2P seedlings that were resupplied with 3 mM phosphate for 3 h were sequenced (Supplemental Table S2). Sequence reads with 100% identity to Arabidopsis pre-miRs were extracted, and identical reads were totaled (Supplemental Table S2; Supplemental Fig. S7; Supplemental File S3) and normalized for each library. A plot of the distribution of read lengths for pre-miR- matching sequences (Supplemental Fig. S7A) illus- trates that these consist almost exclusively of 20- and 21-mers, with the latter being the most abundant, whereas a plot of all genome-matching sequences reveals a substantial number of 24-mers and other lengths (Supplemental Fig. S7B). Comparison also shows a significantly (approximately 30%) higher number of 21-mers in the group of genome-matching sequences, possibly indicating the presence of un- known miRs. The normalized read numbers for miR399s were Figure 3. Nutrient responsiveness of mature miRs. qRT-PCR results are high during P limitation and very low in nutrient- shown for nutrient-replete (white bars), P-limited (black bars), replete (FN) conditions (Fig. 4A), as expected (Bari N-limited (gray bars), and C-limited (hatched bars) Arabidopsis seedlings. et al., 2006; Pant et al., 2008). There was no appreciable Expression levels are given on a logarithmic scale expressed as 40 2 decrease of miR399s at 3 h after Pi readdition, sup- DC , where DC is the difference in qRT-PCR threshold cycle number of T T porting earlier results that suggested that miR399s are the respective miR and the reference gene UBQ10 (At4g05320); rather stable (Bari et al., 2006). SRS also confirmed high therefore, 40 equals the expression level of UBQ10 (the number 40 expression of miR778 and miR827 during P limitation was chosen because the PCR run stops after 40 cycles). The results are averages 6 SE of three biological replicates. Significance of the changes (Fig. 4A). Very few (one or two) or no reads were found during P, N, or C limitation was checked with Student’s t test. P , obtained for miR398a, -447c, -845, or -169h to -n, 0.05 is indicated by circles, and P , 0.01 is indicated by plus signs. whereas these miRs were detectable by qRT-PCR, suggesting that our qRT-PCR analysis was more sen- sitive than the SRS approach. SRS showed three addi- Nucleolytic cleavage of pri-miR169h to -n (Supple- tional miRs (i.e. miR408, miR829, and miR863) to be mental Fig. S1) by the DICER endoribonuclease results 4- to 10-fold more abundant during P limitation. Anal- in identical miR169 molecules, precluding a specific ysis by qRT-PCR supported the increase in levels of assay. Nonetheless, since all seven pri-miR sequences miR408 but was unable to confirm the higher abun- were less abundant in N limitation (Fig. 2), we were able dance of miR829 or miR863 (Supplemental Fig. S8). to confirm lower miR169h to -n levels in N limitation (Fig. Surprisingly, sequence reads representing star (*) 3). The moderate induction found for two of these pri- strands of some of the nutrient-responsive miRs (i.e. miR169 species during P limitation was not supported miR398a*, miR399a*, miR399c*, miR399d*, miR399f*, by the miR169h to -n assay; on the contrary, mature and miR778*) were found to be present in numbers miR169h to -n showed a significant (approximately similar to, or exceeding (up to 200-fold in the case of 3-fold) decrease in P limitation. A second assay designed miR398a*), those for the corresponding miRs (Fig. 4A; for miR169a to -g also indicated lower abundance during Supplemental File S3), whereas others (e.g. miR399b*, N and P limitation (Fig. 3), although no clear change was miR827*, and miR863*) were absent. The presence of detected for the corresponding pri-miRs. star strands and their specific induction by Pi limita- We were also able to confirm a slight induction tion was confirmed by qRT-PCR (Supplemental Fig. (DDC w 2) of miR447 during P but not N limitation, S8). These results raise the question of the biological whereas the induction of miR156 during N limitation, function of these abundant miR*s. Another interesting as suggested by pri-miR156e and -156g (Fig. 2), was observation from the SRS was that the sequence reads not confirmed (Fig. 3). Again, a nonspecific assay and corresponding to miR778 and miR863 do not perfectly high expression of other pri-miRs from the same agree with the annotated miR sequences determined family (Supplemental File S2) probably account for from flower samples (Fahlgren et al., 2007) but look to this discrepancy. be shifted by one or several nucleotides (Supplemental File S3). Furthermore, the miR778/miR778* duplex P Status-Responsive miRs Detected by Small appears to be unusual in having a 5# overhang that is RNA Sequencing seven nucleotides long (Supplemental File S3). These results from Arabidopsis seedlings could indicate that As an independent approach to verify the P respon- siveness of mature miRs, we used small RNA se- the cleavage site of some pre-miRs by DICER proteins Plant Physiol. Vol. 150, 2009 1545 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Figure 4. P-responsive miRs as detected by small RNA sequencing. A, The number of normalized sequence reads (ppm) is shown for all miRs and their corresponding star (*) strands that show at least a 5-fold change between P limitation (black bars) and P-replete (white bars) conditions and for which at least 10 total reads were scored in at least one condition (Supplemental File S3). Normalized read numbers are also shown for a time point of 3 h after resupply of 3 mM potassium phosphate to previously P-limited seedlings (3hP; gray bars). Missing bars indicate zero reads. B, Putative miR2111a and -b precursor sequences and structures. The sequences of mature miR2111 and the presumed star strands are shown in boldface on the top and bottom strands, respectively. The number of absolute reads for the three conditions tested (see A) is indicated. C, qRT-PCR expression of the 82- and 103-nucleotide-long pri-miR2111a and -b amplicons in various conditions (FN, full nutrients; 2P, P limitation; 3hP, 3 h of Pi readdition; 2N, N limitation; 2C, carbohydrate limitation). Expression levels are plotted on a logarithmic scale as described in the legend to Figure 3. The results are averages 6 SE of three biological replicates with two technical replicates for each. is not totally fixed and may change with developmen- dence on P status (Fig. 4A). It is highly abundant tal stage. during P limitation but almost absent in nutrient- replete seedlings and so closely resembles the behav- ior of known P-responsive miRs (miR399s, miR778, Discovery of a Novel P Status-Responsive miR and miR827) in these conditions. When Pi is resup- The SRS data were analyzed further to look for novel plied to Pi-limited seedlings, the abundance of the P status-dependent miRs. Candidate miRs were pre- novel miR2111 fell by approximately 2-fold within 3 h dicted using a miRDeep algorithm (Friedlander et al., of Pi readdition (Fig. 4A). This response differs from 2008) that was adapted for plant miR precursor se- those of miR399s and miR827, which do not fall ¨ rapidly after Pi readdition, but is similar to that of quences (P. May and M. Friedlander, unpublished miR778, suggesting that the biological half-lives of data). Among the predicted miR candidates was one, named miR2111, that showed a very strong depen- miRs can vary considerably. 