Novel and conserved miRNAs in the halophyte Suaeda maritima identified by deep sequencing and computational predictions using the ESTs of two mangrove plants

The spatial and temporal regulation of gene expression in response to environmental cues are important factors for determining plant survival and adaptability leading to development of an ecotype [1]. Transcriptome studies have shown that plants challenged with an abiotic stress present altered expression levels of a number of genes involved in a broad spectrum of biochemical, cellular and physiological processes, including genes providing resistance to the abiotic stress [2–4]. Alterations to gene expression may occur because of regulatory mechanisms working at several levels. The two most well-known and studied categories of regulation are those involving transcription factors and small RNAs (sRNAs). While transcription factors work at the transcription level, sRNAs regulate gene expression at both the transcription and post-transcription levels. The importance of transcription factors in the regulation of gene expression and their associated mechanisms has been known since the 1980s [5–7] and can be illustrated by the presence of a massive number of transcription factor families and the extensive combinatorial control of gene expression by multiple transcription factors. However, the regulatory role of sRNAs in gene expression was only discovered two decades later after observing the reduced accumulation of gene products (mRNA) upon the introduction of dsRNA homologous to the gene into the tissue [8, 9]. As per our current understanding, sRNAs are typified by a large and growing class of ~22-nucleotide (nt)-long non-coding RNAs that function in association with Argonaute (Ago)-family proteins [10].

In plants, sRNAs exhibit an unexpected complexity and are classified based on their biogenesis and the structure of the genomic loci from which they are transcribed [11]. At present, sRNAs are distinguished as microRNAs (miRNAs) and three classes of endogenous small interfering RNAs (siRNAs), specifically trans-acting siRNA (ta-siRNA), heterochromatic siRNA (hc-siRNA) and natural antisense siRNA (nat-siRNA) [11, 12]. Among these, miRNA-guided post-transcriptional gene regulation constitutes one of the most conserved and well-characterized gene regulatory mechanisms, and it is important for development, stress responses and a myriad of other biological processes in eukaryotes [11, 13, 14]. The involvement of miRNAs increases the complexity of gene regulation processes because miRNAs generally have multiple targets, including transcription factors, and transcription factors also regulate the expression of pri-miRNA [15]. This complexity is further increased because miRNAs have been discovered to regulate gene expression at the transcriptional level [16]. In one of the first records of miRNA involvement in transcriptional gene silencing, Bao et al. [17] showed that mutations in the Arabidopsis PHABULOSA (PHB) and PHAVOLUTA (PHV) transcription factor genes, which affect the ath-miR165/166 complementary site in the processed mRNAs, presented decreased methylation of the gene downstream of the complementary site compared with that of the wild type. Similarly, Kim et al. [18] reported that the expression of POLR3D (DNA-directed RNA polymerase III subunit RPC4) was negatively regulated by miR320 mediated through histone methylation. However, Place et al. [19] observed the presence of a miR373 target site in the promoters of E-cadherin and cold-shock domain-containing protein C2 (CSDC2), and this miRNA induced the expression of both genes. In fact, miR373 represents the first evidence that miRNAs targeting promoters can enhance the synthesis of RNA or RNA activation (RNAa) in a manner similar to that shown by small activating RNAs (saRNA). Recently, Huang et al. [20] showed that three miRNAs (miR744, miR1186 and miR466d-3p) induced the expression of Ccnb1 (Cyclin B1) in mouse cell lines. In addition, miRNAs have also been reported to induce upregulation of translation of target mRNAs [21–23], making these molecules highly versatile components of gene regulation and function.

The regulatory roles of miRNAs in biotic and abiotic responses in plants have also been suggested [11, 24–27]. Sunkar and Zhu [28] were the first to demonstrate the upregulation of miR393, miR402, miR397b and miR319c by at least one of the following stresses: drought, cold and salt. Similarly, Katiyar-Agarwal et al. [29] described a novel class of bacteria-induced sRNAs in Arabidopsis thaliana. However, although further work on miRNA responses to biotic stress in plants has been conducted, particularly for pathogen-host specificity, assessments of such responses with regard to abiotic stress have mostly considered test species that are either intolerant or naturally not tolerant to a particular type of stress. To the best of our knowledge, the current understanding of miRNA expression in halophytes, the naturally salt tolerant plants, is limited to three species: Avicennia marina [30], Salicornia europaea [27] and Salicornia brachiata [31]. Deep sRNA sequencing has only been performed for the former two species, whereas high-throughput sRNA sequencing data are available for innumerable numbers of glycophytes.

