11 publications mentioning mdm-MIR167h

Open access articles that are associated with the species Malus domestica and mention the gene name MIR167h. Click the [+] symbols to view sentences that include the gene name, or the word cloud on the right for a summary.

1

Recent study using high osmotic stress in Arabidopsis has shown that 47 ath-miR167 also targets IAR3, an evolutionary conserved target, other than ARF6 and ARF8, which suggests that there might be additional target(/s) other than targets validated in natural/control condition.

Since both gma-miR167h/i (UmiR167-7 and UmiR167-8) have only 77% and 85% complementarity with target Gma-MCCC and Gma-LRPK respectively (Fig. 8E–F), we cannot rule out the translational inhibition of these targets, instead of predicted cleavage.

This suggests the miR167 mediated cleavage of novel Mdm-CNBL10 target mRNA and possible translational inhibition of novel Gma target mRNAs (as above).

This suggests that changes in the spatiotemporal expression of miRNA or predicted target genes under stress or treatment may lead to validation of additional miR167 mediated target cleavage and regulation of biological processes.

To understand conservation and diversification of miR167 target sites and the coevolution of miR167s and their target genes, we have done phylogenetic analysis with 27 UTSs along with total 44 selected target genes of 14 UmiR167s.

Therefore, the uniqueness and sequence variation of miR167 sequences (on the basis of similarity) as well as their complementary target sequences are very important to understand the coevolutionary pattern of miR167s, their corresponding targets, and miR167 mediated gene regulation.

Previous experimental studies on Arabidopsis miR167 37 have reported that over expression of only ath-miR167a (among four ath-miR167s) showed arrested flower development, similar to mutants of target arf6-2 and arf8-3 plants.

Although this gene is not regulated by miR167 in Arabidopsis, the Mdm-CNBL10 homolog is targeted by mdm-miR167a due to sequence diversification in both miRNA and target site.

As miR167 is a crucial family of plant miRNA implicated in multiple biological processes including gametophyte development, flower development and adventitious root development, we have attempted to trace back the evolutionary relationship of miR167 family members (as registered in miRBase database registry) and their target sequences among the land plants.

We have used the psRNATarget web tool 46 for the identification of the novel target of the unique miR167 (Table 2).

Phylogenetic tree was reconstructed using predicted target sequences and their unique miR167 binding sites to reveal the critical sequence variation in miRNA target sites.

Each of other UmiR167s (UmiR167-7/8/9/11/12/14) are shared by only 1 miR167 sequence and also targets 1 gene except the UmiR167-12 (targeting 2 genes).

A unique miR167 sequence may target transcripts of many genes, which may share the same miR167 complementary site (unique target sequence) despite their sequence variation at the whole gene level.

The largest number of miR167 sequences (54) from 21 different species has the UmiR167-1 sequence, which has 3 UTSs (Unique Target Sequences), and is predicted to target transcripts from total 19 genes (Table 2).

The complementary sequences of UmiR167 with their target genes, unique miR167 binding sites (on target mRNAs) were identified and numbered (Table 2).

This suggests that besides miR167 sequences, the common target ARFs have also undergone sequence diversification resulting in ARF6/8 that are not targeted by miR167s in some cases.

Predicted targets of unique miR167 sequences using psRNATarget web server tool.

These unique miR167 (UmiR167) sequences were used for the prediction of targets of miR167s from all species (analyzed here) using psRNATarget tool 46.

We have predicted that the unique miR167 sequences might target genes other than ARF6 and ARF8, which are the proved targets of miR167 in A. thaliana.

Based on complementarity of unique miR167 sequences with their target genes, unique miR167 binding sites (on target mRNAs) were identified and numbered (as 1, 2, 3 etc.

Variation in either of miR167 or its complementary target sequence may lead to functional diversity of miR167 mediated regulatory processes (Fig. 7).

The absence of homologs of japonica Osa-ARF6 and Osa-ARF (targets of japonica UmiR167-1) (Table 2) in indica cultivars suggest the functional diversification of miR167 regulated ARFs.

On the other hand, conserved target genes may undergo sequence variation in the miR167 binding sites.

It is well known that ARF6/8 and IAA-Ala Resistant3 (IAR3) are evolutionary conserved targets of miR167 in Arabidopsis 47.