1546 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs The DNA sequence of miR2111 is present twice in was previously reported by Giavalisco et al. (2006) and the Arabidopsis genome: upstream of At5g02040, and Buhtz et al. (2008). between At3g09280 and At3g09290. For these loca- Further analysis revealed that in addition to Bna- tions, the algorithms miRDeep and miRCat (http:// miR399 (Fig. 5B; Pant et al., 2008), a Bna-miR399-like srna-tools.cmp.uea.ac.uk) predicted two precursor se- sequence, Bna-miR2111, a Bna-miR2111-like sequence, quences/structures, named pre-miR2111a and -b (Fig. and an Ath-miR827-like sequence are highly abundant 4B). The observation that the read numbers of the two in rapeseed phloem sap during P limitation, while no star strands (miR2111a* and miR2111b*) are similar sequences homologous to the P-responsive Ath- and that they decrease after Pi readdition or are almost miR778 were found. We were also able to determine absent in P-replete conditions reveals that both loci that (1) two sequences with homology to Ath- have P-responsive expression and contribute to a miR399d* and Ath-miR399f*, (2) Bna-miR2111a* and similar extent. The strong P limitation response of Bna-miR2111b*, and (3) five miR2111*-like sequences miR2111 was confirmed by qRT-PCR (Supplemental were clearly present in phloem sap during P limitation Fig. S8). In addition, PCR products designed to en- but absent or much less abundant in phloem sap of compass larger stretches (82 and 103 nucleotides) of nutrient-replete or N-limited rapeseed (Fig. 5B). Fur- pre-miR2111a and -b could be amplified from oligo thermore, we found Bna-miR169m to be present in (dT)-primed cDNA pools, showing that miR2111 is phloem sap of nutrient-replete plants, and the relative derived from poly(A)-tailed primary transcripts, and abundance of this species strongly decreased in these also displayed strong P responsiveness (Fig. 4C). phloem sap of N-limited or P-limited rapeseed plants Furthermore, both the mature miR2111 and its pri- (Fig. 5B). These results pinpoint miR2111s, miR2111*s, miRs displayed specificity for P, as N or C limitation miR399*s, miR827, and miR169 as novel candidate did not affect the expression levels (Fig. 4C; Supple- long-distance signals for reporting P or N status in the mental Fig. S8). We were also able to find potential plant system and suggest an expansion of the miR2111 orthologs of pre-miR2111s in rapeseed (Brassica napus; family in rapeseed. Supplemental Fig. S9). The rapeseed 2111a and 2111b precursors share 85% and 83% identity at the DNA Target Predictions for Nutrient-Regulated miRs level with their Arabidopsis counterparts and are predicted to fold into stable extended hairpin struc- To identify candidate miR target genes, we used tures by the miRCat algorithm. PCR primer pairs for four prediction algorithms: miRU (Zhang, 2005), target Arabidopsis pri-miR2111a and -2111b were added to search in WMD2 (Ossowski et al., 2008), the prediction the qRT-PCR platform, thus bringing the number tool in the UEA plant sRNA toolkit (Moxon et al., of MIRNA genes represented to 177 (Supplemental 2008b), and PITA (Kertesz et al., 2007). In addition, we File S2). mined experimental data from degradome studies (Addo-Quaye et al., 2008; German et al., 2008). The first three algorithms score the complementarity of MiRs with P or N Status-Dependent Abundance in miR and target RNA sequences based on established Rapeseed Phloem Sap rules (e.g. the “seed rule”; Allen et al., 2005), thereby MiR399 was previously found to be highly abun- also exploring the observation that plant miRs seem to dant in phloem sap from rapeseed and pumpkin bind almost perfectly to their cognate mRNAs (Cucurbita maxima) during P limitation and to consti- (Rhoades et al., 2002; Lai, 2004). In contrast, PITA tute a shoot-derived long-distance signal for the reg- assesses site accessibility in miR target recognition. ulation of plant Pi homeostasis (Buhtz et al., 2008; Lin This includes prediction of target RNA secondary et al., 2008; Pant et al., 2008). To test whether additional structure and calculation of the free energy gained miRs show nutrient status-dependent abundance in from the formation of the miR-target duplex and the phloem sap, we collected sap from nutrient-replete, energetic cost of unpairing the target to make it acces- P-limited, and N-limited rapeseed plants and pre- sible to the miR. All candidate miR targets predicted pared small RNA libraries for sequencing. with miRU, WMD2, or the UEA plant sRNA toolkit, Analysis of the resulting small RNA reads (Supple- and up to 20 best targets from the PITA analysis, are mental Table S3) showed that sequences homologous shown in Supplemental File S4, with a selection of the to miRs known to be present in the phloem (e.g. predicted targets shown in Table I. miR156, miR159, and miR167) were present in the In addition to the confirmed miR399 target PHO2,a libraries, whereas sequences homologous to miR171, potential target of miR399b/c predicted by all algo- which is abundant in leaf or stem tissue but undetect- rithms is the receptor kinase gene ACR4, which re- able in phloem (Yoo et al., 2004; Buhtz et al., 2008), stricts formative cell divisions in the Arabidopsis root were completely absent (Fig. 5A). This indicates that (De Smet et al., 2008). Another repeatedly predicted the phloem sap samples were not noticeably contam- miR399b/c target is At4g00170, encoding a vesicle- inated with small RNAs from stem tissue. Small RNA associated membrane protein. A DEAD box helicase species found in the libraries should thus represent gene (At4g09730) appears as a candidate miR399a/d authentic phloem constituents. The high purity of target from analysis with conventional prediction al- rapeseed phloem sap obtained using the same method gorithms, whereas PITA analysis identifies IPS1,a Plant Physiol. Vol. 150, 2009 1547 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Figure 5. miRs with P or N status-dependent abundance in rapeseed phloem sap. A, Abundance of miRs known to be present (miR156, miR159, and miR167) or absent (miR171) in rapeseed phloem sap. B, MiR and miR* sequences with P or N status- dependent abundance. Depicted are normalized read numbers as detected by Solexa sequencing in small RNA libraries prepared from phloem sap samples of P-starved (black bars), nutrient-replete (white bars), and N-starved (gray bars) rapeseed plants. The significance of the changes between P-starved (or N-starved) and nutrient-replete conditions was analyzed with a x test and Benjamini-Hochberg P value correction; P , 0.0003 is marked with circles. Sequences of the species shown are as follows: 169m, 5#-TGAGCCAAAGATGACTTGCCG-3#; 399, 5#-TGCCAAAGGAGATTTGCCCGG-3#; 399-like, 5#-TGC- CAAAGGAGATTTGTCCGG-3#; 399*-like 1, 5#-GGGCGAATACTCTTATGGCAGA-3#; 399*-like 2, 5#-GGGCAAGATCTC- TATTGGCAGA-3#; 2111, 5#-TAATCTGCATCCTGAGGTTTA-3#; 2111-like, 5#-TAATCGGCATCCTGAGGTTTA-3#; 2111a*, 5#-ATCCTCGGGATACAGATTACC-3#; 2111b*, 5#-GTCCTCGGGATGCGGATTACC-3#; 2111*-like 1, 5#-ATCCTCGGGATG- CGGATTACC-3#; 2111*-like 2, 5#-ATCCTCGGGATACGGATTACC-3#; 2111*-like 3, 5#-ATCCTCGGGACACAGATTACC-3#; 2111*-like 4, 5#-GACCTCAGGATGCGGATTACC-3#; 2111*-like 5, 5#-TCCTCGGGATACAGATTACC-3#; 156, 