The responsiveness of most miRNAs to abiotic stresses in glycophytes has been found to be species specific rather than stress specific although they might be involved in stress tolerance [13, 24, 26, 32–34]. It may be that the miRNA-mediated regulatory requirements for stress tolerance could vary with different species. Moreover, the majority of stress-related miRNA in miRNA databases were identified in model plant species, such as Arabidopsis lyrata, A. thaliana, Brachypodium distachyon, Glycine max, Medicago truncatula, Nicotiana tabacum, Oryza sativa, Physcomitrella patens, Populus trichocarpa, Solanum tuberosum, Sorghum bicolor, Vitis vinifera and Zea mays, and they constitute 5380 miRNAs out of the 8455 miRNAs from 73 plant species in miRBase 21.0 (http://www.mirbase.org) [35] and 11183 miRNAs out of the 15041 miRNAs from 150 plant species in the Plant Non-coding RNA Database (PNRD) (http://structuralbiology.cau.edu.cn/PNRD/) [36]. The miRNAs identified in the halophytes S. brachiata and S. europaea have yet to be included in miRBase, whereas those identified in A. marina are only partially included.

Because of ambiguities in the available information on the miRNAs involved in tolerance to salinity, which is a serious abiotic stress affecting crops worldwide, as well as the availability of limited information regarding salt-responsive miRNAs in naturally salt-tolerant plants, the present work was conducted to identify miRNAs and their targets and study their responses to salt application in S. maritima, a halophyte that grows naturally along the seashore and has been described as a well-suited plant for studying salt stress response because of its ability to grow in the presence and absence of salt [2]. The halophyte S. maritima has been exploited worldwide for the physiological and molecular characterization of salt tolerance in plants [37–39]. Recently, the possible involvement of antioxidative machinery in salt tolerance in S. maritima was also investigated [40, 41].

The current study was also designed to explore the possibility of identifying miRNAs in S. maritima using the EST database for two mangrove plants, Rhizophora mangle and Heritiera littoralis, to establish whether the transcriptome data of one species could be utilized to identify miRNAs in other species with confidence. In the present study, seven novel salt-responsive miRNAs and several conserved miRNAs were identified and validated experimentally in test plants. This study also investigated the response of these novel miRNAs to NaCl application and/or their presence in taxonomically and ecologically diverse plant species, including two rice cultivars (salt-tolerant O. sativa cv. Pokkali and non-tolerant O. sativa cv. Badami) and several plant species growing naturally under the direct influence of seawater, such as Sesuvium portulacastrum, Cyperus arenarius, Ipomoea pes-caprae and S. maritima.

The identification of 134 conserved miRNAs (Fig. 1, Additional file 4) from nearly 17 million total reads (Additional file 3) in the current study is comparable to the identification of 115 conserved miRNAs from nearly 15 million reads in Cunninghamia lanceolata [47] and 210 conserved miRNAs from nearly 13 million total reads in S. europaea [27]. Most of the miRNAs identified belonged to the glycophytes such as A. lyrata, A. thaliana, G. max, O. sativa, P. trichocarpa, Z. mays and additional species for which the miRNA database is relatively large compared to the other plant species, suggesting the need to create miRNA and EST databases containing as many diverse groups of plant species as possible. The family diversity of miRNAs (Fig. 1) was similar to those reported in the halophytes such as Avicennia marina (193 conserved miRNAs distributed to 36 families) and S. europaea (210 conserved miRNAs distributed to 51 families), with miRNA to miRNA family ratios of ~5.3 and ~4.1, respectively [27, 30]. In contrast, studies have shown much lower miRNA to miRNA family ratios in a range from 1.5 to 2.0 [52–56] for terrestrial plants other than halophytes. NaCl-induced changes observed in the miRNA abundance in several miRNA families, particularly miR172, miR408, miR399, miR166, miR165 and miR535 (Fig. 1), is indicative of an altered metabolism in the test plant in the presence of salt, which was also observed in other studies, including halophytes and non-halophytes [27, 57].

The expression of several of the miRNAs such as miR166b, miR166e, and miR166m that exhibited salt-induced decreases in abundance in S. maritima (Fig. 2a) revealed increases in abundance in rice in response to salt treatment [58]. The reverse was true for other miRNAs, such as miR156a, miR159a, miR159c, miR168a, miR169a, miR169b, miR171b, miR172c, miR319a and miR398a (Fig. 2a, b) [58]. However, the expression levels of at least nine out of the 12 miRNAs found common in S. maritima and S. europaea [27] were comparable, with five (miR166a, miR164a, miR165a, miR166e, miR165a) showing downregulation and four (miR168b, miR169a, miR171b, miR396b) showing upregulation in both of the species. These observations suggested that the response of miRNAs to environmental change might differ with the plant species and that the level of difference would depend on the difference in their metabolism.