Similarly, UmiR167-6, shared by 6 miR167 sequences from 4 species, is having only 1 UTS, and it targets 6 genes (Table 2).

Though 8 miR167 sequences from 6 species share UmiR167-3, it has only one UTS belonging to one target gene.

Validation of novel non-conserved target of miR167.

Identification of Unique miR167 and Unique Target genes.

ML Phylogeny of corresponding target sequences and miR167 using MEGA5 to show relatedness.

Further, a phylogenetic analysis was done to understand the evolutionary pattern and reveal critical sequence variation in miR167 target sites.

Therefore, coevolution of both miR167 and their respective target sequences played important role in the functional diversification among diverse species.

Computational identification provided an extensive list of potential miR167 targets for diverse plant species (Table 1).

5′RLM-RACE PCR based validation of novel miR167 targets.

The availability of well annotated complete genome sequences of diverse mo del land plants such as P. patens, O. sativa A. thaliana and Z. mays (as described in the materials & methods) have enabled the comparative genomics studies to explore the evolutionary relationship of the pre-MIR167 gene family and their targets across diverse plant species.

In course of evolution of miR167 sequences, the complementary sequence of target genes has also been subjected to evolutionary selection pressure.

Our analysis suggests that gma-miR167h targets METHYL-CROTONOYL-CoA CARBOXYLASE (MCCC), which is not homologous to any ARFs, as observed through BLAST, indicating the possible functional diversification of gma-miR167s in G. max (Table 2).

The UmiR167-5 is shared by 7 miR167 sequences from 5 species and has 6 UTSs belonging to 7 target genes.

Not only at mature and precursor level, but target genes of bra-miR167 are also conserved.

Interestingly, only 2 miR167 sequences share the UmiR167-4 from 2 species and is having 5 UTSs belonging to 6 target genes.

It was evident from our phylogenetic analysis that gma-miR167h/i and mdm-miR167a are clustered with 3′ derived miR167s rather than 5′ derived miR167s (Fig. 3) and predicted to cleave non-conserved target mRNAs.

Therefore, we suggest that in monocots (except O. sativa), miR167 mediated gene regulation is least affected during the course of evolution resulting into their functional conservation including auxin signaling, (Fig. 7).

This is likely to provide functional diversification of miR167 mediated gene regulation and stress response in apple.

This functional diversification is caused by mutation in critical region of mature miR167 sequences, as we have earlier shown for ppt-miR166m 44.

So it could be assumed that not all miR167* are degraded during the processing of precursor, rather they function as miR167-3p as evident from deep sequencing results in miRBase (http://www.

Therefore, a separate phylogenetic tree was reconstructed using reverse complementary sequences of miR167-3ps (miR167-3p-RC) along with other miR167s (Fig. 4).

Phylogenetic analysis of mature miR167 sequences.

Interestingly, we have observed that the gma-pre-MIR167h and gma-pre-MIR167i are conserved and fall in clade IV within the group II, which is similar to their mature gma-miR167h and gma-miR167i.

The branch length suggests that ath-miR167c and aly-miR167c-5p evolved faster with less substitution rate than acq-miR167 and cme-miR167e.

We retrieved one hundred and fifty three mature miR167 and their precursor MIR167 (pre-MIR167) sequences from thirty three diverse plant species including basal plants like moss (Physcomitrella patens), monocot (Oryza sativa, Zea mays etc. )

Interestingly, the miRNAs represented as miR167-3p formed another group.

The clade I of group II has further diverged through the duplication of sequences in the evolutionary process to a major sub-clade consisting of twenty five pre-MIR167 genes.

Interestingly, only ptc-miR167h-5p clustered with osa-miR167e/i-3p-RC, zma-miR167h/i-3p-RC and bdi-miR167d-3p-RC (Fig. 4).

Similarly, the diversity in sequences (within the mature miRNA part of some species such as gma-miR167h/i and mdm-miR167a) from their corresponding other mature miR167 family members are due to critical changes in the mature sequence in the course of evolution (Table 2).

The UmiR167-10 is shared by bna-miR167a/b and UmiR167-13 is shared by mdm-miR167h/i/j within their respective species B. napa and M. domestica only.