5#-TGACAG- AAGAGAGTGAGCAC-3#; 159, 5#-TTTGGATTGAAGGGAGCTCTA-3#; 167, 5#-TGAAGCTGCCAGCATGATCTA-3#; 171, 5#-TGATTGAGCCGCGCCAATATC-3#. miR399 target-like interactor (Franco-Zorrilla et al., the CSD2 miR398a duplex (Brodersen and Voinnet, 2007). There are also clear target predictions for 2009). miR399*s, with At3g11130, encoding a clathrin heavy chain and thus another protein involved in vesicle transport in the secretory pathway, being a potential DISCUSSION target of miR399d*. One of the best potential targets of miR399f* is the CLAVATA3-related At3g25905/CLE27, qRT-PCR of pri-miRs Is Suitable for Discovery of a gene encoding a small peptide that is highly ex- Stimulus-Dependent miRs pressed in shoot apices (Sharma et al., 2003). Obvious targets of miR827 are the E3 ligase gene The role of miRs during the adaptation of plants to NLA (At1g02860) and its homolog At1g63010 (Table I; abiotic and nutritional stresses is a field that attracts Fahlgren et al., 2007). Also, miR2111 is predicted to increasing interest (Chiou, 2007; Sunkar et al., 2007), target an E3 ligase gene (At3g27150) and a calcineurin- but information on stress-dependent expression of like phosphoesterase gene (At1g07010). A likely target MIRNA genes is limited. To overcome this situation, of miR778 is the histone methyltransferase gene we developed a qRT-PCR platform for quantification SUVH6 (At2g22740), with SUVH5 (At2g35160) being of almost all known Arabidopsis pri-miRs. The results a possible target as well. Interestingly, miR2111a* and obtained with the platform and from targeted assays miR2111b* also appear to target genes required for for mature miRs show that the responses of pri-miRs chromatin remodeling/modification (i.e. At2g23380/ frequently match the responses of the biologically CURLY LEAF and At2g28290/SPLAYED). active mature species, thus validating pri-miR profil- Confirmed targets of miR398a include two Cu/Zn ing as a useful discovery tool. It is also rapid, sensitive, superoxide dismutase genes (CSD1 and CSD2) and and inexpensive and should represent an attractive COX5b.1 (see introduction). The prediction algorithms initial approach for the investigation of MIRNA gene detected CSD1, COX5b.1,and At1g12520/CCS1, a chap- expression. The usefulness of the pri-miR platform is erone that activates CSD, as potential targets of also demonstrated by comparative analysis of the hyl1 miR398a (Table I). CSD2 was not found, most likely mutant and wild-type plants, revealing accumulation due to a bulge and GU wobble in the seed region of of specific pri-miRs in hyl1 and thus an important role 1548 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Table I. Selected targets of P- and N-responsive miRs/miR*s miR/miR* Stimulus Predicted Target Gene(s) Gene Product(s) abcdef 169a +N (?) Several NF-YA Several nuclear factor Y A subunits abcdef 169b/c +N (?) Several NF-YA Several nuclear factor Y A subunits abce At5g42120 Lectin protein kinase family protein abcdef 169d to -g +N (?) Several NF-YA Several nuclear factor Y A subunits ae be At1g70700 and At1g48500 Jasmonate-ZIM domain proteins 4 and 9 abcdef 169h to -n +N, +P Several NF-YA Several nuclear factor Y A subunits ab At5g42120 Lectin protein kinase family protein bdef 398a +P, +N, +C At1g08830 Cu/Zn superoxide dismutase CSD1 be At1g12520 Superoxide dismutase chaperone CCS1 cd At3g15640 Cytochrome c oxidase subunit COX5b.1 abcdef 399a 2P At2g33770 E2 conjugase PHO2 ac At4g09730 ATP-dependent RNA helicase At4g27850 Pro-rich protein abcdef 399b/c 2P At2g33770 E2 conjugase PHO2 abce At3g59420 ACR4 kinase abc At3g54700 Phosphate transporter bce At4g00170 VAMP family protein At4g27850 Pro-rich protein At3g09922 IPS1 abcdef 399d 2P At2g33770 E2 conjugase PHO2 ac At4g09730 ATP-dependent RNA helicase abcdef 399f 2P At2g33770 E2 conjugase PHO2 At3g09922 IPS1 abcde 778 2P At2g22740 Histone methyltransferase SUVH6 At2g35160 Histone methyltransferase SUVH5 At5g51980 WD40-repeat protein abcde 827 2P At1g02860 SPX E3 ligase NLA abce At1g63010 SPX E3 ligase abce 2111 2P At3g27150 F box protein abcde At2g23370 Unknown protein abcde At1g07010 Calcineurin-like phosphoesterase ae 398a* +P, +N, +C At4g00950 MEE47 (maternal effect embryo arrest 47) abce At5g06120 Ran-binding protein abe 399c* 2P At5g64470 DUF231 protein ae 399d* 2P At3g11130 Clathrin heavy chain abe 399f* 2P At3g25905 CLAVATA3/ESR-RELATED27 abe 778* 2P Ag1g69610 Ribosomal protein 2111a* 2P At2g28290 SWI2/SNF2-like protein SPLAYED abce 2111b* 2P At2g23380 SET domain protein CURLY LEAF abe At1g60380 NAC domain transcription factor 24 a b c d Predicted by miRU. Predicted by WMD2. Predicted by UEA toolbox. Predicted by e f degradome data. Predicted by PITA. Experimentally confirmed. for the HYL1 double-stranded RNA-binding protein platform as new MIRNA genes are discovered. In this in selective pri-miR maturation (Szarzynska et al., regard, we invite readers to send us relevant sequence 2009). We found only a few examples where the P or information. N response of pri-miRs was not confirmed for the corresponding mature miRs. Although this could have Evidence for miR2111 Being a True miR been due to technical reasons, regulation or attenua- tion at the level of miR maturation could also account The novel Pi-responsive miR2111 was revealed by a for these differences. version of miRDeep optimized for analysis of plant It is likely that the number of recognized Arabi- sequences (P. May, unpublished data), but this alone is dopsis MIRNA loci will further increase (Lindow and not conclusive proof that it is a true miR. A number of Krogh, 2005; Lindow et al., 2007), especially with revised criteria exist for annotation of plant miRs deeper analysis of SRS data, possibly reaching num- (Meyers et al., 2008). The primary and only criterion bers similar to those currently known for rice (Oryza that is necessary and sufficient for annotation as a miR sativa; 269), zebra fish (337), mouse (472), and human is conclusive evidence of precise biogenesis from a (678; http://microrna.sanger.ac.uk). The easily scal- qualifying stem loop, and this criterion is met by able qRT-PCR approach allows regular updates of the miR2111. First, miR2111 was shown to be derived from Plant Physiol. Vol. 150, 2009 1549 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. longer poly(A)-tailed precursor transcripts, as it was two or more complementary algorithms, giving possible to obtain precursor PCR amplicons from greater confidence in the predictions. Target analysis cDNA pools generated by RT with an oligo(dT) primer with PITA also indicated that inclusion of thermody- (Fig. 4C). Second, the two miR2111 precursors fold into namics of RNA-RNA interactions can change the characteristic hairpin structures with stabilities typical results greatly (Hofacker, 2007). Only PITA correctly for known miR precursors. Third, there is precise identified IPS1, a noncoding RNA that qualifies as excision of 21-nucleotide miR2111/miR2111* duplexes a bona fide miR399 target-like interactor (Franco- from the two stem-loop precursors (Supplemental File Zorrilla et al., 2007), emphasizing the complementarity S3). Fourth, miR2111 and miR2111a/b* are derived and predictive power of PITA. The number of poten- from opposite stem arms and form duplexes with tial targets identified with PITA (Supplemental