None of the novel miRNAs identified in S. maritima was identical to those identified in the halophytes S. europaea [27], S. brachiata [31] and A. marina [30], suggesting that the database of plant miRNAs is not yet saturated and that there is a need for further miRNA identification in more halophytes. However, the unavailability of genome/transcriptome information often limits such attempts. To overcome this disadvantage, Khraiwesh et al. [30] used the Mangrove Transcriptome Database (MTDB - http://mangrove.illinois.edu), whereas Singh and Jha [31] relied on the genome databases of species other than S. brachiata at NCBI. In fact, identification of the precursors of as many as seven novel miRNAs in S. maritima out of the 13 novel miRNAs identified using the ESTs of R. mangle and H. littoralis that met all the criteria (Table 2, Additional files 4 and 5) revealed that the ESTs of one plant species can be used in the computational predictions of miRNAs in another species, even taxonomically separated species, if they were derived from similar habitats. However, metabolism may be more important than habitat for determining similarities in the miRNAomes of two species because while all seven miRNA precursors in S. maritima were also present in R. mangle and both species accumulated high levels of Na⁺ in their tissues [42, 59], only one of the precursors (sma-mir4) was present in H. littoralis, which accumulated low amounts of Na⁺ in its tissues [59].

Most of the novel miRNAs identified were low in abundance (Fig. 4) and expressed only in response to NaCl (sma-miR1, -miR3, -miR4 and -miR5), unlike the conserved miRNAs. Feng et al. [27] described the low-abundance novel miRNAs in S. europaea that were expressed only in the salt-treated plants as species-specific miRNAs that could be important determinants of salt tolerance in the plant.

The influence of NaCl on the expression levels of miRNAs considered for Northern analysis (Fig. 3) has been reported to vary with species [24, 32, 60], indicating a species-specific response of these miRNAs. However, because of the presence of these miRNAs in a wide taxonomic group of plant species and their responsiveness to a wide range of abiotic stresses (Table 1), their involvement in protecting plants against stress cannot be ruled out. Similarly, the salt responsiveness of the novel miRNAs, particularly of sma-miR2 and sma-miR7 showing downregulation in response to the NaCl treatment (Fig. 5), indicates that their metabolic functions are important for salt tolerance in the plant. A significant decrease in sma-miR7 expression in S. maritima showed no changes in ETR, whereas significant increases in the miRNA expression level in the rice cultivars showing decreases in ETR in response to NaCl (Fig. 6, Additional file 8) provide circumstantial evidence of involvement of sma-miR7 in salt tolerance.

Most of the targets of the conserved miRNAs were downregulated after NaCl treatment (Fig. 7), which is consistent with the increased expression of the miRNAs after NaCl treatment, except for that of sma-miR396b (Fig. 3). The decrease in the expression of at least a few of the targets might be important in the context of salt tolerance. For example, the decrease in the cellular level of lipid phosphate phosphatase (LPP) might lead to acceleration of the phosphorylation signaling [61], which, in turn, could function to accelerate biochemical processes linked to salt tolerance. Similarly, the decreases in the cellular levels of ubiquitin-conjugating enzyme (E2) and F-box proteins might function to slow the degradation process of the cellular proteins [62–64] associated with salt tolerance in the plant. Hence, the upregulation of these miRNAs in response to NaCl application could be of great ecological significance in terms of adaptations of S. maritima to NaCl.

Functional annotation of the targets of the novel miRNAs (Table 3), most of which were upregulated (Fig. 7), revealed that these targets and the associated miRNAs could be important for plant adaptations to adverse environments, a possibility that could be particularly true for sma-miR7, which was downregulated in S. maritima and upregulated in the rice cultivars (Fig. 6), and targeted ARF family proteins in S. maritima, R. mangle and H. littoralis (Table 3). Although the reports of involvement of auxins in salt resistance in plants are contradictory [65], their involvement is supported by the greater expression of sma-miR7 in the salt-sensitive Badami compared with that of the moderately salt tolerant Pokkali in response to NaCl application, along with the salt-induced downregulation of sma-miR2 targeting SEUSS-LIKE (SLK) protein in R. mangle, which is required for a proper response to the phytohormone auxin [66]. Moreover, the downregulation of seu-miR160 and seu-miR5 targeting ARF genes in S. europaea in response to NaCl application [27] further supports a possible role for auxin in salt tolerance in plants.