This tree is similar to Fig. 3, but contains reverse complementary sequences of miR167-3p (highlighted, encircled portion) along with three miR167-5ps namely, gma-miR167h/i and mdm-miR167a.

Interestingly, our phylogenetic analysis suggests that sequences of Brassica rapa is conserved at mature as well as precursor (miR167/MIR167) level and the event of duplication had occurred with a slow rate of substitution, since their precursor sequences are of almost equal length with maximum similarity.

The csi-pre-MIR167a, ctr-pre-MIR167 and ccl-pre-MIR167a/b also have a common ancestor.

Similarly, mdm-miR167a clustered with ptc-miR167f-3p, ptc-miR167g-3p and ptc-miR167h-3p (Fig. 3).

This is due to the less conservation of pre-MIR167 of sequences, as shown in the alignment statistics (Fig. 2).

We have observed huge variation in the length of mdm-pre-MIR167 sequences, which is another cause of their diversification at precursor level in addition to the variation in the present sequences.

Our analysis on evolutionary relationship among miR167 sequences shows that the mature miR167 family members, except gma-miR167i, gma-miR167h and mdm-miR167a which together produced a different group, are conserved and clustered in a single clade (Fig. 3).

Unlike its mature miR167s, which are mostly conserved, the pre-MIR167 sequences of A. thaliana (ath-pre-MIR167) and A. lyrata (aly-pre-MIR167) are distributed in four clusters in both group I and II, and have diverged more (Fig. 6).

Similarly ~0.25 fraction of the total sequences in MIR167 (B) have >22% sequence identity.

The first divergence from common ancestral sequence was of acq-miR167 and cme-miR167e, whereas the subsequent evolutionary divergence has resulted in the separation of ath-miR167c and aly-miR167c-5p (two orthologous miR167 sequences).

This tree shows that all the miR167s, which are processed from the 3′ end of the irrespective pre-MIR167, made one cluster as in group I (Fig. 3).

The miR167-3ps are thought to be processed from 3′ end of stem loop precursors, which are complementary to miR167-5p counterparts (Fig. 4).

Unlike the miR166 sequences, where sequences were intermingled in Multiple Sequence Alignment (MSA) 45, the miR167 sequences from different species (as specified in Table 1) taken for our studies are aligned at a distinct position (Fig. 1).

The topology of both ML and NJ phylogenetic tree for miR167 family members was found to be mostly similar, except changes in position of some members (Fig. 3 and Supplementary Fig. S1).

Likewise, another branch has cme-pre-MIR167b diverged from vvi-pre-MIR167c and sly-pre-MIR167.

In the sub-clade of mdm-pre-MIR167a/h/i, ahy-pre-MIR167 has evolved with higher rate of substitution from gma-pre-MIR167e/f.

A total of 14 unique miR167 (UmiR167-1 to UmiR167-14) sequences have been identified, where 6 UmiR167s are shared by multiple species, 2 UmiR167 have more than 1 miR167 sequences of same species, and 6 UmiR167 sequences have only one miR167 sequence (Table 2).

We have reconstructed the phylogeny of miR167 sequences for studying their sequence conservation and diversification among diverse plant species.

This might be due to the separate evolution pattern of gma-miR167h/i and mdm-miR167a, because these were processed from 3′ end.

The ML tree is divided into two groups, Group I (comprising of twenty three pre-MIR167 sequences) and remaining sequences in Group II (Fig. 6).

The results show that miR167-3p-RC sequences have also clustered together (Fig. 4) as their corresponding miR167-3p sequences (Fig. 3).

Likewise, in clade III, sbi-pre-MIR167d, osa-pre-MIR167d, bdi-pre-MIR167c, osa-pre-Mir167h, zma-pre-MIR167e, osa-pre-MIR167g, osa-pre-MIR167f, sbi-pre-MIR167h, and zma-pre-MIR167j have a common ancestor, sharing with zma-pre-MIR167f, sbi-pre-MIR167g and osa-pre-MIR167j.

To identify the number of miR167 sequences available, we used the miRNA registry database (miRBase version 19, http://microrna.