File S4) typical 3# overhangs of two-nucleotide length. Fifth, further suggests that an individual plant miR could base pairing between miR2111 and the star strands is also have a larger range of action, for example, by extensive (Fig. 4B). Finally, the observed small RNA inducing widespread changes in protein synthesis, as abundance corresponds entirely to the duplexes. An- recently reported for several human miRs (Selbach cillary criteria for plant miR annotation include (1) the et al., 2008). existence of target genes, (2) conservation between Three P starvation-inducible miRs (miR399, species, and (3) biogenesis that is dependent on miR2111, and miR827) have confirmed or likely target DICER-like (DCL) proteins. There are obvious poten- genes involved in protein degradation via the 26S tial target genes for miR2111 (Table I), although the proteasome. MiR827 targets the E3 ligase gene NLA biological significance of these being targeted by (At1g02860). NLA transcript also drops 2- to 3-fold miR2111 is as yet unknown. We also detected pre- during P limitation when miR827 is highly expressed miR2111 homologs and found mature miR2111 in (Morcuende et al., 2007). NLA is crucial for anthocy- rapeseed (Fig. 5; Supplemental Fig. S9). Although the anin synthesis, and the nla mutant displays severely dependence of miR2111 biogenesis on DCL1 or DCL4 reduced anthocyanin content and early leaf senescence still needs to be tested, there is already overwhelming during N limitation (Peng et al., 2007a, 2007b, 2008). evidence that miR2111 is a true miR. However, during P limitation or during simultaneous P and N limitation, the mutant displays wild-type-like anthocyanin levels and no leaf senescence (Peng et al., More Widespread Regulation by Small RNAs during 2008), showing that the signal derived from P limita- P Limitation tion is sufficient to induce anthocyanin production in So far, miR399s have been the only small RNA species known to strongly increase during P limitation (see introduction). Five MIRNA399 genes that encode the slightly different mature miR399s exist in Arabi- dopsis. Still the only confirmed target of miR399s is PHO2, while IPS1 is a miR399 interactor (Franco- Zorrilla et al., 2007). This situation and expansion of the MIRNA399 gene family in other plant species such as Medicago truncatula (F. Krajinski, personal commu- nication) and rice (Lindow et al., 2007), however, indicate a larger miR399 regulatory network. The new finding that at least four additional miRs and several miR*s with strong P status-dependent expres- sion exist now suggests that regulation/signaling by small RNAs during P limitation is even more wide- spread. Regulatory activity of miR* species and their presence in argonaute complexes have been demon- strated previously (Mi et al., 2008; Okamura et al., 2008). The fact that the SRS read numbers for some miR* sequences clearly exceed the numbers for the corresponding miRs indicates that they are not merely by-products that are slowly degraded. Target Predictions and Biological Processes Potentially Figure 6. Hypothetical models for miR827 and miR169 functions. A, Affected by P-Regulated Small RNAs Model showing cross talk between P limitation and N limitation signaling pathways that affect anthocyanin synthesis. B, Model inte- We used several prediction algorithms and mined grating published features of the systemic regulation of nodulation by degradome data (Table I) to determine likely targets of CLE (Okamoto et al., 2009), SUNN/HAR1 (Krusell et al., 2002), and P-regulated small RNAs. This combinatorial approach HAP2 and miR169 (Combier et al., 2006), with novel results concern- revealed several miR targets that were predicted by ing miR169 described in this work. 1550 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs the nla mutant. The link between P limitation and NLA expressing miR169 show enhanced leaf water loss and provided by miR827 (Fig. 6A) suggests that NLA are more sensitive to drought stress, whereas NFYA5 activity is actively down-regulated during P limita- overexpressers show the opposite phenotypes (Li tion. This could indicate that plants select one or the et al., 2008). In addition to the effects of the nitrate other input signal depending on nutrient conditions transporter CHL1 (Guo et al., 2003) or nitrate reductase- and, therefore, the existence of hierarchies in the mediated nitric oxide generation (Desikan et al., 2002) interplay of macronutrient regulatory networks. on stomatal opening, low expression of miR169 during MiR2111 is predicted to target the E3 ligase gene N limitation could thus contribute to drought tolerance At3g27150 (Table I). At3g27150 displays strictly root- of N-limited plants (Lodeiro et al., 2000; Castaings et al., specific expression in large-scale transcriptome data 2008). sets like AtGenExpress (Schmid et al., 2005), suggest- In legume species, nodule development is depen- ing that it functions in the root. This is interesting in dent on the presence of previously established nodules the context of the high abundance of miR2111 in and N/nitrate availability, creating a root-to-shoot phloem sap during Pi limitation (Fig. 5), suggesting signal that activates the CLAVATA1-like receptor ki- another systemic regulatory circuitry, analogous to the nase SUNN in M. truncatula or HAR1 in Lotus japoni- miR399-PHO2 paradigm (Pant et al., 2008). cus. A recent report suggests that a nitrate-induced Regulation of chromatin status appears to be an- CLAVATA3/ESR-related (CLE) peptide is this root-to- other biological process influenced by P limitation- shoot signal (Fig. 6B; Okamoto et al., 2009). HAR1 induced miRs, as suggested by the best predicted exerts negative shoot control of root nodulation target genes of miR778, miR2111b*, and miR2111a*, (Krusell et al., 2002; Nishimura et al., 2002) through a namely SUVH6 (At2g22740), encoding a SET domain- shoot-to-root signal that might include auxin transport containing histone methyltransferase; At2g23380/ (van Noorden et al., 2006). It is also known that the CURLY LEAF, a SET domain gene required for his- miR169 target gene HAP2-1 in M. truncatula is a key tone methylation and genetic imprinting (Schubert regulator for the differentiation of nodule primordia et al., 2006); and At2g28290/SPLAYED, encoding a (Combier et al., 2006; Fig. 6B). MiR169 overexpression chromatin-remodeling complex subunit required for or knockdown of HAP2-1 leads to a developmental maintenance and identity of the shoot apical meristem block of nodule formation (Combier et al., 2006). (Kwon et al., 2005). Repression of miR169s by N limitation, as detected MiR398a is strongly reduced in P, N, and C limita- in our experiments, points toward a potential mecha- tion (Fig. 3), indicating a more general response to nistic link between low N status and nodule develop- nutrient stress. Repression of miR398a by C limitation ment in legumes. High abundance of miR169 in phloem also correlates with its induction by Suc (Dugas and sap during N-replete growth and the sharp decrease Bartel, 2008). Targets of miR398a include CSD1 and during N and P limitation (Fig. 5) also flags miR169 as a CSD2 (see introduction). CSD is required for detoxifi- potential long-distance signal (Fig. 6B) that is able to cation of reactive oxygen species