The possible involvement of sma-miR7 in salt resistance is further emphasized by the occurrence of another common target of the miRNA in S. maritima and R. mangle/H. littoralis, the vacuolar protein sorting-associated protein (Table 3), a constituent of ESCRTs (Endosomal Sorting Complexes Required for Transport), which are involved in mono-ubiquitination mediated endocytic trafficking of plasma membrane proteins to the vacuole followed by lysosomal degradation [67]. Moreover, this miRNA also targets the RING finger protein E3 ligase (Table 3), which serves as a unit of the ubiquitin/26S proteasome system. The biological role of the ubiquitin/26S proteasome system has been investigated in relation to the abiotic stress tolerance of plants and has been found to be a positive regulator of stress tolerance [63]. The suppression of the expression of the miRNA (seu-miR396h/I) that targets ubiquitin-specific proteases (UBPs) in S. europaea also strongly suggests a possible role for the ubiquitin/26S proteasome system in salt tolerance in plants [27]. Nevertheless, the activity of both plasma membrane proteins and cytosolic proteins, might also be influenced by regulating the abundance of active protein, which is modulated by the phosphorylation process as indicated by the downregulation of sma-miR5 in response to NaCl application (Fig. 5) and the presence of the target serine/threonine kinase family protein in S. maritima and both mangrove plants (Table 3). The importance of protein phosphorylation in salt tolerance has also been indicated by Feng et al. [27] in S. europaea.

The salt tolerance process in halophytes is mediated by many proteins in various ways [3, 27, 68]. Feng et al. [27] stressed the important roles of the proteins such as squamosa promoter-binding-like protein (SPL) and nuclear factor Y subunit A (NF-YA), among others, in the morphological adaptation of the halophyte S. europaea to salt stress, which also seems to hold true for S. maritima, because the miRNAs (sma-miR156b, -157d and -169g) targeting these transcripts (Additional file 11) were downregulated in response to NaCl (Fig. 2). Moreover, salt application downregulated the expression of miR-169g targeting NF-YA in the halophyte S. brachiata [31]. Feng et al. [27] further related salt tolerance in S. europaea to the increase in protein turnover that occurs as a result of the miRNA-regulated upregulation of the F-box protein. Salt application has also been reported to increase F-box protein transcripts in S. maritima and S. fruticosa [3, 68], and the increase in S. maritima could have been caused by the downregulation of certain miRNAs, such as sma-miR157d and sma-miR156b (Fig. 2), which target the mRNAs encoding F-box proteins (Additional file 11). These results suggest the important role of the ubiquitin/26S proteasome system in maintaining the required protein turnover demanded by the metabolic requirements in the presence of salt [27]. In addition, Sahu and Shaw et al. [3] indicated accumulation of glycinebetaine as an important component of salt tolerance in S. maritima, and the salt-induced downregulation of sma-miR169g (Fig. 2) supports the same because the miRNA targets phosphoethanolamine N-methyltransferase (Additional file 11), which is involved in glycinebetaine biosynthesis. Nevertheless, the overall salt tolerance process in halophytes could be under the control of regulatory kinases [27, 68], a possibility that is also supported by the current study in which NaCl induced the downregulation of sma-miR157d and -miR5 (Figs. 2 and 5), which target serine-threonine kinase (Additional file 11, Table 3).

Although the regulatory role of miRNAs cannot be generalized, the expression of sma-miR2 and sma-miR5 only in plants growing in saline environments, such as S. maritima, S. portulacastrum, C. arenarius and I. pes-caprae (Additional file 9), does suggest that these miRNAs might be regulating certain biochemical processes common in the plants that are important for their survival in the given saline ecological conditions. In this regard, it is important to note that sma-miR2 targets 4-coumarate-CoA ligase in S. maritima (Table 3), an enzyme that plays a pivotal role in the biosynthesis of plant secondary compounds with essential functions in plant development and environmental interactions [69].

This study identified 134 conserved and thirteen novel miRNAs in S. maritima, a halophyte. More importantly, the study demonstrated that the transcriptome of one species can be successfully used to computationally predict miRNAs in other species inhabiting similar environments, even if they are taxonomically distantly related. However, metabolism could be an even more important determining factor of similarity in miRNAomes of two species than the habitat. This study further revealed that a high ratio of known or conserved miRNAs to the number of families to which they belong could be considered a potential indicator of further miRNA identifications in a species, a result that is supported by the lack of identities among the novel miRNAs identified in the three halophytes, including S. brachiata [31], S. europaea [27] and S. maritima (this study). Furthermore, the responsiveness of many of the conserved miRNAs and most of the novel miRNAs in particular to salt suggested their important roles, in the metabolic adjustments to facilitate growth in a saline environment. However, any role for miRNAs in salt tolerance in halophytes may not be straightforward because several metabolic pathways could be acting either in parallel or in sequence during the process [3, 27, 68]. Of course, the miRNAs specifically expressed in plants growing under the influence of seawater, such as sma-miR2 and sma-miR5, could be important for the plants’ survival in a saline environment. Although miRNA-mediated molecular mechanisms for salt tolerance in plants have yet to be established, the target prediction of novel miRNAs provided an indication that ubiquitin-mediated alterations in the abundance of plasma membrane and cytosolic proteins and phosphorylation-mediated alterations of their active states could be important miRNA-regulated biochemical processes for salt tolerance in halophytes.