The clustering of the mdm-pre-MIR167a with the other precursor sequence of apple (mdm-MIR167) species such as mdm-pre-MIR167j, h, i (Fig. 6, clade VIII) suggests that the members of the pre-MIR167 have evolved through probable duplication of same pre-MIR167 sequence and exist as ortholog or homolog in other species 49.

Three of the mature miR167s pairs ath-miR167c/aly-miR167c, gma-miR167h/i and acq-miR167/cme-miR167e have maintained their conservation at the precursor level in clade II, IV and V, respectively (Fig. 6).

Since there are no supporting experimental evidences for the evolutionary relationship of miR167 family till now, it was imperative to study the phylogenetic evolution of miR167.

As we have suspected, all these miR167-3p-RCs formed a separate group from miR167-5p similar to their corresponding miR167-3p sequences (Fig. 4 and Fig. 3).

An unrooted ML phylogeny of miR167s along with reverse complementary sequences of miR167-3p using MEGA5.

The group I contains three miR167-5p (processed from 5′ of precursor) sequences such as gma-miR167h/i and mdm-miR167a along with all miR167-3ps.

In those cases, we have verified miR167 using both RNAshape 55 and Mfold RNA secondary structure prediction 45 tool, using the few parameter settings for secondary structure prediction.

We have also observed that two conserved pre-MIR167s, namely ath-pre-MIR167a and aly-pre-MIR167a, shares the common ancestor with the other two plant pre-MIR167 precursors of Brassica rapa (bra-pre-MIR167s) and Brassica napus (bna-pre-MIR167s).

In group I, gma-miR167h and gma-miR167i have been separated from all other miR167 members in that species (Group II), and made a single cluster in group I (Fig. 3).

This is also evident from our phylogenetic analysis that gma-miR167h/i and mdm-miR167a are not clustered with 5′ derived miR167s, however, clustered with 3′ derived miR167s (Fig. 3), whilst in the phylogenetic analysis of pre-MIR167s these gma-pre-MIR167h/i and mdm-pre-MIR167a are clustered close to those pre-MIR167s which give rise to 3′ derived miR167s (Fig. 6).

The ClustalW alignment in the MEGA5 shows many unique miR167 sequences; each unique mature miR167 sequence may be derived from multiple precursor sequences (Fig. 1).

Similar paralogous sequences like zma-pre-MIR167h and zma-pre-MIR167i are also found in group I. Among ten rice osa-pre-MIR167 sequences, only osa-pre-MIR167e is found in group I with higher substitution rate (Fig. 6).

This suggests that the miR167-3p mature sequences were processed from the 3′ end of their precursor sequences and represented as separate miR167-3p, rather than miR167*.

The miR167 entries in miRBase were further verified using BLAST search in NCBI, (http://www.

Phylogenetic analysis of pre- MIR167 sequences.

The clustering of these sequences in group I, where all the miR167 sequences are processed from 3′ end of stem loop precursors, suggests that these sequences are also processed from 3′ end.

Deletion or addition of sequences during duplication process might have caused the changes in the length of precursors/genes during evolution of mdm-pre-MIR167 family.

This indicates that the ppt-miR167, with faster and higher substitution rate, had undergone sequence diversification from others in group II.

The group I clade of ML tree supported thirty miR167 sequences and rest clustered in group II (Fig. 3).

The key word “miR167” was used as query against miRBase to search the miR167 family members in each plant species.

Percentage Identity of aligned sequences, using Kalmogorov-Smirnov statistical test in GeneDoc (version 2.7), shows that ~0.25 fraction of mature miR167 sequences have ~90% sequence identity.

Sequences in group-I (Fig. 3) show that gma-miR167h/i and mdm-miR167a are distantly separated, but in Fig. 4, these three are clustered in group I and separated from group II consisting of all other miR167 family members of the respective species.

We have further cross verified the precursor MIR167 sequences by Mfold 44 and RNAshape software tools and found that gma-miR167h, gma-miR167i and mdm-miR167a are processed from the 3′ end of their respective precursor sequences (Fig. 5).

We have identified 153 mature miR167 sequences from thirty three different plant species using miRBase Registry database (Table 1).