that increase during report shoot N and P status to the roots, similar to the nutrient limitations and other environmental stresses role of miR399 (Pant et al., 2008). It will be interesting to (e.g. heat or drought; Apel and Hirt, 2004; Shin et al., test whether miR169 abundance is increased by nitrate/ 2005; Sunkar et al., 2006). Therefore, down-regulation N in legumes and whether miR169 expression and/or of miR398a leading to higher CSD activity would be an phloem abundance is dependent on SUNN/HAR1 to appropriate response to nutrient stress. establish if this is a novel shoot-to-root signal for the control of nodule differentiation. Potential Biological Impact of miR169 Regulation by CONCLUSION N Availability MiRs are emerging as increasingly interesting (sys- The targets of miR169s are several HAP2 trans- temic) regulators during mineral nutrient stress in cription factors (i.e. nuclear factor YA subunits plants. The discovery of new nutrient-dependent miRs [NF-YA]; Table I; Combier et al., 2006; Fahlgren et al., opens up the possibility of testing their roles and those 2007; Li et al., 2008). Transcripts of several of these of their predicted targets during adaptation of plants genes, including At3g05690/NF-YA2, At1g54160/ to nutrient deficiency. The qRT-PCR platform de- NF-YA5, At3g14020/NF-YA6, At1g72830/NF-YA8,and scribed here serves as a useful initial approach to test At5g06510/NF-YA10, increase during N and P limita- the response of annotated miRs in a given biological tion (Supplemental Fig. S10; Scheible et al., 2004; qRT- scenario, providing opportunities to discover new PCR confirmation not shown), thereby showing the signaling and regulatory networks. opposite response compared with miR169. In Arabidopsis, miR169 was reported to influence drought resistance via inhibition of the A5 subunit of MATERIALS AND METHODS NF-Y, a ubiquitous transcription factor that is highly Plant Materials expressed in guard cells and crucial for the expression of a number of drought stress-responsive genes (Li Nine-day-old nutrient-replete and N-, P-, or C-limited wild-type Arabi- et al., 2008). Nfya5 knockout mutants and plants over- dopsis (Arabidopsis thaliana Col-0) seedlings were grown in sterile liquid Plant Physiol. Vol. 150, 2009 1551 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. cultures as described previously (Scheible et al., 2004; Morcuende et al., 2007; 300 mL of EBR buffer (50 mM Mg acetate, 0.5 M ammonium acetate, 1 mM Osuna et al., 2007). The physiological status of the plant materials was EDTA, and 0.1% SDS) for 10 to 16 h at 20Cto25C (300 rpm). After phenol/ confirmed by expression analysis of marker genes (Supplemental Fig. S5) chloroform and chloroform extraction, the aqueous phase was mixed with prior to qRT-PCR analysis. Rapeseed (Brassica napus ‘Drakkar’) was germi- 1 mL of glycogen and 900 mL of 96% (v/v) ethanol, then cooled to 220C for nated and grown hydroponically (Buhtz et al., 2008) in a full-nutrient solution 2 h and centrifuged (25 min, 16,000g,4C). The RNA pellet was washed twice containing 4 mM N and 0.5 mM P. After 47 d, plants were divided into three with 75% (v/v) ethanol and dissolved in 6 mL of water. sets, and one set was supplied with lower N (2 mM KNO ) nutrient solution, 5# and 3# RNA adaptor ligations with RNA primers, RT, and PCR were one with low P (0.1 M KPi) nutrient solution, and one with full-nutrient performed according to Lu et al. (2007), except for a 3# RNA adaptor 3# end solution. From day 61 onward (i.e. approximately 1 week before flowering modification consisting of a C3 hydrocarbon spacer (Biomers.net). The PCR started), low-N, low-P, and FN plants were supplied with nutrient solutions (25 mL) was terminated with 75 mL of stop buffer (10 mM Tris-HCl, pH 8.0, containing no N, no P, or full-nutrient solution, respectively. N- or P-limited 1mM EDTA, and 0.4 M ammonium acetate). After phenol (pH = 8.0)/ plants developed clear signs of N or P starvation (e.g. reduced leaf biomass, chloroform extraction, 1 mL of glycogen (Roche; 20 mg mL ) and 300 mLof earlier flowering, reduced chlorophyll in 2N leaves; data not shown). Phloem ethanol were added to precipitate cDNA. The cDNA was denatured in loading sap was sampled between days 72 and 82 as described previously and was not buffer II (Ambion) and separated on an 8% polyacrylamide/7 M urea gel. The significantly contaminated by cell sap from other tissues (Giavalisco et al., cDNA band was eluted and precipitated as above. The pellet was washed 2006; Buhtz et al., 2008). with 70% (v/v) ethanol, air dried, and dissolved in 14 mL of water. The cDNA concentration was measured using a NanoDrop ND-1000 and checked by 15% polyacrylamide/7 M urea gel electrophoresis with oligonucleotides of known Primer Design and qRT-PCR Analysis concentration. Quality control was performed by TOPO cloning and Sanger sequencing of several plasmid clones (Lu et al., 2007). Illumina-Solexa A first set of pri-miR primers (pri-miR156 through -404) was designed by sequencing was performed at GATC Biotech for Arabidopsis and at FASTERIS Eurogentec. Primers for pri-miR405 through -870, and primers that replaced for rapeseed libraries. malfunctioning primers from the first set were designed using Primer Express 2.0.0 (Applied Biosystems) and Oligo 6.71 (Molecular Biology Insights). To Analysis of Deep Sequencing Results ensure maximum specificity and efficiency during PCR amplification of pri- miR cDNA under a standard set of reaction conditions (Fig. 1A), a stringent set Sequencing reads of lengths between 15 and 32 nucleotides were used after of criteria was used for primer design. This included predicted melting trimming sequence adapters and low-complexity regions [e.g. poly(A)] and temperatures of 61C 6 2C, limited self-complementarity, and PCR amplicon after removing reads of low quality (containing n runs, where n . 12). The lengths of 50 to 150 bp. Secondary hits were minimized by aligning primer read sets from the different conditions were subsequently mapped onto the candidates to all known Arabidopsis transcript sequences via BLAST searches Arabidopsis genome (TAIR8 assembly) using RazerS software. RazerS is an and eliminating primer pairs with more than the specific hit. Stem-loop efficient and generic read-mapping tool allowing the user to align reads of sequences for which no satisfactory primers could be found were elongated by arbitrary length using either the Hamming distance or the edit distance. 100 bp of flanking genomic sequence on each side before primer design was RazerS is part of the generic sequence analysis library Seqan (Doring et al., reinitiated. Annealing sites of the primers on the pri-miR sequence are 2008). Only perfect matches to the genome (i.e. full-length alignments with highlighted in Supplemental File S1. Sequences of the qRT-PCR primers are 100% identity) were retained. To investigate the read distributions for avail- given in Supplemental File S2. Cartridge-purified primers were purchased able TAIR8 annotations of genes and transposable elements (available from from Eurogentec, mixed with the corresponding forward or reverse primer ftp://ftp.arabidopsis.org/home/tair/Genes/TAIR8_genome_release) and to upon arrival to a final concentration of 50 mM each, arrayed on 96-deep-well find statistically significant changes in read distributions, we used the x test plates, and frozen at 280C for long-term storage. Working stocks (0.5 mM)of together with Benjamini-Hochberg P value correction. The x test is known to each primer pair were prepared from the storage stocks in two serial 10-fold have a good predictive power and robustness for gene expression analysis dilution steps and kept at 220C for short-term storage and used within 2 (Man et al., 2000). Normalization of small RNA data was performed by weeks. dividing the read number of each individual small RNA sequence by the RNA isolation, cDNA synthesis, and qRT-PCR analysis were carried out as number of redundant reads (15–32 nucleotides) in each library (Supplemental described previously by Czechowski et al. (2004, 2005) and Udvardi et al. Tables S2 and S3). (2008). Mature miR expression was analyzed using the method of Chen et al. (2005), as described (Pant et al., 2008). MiR-specific RT stem-loop primers are given in Supplemental Table S1. Primer sequences for marker genes are given MiR Prediction in the legend to Supplemental Figure S5. Potential miRs together with their precursor sequences were predicted using the miRDeep software tool (Friedla¨nder et al., 2008). The miRDeep Isolation of Small RNAs, Library Preparation, and algorithm was adjusted to plant precursor structures to take account of the Deep Sequencing following features of some plant miRs: (1) longer pre-miRs; (2) pre-miRs that contain more than one miR sequence; (3) the more diverse read distribution of Total RNA was isolated with Trizol reagent (Invitrogen) supplemented sequenced small RNAs on plant pre-miR sequences; and (4) nonhairpin pre- with 0.5% (w/v) N-lauroylsarcosine sodium salt, 3 mM b-mercaptoethanol, miR structures (P. May, unpublished data). and 5 mM EDTA. After phase separation, one phenol/chloroform and two chloroform extractions were performed. The aqueous phase (500 mL) was Target Gene Predictions mixed with 3 mL of glycogen (Roche; 20 mg mL ) before RNA was precip- itated with 625 mL of ethanol and 250 mLof0.8 M sodium citrate/1.2 M sodium Candidate miR target genes were determined using publicly available chloride. Samples were incubated for 30 min at room temperature and then prediction algorithms, including miRU (Zhang, 2005), the target search in centrifuged (25 min, 16,000g,4C). The precipitate was washed with 80% (v/v) WMD2 (Ossowski et al., 2008), the prediction tool in the UEA plant sRNA ethanol, air dried, and dissolved in 2 mM Tris-HCl (pH 7.5). Efficiency of small toolkit (Moxon et al., 2008b), and PITA (Kertesz et al., 2007). The programs RNA extraction and total RNA quality was checked by northern-blot hybrid- were used with their default settings. ization with a P-labeled oligonucleotide complementary to miR399 (Bari et al., 2006). RNA concentration was measured with a NanoDrop ND-1000 MiRBase accession numbers for all annotated Arabidopsis miRs (NanoDrop Technologies), and integrity was measured with an Agilent-2100 are available at http://microrna.sanger.ac.uk/cgi-bin/sequences/mirna_ Bioanalyzer (Agilent Technologies; RNA 6000 NanoChips). summary.pl?org=ath. GenBank accession numbers for the novel miR2111 Total RNA from three independent biological replicates (3 3 20 mg) was sequences described in this work are FN391952 (Ath-miR2111b), FN391950 mixed with 23 loading buffer II (Ambion), denatured for 2 min at 90C, and (Ath-miR2111a), FN391951 (Bna-miR2111b), and FN391953 (Bna-miR2111a). separated on a 15% polyacrylamide/7 M urea/13 TBE gel at 300 V. Synthetic, phosphorylated 18-mer and 24-mer RNA markers (Biomers.net) and a 10-bp DNA ladder (Invitrogen) were used to localize small RNAs (18–30 nucleo- Supplemental Data tides) as well as ligation and PCR products on gels stained with SYBR Gold (Invitrogen). RNA and PCR products were eluted from polyacrylamide gels in The following materials are available in the online version of this article. 1552 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Supplemental Figure S1. Genome arrangement and sequence similarity of organidentitybyamicroRNA and its APETALA2-liketargetgenes. miR169i to -n precursors. Plant Cell 15: 2730–2741 Aung K, Lin SI, Wu CC, Huang YT, Su CL, Chiou TJ (2006) Pho2, a Supplemental Figure S2. Strong induction of mir395 primary transcripts phosphate overaccumulator, is caused by a nonsense mutation in a during sulfur limitation. microRNA399 target gene. Plant Physiol 141: 1000–1011 Axtell MJ, Bartel DP (2005) Antiquity of microRNAs and their targets in Supplemental Figure S3. Melting curves of pri-miR amplicons land plants. Plant Cell 17: 1658–1673 Supplemental Figure S4. Sequencing results of four pri-miR amplicons. Bari R, Pant BD, Stitt M, Scheible WR (2006) PHO2, microRNA399, and PHR1 define a phosphate-signaling pathway in plants. Plant Physiol Supplemental Figure S5. Marker gene expression in nutrient-limited 141: 988–999 Arabidopsis seedlings. Bartel DP (2004) MicroRNAs: genomics, biogenesis, mechanism, and Supplemental Figure S6. Strong overexpression of miR399d mimics function. Cell 116: 281–297 molecular phenotypes of pho2 mutants. Bracht J, Hunter S, Eachus R, Weeks P, Pasquinelli AE (2004) Trans- splicing and polyadenylation of let-7 microRNA primary transcripts. Supplemental Figure S7. Number and length distribution of small RNA RNA 10: 1586–1594 sequences. Brodersen P, Sakvarelidze-Achard L, Bruun-Rasmussen M, Dunoyer P, Supplemental Figure S8. qRT-PCR verification of P limitation-induced Yamamoto YY, Sieburth L, Voinnet O (2008) Widespread translational small RNA species. inhibition by plant miRNAs and siRNAs. Science 320: 1185–1190 Brodersen P, Voinnet O (2009) Revisiting the principles of microRNA Supplemental Figure S9. miR2111 precursors from rapeseed. target recognition and mode of action. Nat Rev Mol Cell Biol 10: 141–148 Supplemental Figure S10. Nutrient-dependent expression of HAP2 genes Buhtz A, Springer F, Chappell L, Baulcombe DC, Kehr J (2008) Identifi- in Arabidopsis. cation and characterization of small RNAs from the phloem of Brassica napus. Plant J 53: 739–749 Supplemental Table S1. Primers used for RT of mature miR and qPCR Cai X, Hagedorn CH, Cullen BR (2004) Human microRNAs are processed quantification. from capped, polyadenylated transcripts that can also function as Supplemental Table S2. Read numbers from small RNA sequencing of mRNAs. RNA 10: 1957–1966 Arabidopsis libraries. Castaings L, Camargo A, Pocholle D, Gaudon V, Texier Y, Boutet-Mercey S, Taconnat L, Renou JP, Daniel-Vedele F, Fernandez E, et al (2008) The Supplemental Table S3. Read numbers from small RNA sequencing of nodule inception-like protein 7 modulates nitrate sensing and metab- rapeseed phloem sap. olism in Arabidopsis. Plant J 57: 426–435 Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT, Barbisin Supplemental File S1. Arabidopsis miR precursor sequences and primer M, Xu NL, Mahuvakar VR, Andersen MR, et al (2005) Real-time annealing sites. quantification of microRNAs by stem-loop RT-PCR. Nucleic Acids Res Supplemental File S2. Quantitative real-time PCR results for all investi- 33: e179 gated pri-miR species. Chiou TJ (2007) The role of microRNAs in sensing nutrient stress. Plant Cell Environ 30: 323–332 Supplemental File S3. Structures and small RNA reads of P-responsive Chiou TJ, Aung K, Lin SI, Wu CC, Chiang SF, Su CL (2006) Regulation of miRs. phosphate homeostasis by microRNA in Arabidopsis. Plant Cell 18: Supplemental File S4. Target gene predictions. 412–421 Chuck G, Candela H, Hake S (2009) Big impacts by small RNAs in plant development. Curr Opin Plant Biol 12: 81–86 Note Added in Proof Combier JP, Frugier F, de Billy F, Boualem A, El-Yahyaoui F, Moreau S, Vernie´ T, Ott T, Gamas P, Crespi M, et al (2006) MtHAP2-1 is a key The novel miR described in this work was independently reported by transcriptional regulator of symbiotic nodule development regulated by Fahlgren et al. (Fahlgren N, Sullivan CM, Kasschau KD, Chapman EJ, microRNA169 in Medicago truncatula. Genes Dev 20: 3084–3088 Cumbie JS, Montgomery TA, Gilbert SD, Dasenko M, Backman TW, Givan Czechowski T, Bari RP, Stitt M, Scheible WR, Udvardi MK (2004) Real- SA, et al [2009] Computational and analytical framework for small RNA time RT-PCR profiling of over 1400 Arabidopsis transcription factors: profiling by high-throughput sequencing. RNA 15: 992–1002). To unify the unprecedented sensitivity reveals novel root- and shoot-specific genes. naming, this miR is referred to as miR2111 in the final published version. Plant J 38: 366–379 Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible WR (2005) Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol 139: 5–17 Desikan R, Griffiths R, Hancock J, Neill S (2002) A new role for an old enzyme: nitrate reductase-mediated nitric oxide generation is required ACKNOWLEDGMENTS for abscisic acid-induced stomatal closure in Arabidopsis thaliana. Proc We thank Dr. John Lunn (Max-Planck Institute of Molecular Plant Phys- Natl Acad Sci USA 99: 16314–16318 iology) and Dr. Peter Do¨rner (University of Edinburgh) for proofreading and De Smet I, Vassileva V, De Rybel B, Levesque MP, Grunewald W, Van suggestions on the manuscript. Damme D, van Noorden G, Naudts M, Van Isterdael G, De Clercq R, et al (2008) Receptor-like kinase ACR4 restricts formative cell divisions Received March 27, 2009; accepted May 18, 2009; published May 22, 2009. in the Arabidopsis root. Science 322: 594–597 Doring A, Weese D, Rausch T, Reinert K (2008) SeqAn an efficient, generic C++ library for sequence analysis. BMC Bioinformatics 9: 11 Dugas DV, Bartel B (2004) MicroRNA regulation of gene expression in LITERATURE CITED plants. Curr Opin Plant Biol 7: 512–520 Dugas DV, Bartel B (2008) Sucrose induction of Arabidopsis miR398 Addo-Quaye C, Eshoo TW, Bartel DP, Axtell MJ (2008) Endogenous represses two Cu/Zn superoxide dismutases. Plant Mol Biol 67: 403–417 siRNA and miRNA targets identified by sequencing of the Arabidopsis Emery JF, Floyd SK, Alvarez J, Eshed Y, Hawker NP, Izhaki A, Baum SF, degradome. Curr Biol 18: 758–762 Bowman JL (2003) Radial patterning of Arabidopsis shoots by class III Allen E, Xie Z, Gustafson AM, Carrington JC (2005) MicroRNA-directed HD-ZIP and KANADI genes. Curr Biol 13: 1768–1774 phasing during trans-acting siRNA biogenesis in plants. Cell 121: Fahlgren N, Howell MD, Kasschau KD, Chapman EJ, Sullivan CM, 207–221 Cumbie JS,Givan SA,Law TF,Grant SR,Dangl JL,etal (2007) High- Apel K, Hirt H (2004) Reactive oxygen species: metabolism, oxidative stress, and signal transduction. Annu Rev Plant Biol 55: 373–399 throughput sequencing of Arabidopsis microRNAs: evidence for fre- Aukerman MJ, Sakai H (2003) Regulation of flowering time and floral quent birth and death of MIRNA genes. PLoS ONE 2: e219 Plant Physiol. Vol. 150, 2009 1553 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Pant et al. Franco-Zorrilla JM, Valli A, Todesco M, Mateos I, Puga MI, Rubio- Man MZ, Wang X, Wang Y (2000) POWER_SAGE: comparing statistical Somoza I, Leyva A, Weigel D, Garcı´a JA, Paz-Ares J (2007) Target tests for SAGE experiments. Bioinformatics 16: 953–959 mimicry provides a new mechanism for regulation of microRNA activ- Meyers BC, Axtell MJ, Bartel B, Bartel DP, Baulcombe D, Bowman JL, Cao ity. Nat Genet 39: 1033–1037 X, Carrington JC,ChenX,Green PJ,etal (2008) Criteria for annotation Friedla¨nder MR, Chen W, Adamidi C, Maaskola J, Einspanier R, Knespel of plant microRNAs. Plant Cell 20: 3186–3190 Mi S, CaiT, Hu Y,ChenY,HodgesE, NiF,Wu L, LiS,Zhou H, LongC, S, Rajewsky N (2008) Discovering microRNAs from deep sequencing data using miRDeep. Nat Biotechnol 26: 407–415 et al (2008) Sorting of small RNAs into Arabidopsis argonaute com- FujiiH,Chiou TJ,Lin SI,AungK, Zhu JK (2005) A miRNA involved in plexes is directed by the 5# terminal nucleotide. Cell 133: 116–127 phosphate-starvation response in Arabidopsis. Curr Biol 15: 2038–2042 Morcuende R,BariR,Gibon Y, ZhengW,Pant BD, Bla¨sing O, Usadel B, German MA, Pillay M, Jeong DH, Hetawal A, Luo S, Janardhanan P, Czechowski T, UdvardiMK, StittM,etal (2007) Genome-wide Kannan V, Rymarquis LA, Nobuta K, German R, et al (2008) Global reprogramming of metabolism and regulatory networks of Arabidopsis identification of microRNA-target RNA pairs by parallel analysis of in response to phosphorus. Plant Cell Environ 30: 85–112 Moxon S, Jing R, Szittya G, Schwach F, Rusholme P, Pilcher RL, Moulton RNA ends. Nat Biotechnol 26: 941–946 Giavalisco P, Kapitza K, Kolasa A, Buhtz A, Kehr J (2006) Towards the V, Dalmay T (2008a) Deep sequencing of tomato short RNAs identifies proteome of Brassica napus phloem sap. Proteomics 6: 896–909 microRNAs targeting genes involved in fruit ripening. Genome Res 18: Gifford ML, Dean A, Gutierrez RA, Coruzzi GM, Birnbaum KD (2008) 1602–1609 Cell-specific nitrogen responses mediate developmental plasticity. Proc Moxon S, Schwach F, Dalmay T, Maclean D, Studholme DJ, Moulton V Natl Acad Sci USA 105: 803–808 (2008b) A toolkit for analysing large-scale plant small RNA datasets. Guo FQ, Young J, Crawford NM (2003) The nitrate transporter AtNRT1.1 Bioinformatics 24: 2252–2253 (CHL1) functions in stomatal opening and contributes to drought Nishimura R, Hayashi M, Wu GJ, Kouchi H, Imaizumi-Anraku H, Murakami Y, Kawasaki S, Akao S, Ohmori M, Nagasawa M, et al susceptibility in Arabidopsis.Plant Cell 15: 107–117 Hofacker IL (2007) How microRNAs choose their targets. Nat Genet 39: (2002) HAR1 mediates systemic regulation of symbiotic organ develop- 1191–1192 ment. Nature 420: 426–429 Jones-Rhoades MW, Bartel DP (2004) Computational identification of Okamoto S, Ohnishi E, Sato S, Takahashi H, Nakazono M, Tabata S, plant microRNAs and their targets, including a stress-induced miRNA. Kawaguchi M (2009) Nod factor/nitrate-induced CLE genes that drive Mol Cell 14: 787–799 HAR1-mediated systemic regulation of nodulation. Plant Cell Physiol Kawashima CG, Yoshimoto N, Maruyama-Nakashita A, Tsuchiya YN, 50: 67–77 Saito K, Takahashi H, Dalmay T (2008) Sulphur starvation induces the Okamura K, Phillips MD, Tyler DM, Duan H, Chou YT, Lai EC (2008) The expression of microRNA-395 and one of its target genes but in different regulatory activity of microRNA* species has substantial influence on cell types. Plant J 57: 313–321 microRNA and 3# UTR evolution. Nat Struct Mol Biol 15: 354–363 Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E (2007) The role of site Ossowski S, Schwab R, Weigel D (2008) Gene silencing in plants using accessibility in microRNA target recognition. Nat Genet 39: 1278–1284 artificial microRNAs and other small RNAs. Plant J 53: 674–690 Kidner CA, Martienssen RA (2005) The developmental role of microRNA Osuna D, Usadel B, Morcuende R, Gibon Y, Bla¨sing OE, Ho¨hne M, in plants. Curr Opin Plant Biol 8: 38–44 Gu¨ nter M, Kamlage B, Trethewey R, Scheible WR, et al (2007) Tem- Krusell L, Madsen LH, Sato S, Aubert G, Genua A, Szczyglowski K, Duc poral responses of transcripts, enzyme activities and metabolites after G, Kaneko T, Tabata S, de Bruijn F, et al (2002) Shoot control of root adding sucrose to carbon-deprived Arabidopsis seedlings. Plant J 49: development and nodulation is mediated by a receptor-like kinase. 463–491 Nature 420: 422–426 Palatnik JF, Allen E, Wu X, Schommer C, Schwab R, Carrington JC, Kurihara Y, Watanabe Y (2004) Arabidopsis micro-RNA biogenesis Weigel D (2003) Control of leaf morphogenesis by microRNAs. Nature through Dicer-like 1 protein functions. Proc Natl Acad Sci USA 101: 425: 257–263 12753–12758 Pant BD, Buhtz A, Kehr J, Scheible WR (2008) MicroRNA399 is a long- Kutter C, Scho¨b H, Stadler M, Meins F Jr, Si-Ammour A (2007) Micro- distance signal for the regulation of plant phosphate homeostasis. Plant RNA-mediated regulation of stomatal development in Arabidopsis.Plant J 53: 731–738 Cell 19: 2417–2429 Peng M, Bi YM, Zhu T, Rothstein SJ (2007a) Genome-wide analysis of Kwon CS,Chen C,WagnerD (2005) WUSCHEL is a primary target for Arabidopsis responsive transcriptome to nitrogen limitation and its transcriptional regulation by SPLAYED in dynamic control of stem cell regulation by the ubiquitin ligase gene NLA. Plant Mol Biol 65: fate in Arabidopsis. Genes Dev 19: 992–1003 775–797 Lai EC (2004) Predicting and validating microRNA targets. Genome Biol Peng M, Hannam C, Gu H, Bi YM, Rothstein SJ (2007b) A mutation in 5: 115 NLA, which encodes a RING-type ubiquitin ligase, disrupts the adapt- LeeY,Kim M, HanJ,Yeom KH, LeeS,Baek SH,Kim VN (2004) ability of Arabidopsis to nitrogen limitation. Plant J 50: 320–337 MicroRNA genes are transcribedbyRNA polymerase II.EMBOJ 23: Peng M, Hudson D, Schofield A, Tsao R, Yang R, Gu H, Bi YM, Rothstein 4051–4060 SJ (2008) Adaptation of Arabidopsis to nitrogen limitation involves Li WX,OonoY,Zhu J, He XJ, Wu JM, Iida K, Lu XY,Cui X, Jin H,Zhu JK induction of anthocyanin synthesis which is controlled by the NLA (2008) The Arabidopsis NFYA5 transcription factor is regulated tran- gene. J Exp Bot 59: 2933–2944 scriptionally and posttranscriptionally to promote drought resistance. Ramakers C, Ruijter JM, Deprez RH, Moorman AF (2003) Assumption- Plant Cell 20: 2238–2251 free analysis of quantitative real-time polymerase chain reaction (PCR) Lin SI, Chiang SF, Lin WY, Chen JW, Tseng CY, Wu PC, Chiou TJ (2008) data. Neurosci Lett 339: 62–66 Rhoades MW, Reinhart BJ, Lim LP, Burge CB, Bartel B, Bartel DP (2002) Regulatory network of microRNA399 and PHO2 by systemic signaling. Plant Physiol 147: 732–746 Prediction of plant microRNA targets. Cell 110: 513–520 LindowM,JacobsenA,Nygaard S,Mang Y,Krogh A (2007) Intragenomic Scheible WR, Morcuende R, Czechowski T, Fritz C, Osuna D, Palacios- matching reveals a huge potential for miRNA-mediated regulation in Rojas N, Schindelasch D, Thimm O, Udvardi MK, Stitt M (2004) plants. PLoS Comput Biol 3: e238 Genome-wide reprogramming of primary and secondary metabolism, Lindow M, Krogh A (2005) Computational evidence for hundreds of non- protein synthesis, cellular growth processes, and the regulatory infra- conserved plant microRNAs. BMC Genomics 13: 119 structure of Arabidopsis in response to nitrogen. Plant Physiol 136: Liu HH, Tian X, Li YJ, Wu CA, Zheng CC (2008) Microarray-based analysis 2483–2499 of stress-regulated microRNAs in Arabidopsis thaliana. RNA 14: Schmid M, Davison TS, Henz SR, Pape UJ, Demar M, Vingron M, 836–843 Scho¨lkopf B, Weigel D, Lohmann JU (2005) A gene expression map of ´ ´ ´ Lodeiro AR, Gonzalez P, Hernandez A, Balague LJ, Favelukes G (2000) Arabidopsis thaliana development. Nat Genet 37: 501–506 Comparison of drought tolerance in nitrogen-fixing and inorganic Schmittgen TD, Jiang J, Liu Q, Yang L (2004) A high-throughput method nitrogen-grown common beans. Plant Sci 154: 31–41 to monitor the expression of microRNA precursors. Nucleic Acids Res Lu C, Meyers BC, Green PJ (2007) Construction of small RNA cDNA 32: e43 libraries for deep sequencing. Methods 43: 110–117 Schubert D, Primavesi L, Bishopp A, Roberts G, Doonan J, Jenuwein T, 1554 Plant Physiol. Vol. 150, 2009 Downloaded from https://academic.oup.com/plphys/article/150/3/1541/6108002 by DeepDyve user on 13 July 2021 Novel P- and N-Responsive MicroRNAs Goodrich J (2006) Silencing by plant Polycomb-group genes requires Mueller-Roeber B, Jarmolowski A, Szweykowska-Kulinska Z (2009) dispersed trimethylation of histone H3 at lysine 27. EMBO J 25: 4638–4649 Gene structures and processing of Arabidopsis thaliana HYL1- Selbach M, Schwanha¨user B, Thierfelder N, Fang Z, Khanin R, Rajewsky dependent pri-miRNAs. Nucleic Acids Res 9: 3083–3093 N (2008) Widespread changes in protein synthesis by microRNAs. Szittya G, Moxon S, Santos DM, Jing R, Fevereiro MP, Moulton V, Nature 455: 58–63 Dalmay T (2008) High-throughput sequencing of Medicago truncatula Sharma VK, Ramirez J, Fletcher JC (2003) The Arabidopsis CLV3-like short RNAs identifies eight new miRNA families. BMC Genomics (CLE) genes are expressed in diverse tissues and encode secreted 9: 593 proteins. Plant Mol Biol 51: 415–425 Udvardi MK, Czechowski T, Scheible WR (2008) Eleven golden rules of Shi R, Chiang VL (2005) Facile means for quantifying microRNA expres- quantitative RT-PCR. Plant Cell 20: 1736–1737 sion by real-time PCR. Biotechniques 39: 519–525 van Noorden GE, Ross JJ, Reid JB, Rolfe BG, Mathesius U (2006) Shin R, Berg RH, Schachtman DP (2005) Reactive oxygen species and root Defective long-distance auxin transport regulation in the Medicago hairs in Arabidopsis root response to nitrogen, phosphorus and potas- truncatula super numeric nodules mutant. Plant Physiol 140: 1494–1506 sium deficiency. Plant Cell Physiol 46: 1350–1357 Wu G, Poethig RS (2006) Temporal regulation of shoot development in Sunkar R, Chinnusamy V, Zhu J, Zhu JK (2007) Small RNAs as big players Arabidopsis thaliana by miR156 and its target SPL3. Development 133: in plant abiotic stress responses and nutrient deprivation. Trends Plant 3539–3547 Sci 12: 301–309 Yamasaki H, Abdel-Ghany SE, Cohu CM, Kobayashi Y, Shikanai T, Pilon Sunkar R, Kapoor A, Zhu JK (2006) Posttranscriptional induction of two M (2007) Regulation of copper homeostasis by micro-RNA in Arabi- Cu/Zn superoxide dismutase genes in Arabidopsis is mediated by dopsis. J Biol Chem 282: 16369–16378 downregulation of miR398 and important for oxidative stress tolerance. Yoo BC, Kragler F, Varkonyi-Gasic E, Haywood V, Archer-Evans S, Lee Plant Cell 18: 2051–2065 YM,Lough TJ,Lucas WJ (2004) A systemic small RNA signaling system Sunkar R, Zhu JK (2004) Novel and stress-regulated microRNAs and other in plants. Plant Cell 16: 1979–2000 small RNAs from Arabidopsis. Plant Cell 16: 2001–2019 Zhang Y (2005) miRU: an automated plant miRNA target prediction server. Szarzynska B, Sobkowiak L, Pant BD, Balazadeh S, Scheible W-R, Nucleic Acids Res 33: W701–W714 Plant Physiol. Vol. 150, 2009 1555

Journal

PLANT PHYSIOLOGYOxford University Press

Published: May 22, 2009

There are no references for this article.