There are five conserved paralogous pairs of pre-MIR167 like zma-pre-MIR167c/d, tae-pre-MIR167a/b, mdm-pre-MIR167b/e, ath-pre-MIR167b and aly-pre-MIR167b, gma-pre-MIR167b/d (sister to gma-pre-MIR167a) and ptc-pre-MIR167b/d (Fig. 6, clade I).

In clade II, the precursor sequence zma-pre-MIR167g has diverged with higher rate of substitution from ssp-pre-MIR167b, sof-pre-MIR167a (which also have conserved mature miR167 sequences; Fig. 3) and sof-pre-MIR167b, which are separated from sbi-pre-MIR167e (Fig. 6).

However, we could not validate cleavage of Gma-MCCC and Gma-LRPK by gma-miR167h and i, by RLM-RACE.

Similarly, ptc-miR167h-5p and tae-miR167b evolved faster with higher substitution rate than its own family members present in the same cluster (Fig. 3).

Apart from these, three other sequences are processed from 3′ end of stem loop sequences of gma-miR167h, gma-miR167i and mdm-miR167a, which we have observed in our analysis using the Mfold 44 and RNAshape software tools.

Similarly, ~0.25 fraction of the total/precursor sequences (pre-MIR167) have >22% sequence identity (Fig. 2).

Although precursor gma-pre-MIR167s are clustered in group II, their conservation is discrete and appeared in different clades such as gma-pre-MIR167a/b/d, gma-pre-MIR167h/i and gma-pre-MIR167e/f/g/j in clades I, IV and VIII, respectively (Fig. 6).

The Multiple Sequence Alignment of miR167 sequence was performed using ClustalW (Version 2.0) 56 with default parameter settings in MEGA5 phylogenetic analysis tool 57 (Fig. 1).

The separation of precursor sequences of some conserved mature miR167 sequences in precursor phylogenetic tree such as ath-miR167a/b, sbi-miR167, zma-miR167, ptc-miR167 etc.

Stem-loop structures of three MIR167 precursor sequences.

Further verification by reconstructing the tree with the reverse complementary sequences of these miR167-3ps along with other (5p miR167s) proved that miR167-3ps are indeed separate from 5p.

Each mature miR167 sequence was carefully cross checked for its identification using the plant miRNA database web server tools and the sequence data of miR167 obtained were used in this study.

These identified unique miR167 binding sites, known as UTS, were selected on the basis of highest UPE vale (Table 2).

The gma-miR167h and gma-miR167i map to Gm10: 46574263-46574413 [+], Gm20: 37901842-37901992 [−] chromosomal scaffolds respectively, whereas there is no genomic context annotated to mdm-miR167a (http://www.

ClustalW alignment of one hundred and fifty three miR167 sequences retrieved from miRBase database registry (version 19) using MEGA5.

The phylogeny places the ppt-miR167 separately in the group II, but along with most conserved miR167 sequences from the other species (Fig. 3).

Our phylogenomic analysis has suggested that gma-miR167h/i and mdm-miR167a are processed from 3′ end of their precursors, rather than conventionally known processing from 5′ end.

The MIR167 precursors of clade VIII have diverged into three branches, where the two major branches are mdm-pre-MIR167a/h/i and mdm-pre-MIR167g/f.

The two separate scaffolds revealed that the gma-miR167h and gma-miR167i are neither polycistronic nor formed from alternative splicing to share the same nucleotide sequences.

The gma-pre-MIR167g/h and gso-pre-MIR167a have diverged from lja-pre-MIR167.

Phylogenetic analysis of pre- MIR167 sequencesThe pre-MIR167s showed variations in divergence and clustering of sequences.

2

The expression of miRNA167 exhibited little response during the first 2 days after the GA application but was downregulated on the 6th day, and then subsequently upregulated.

downregulated miRNA167 at first, and then subsequently induced it; resulting in upregulating.

The high levels of expression of miRNA167 and miRNA396, and the low expression of miRNA159, in YF trees appears to have inhibited cell division and reduced internode length.

The expression of miRNA159, miRNA167, and miRNA396 in YF and CF shoot tips exhibited their highest level of expression during the period of slow shoot growth.

The potential targets of miR167, miR160, and miR159 play a role in auxin response and the auxin signaling pathway may also participate in SAM development.

miRNA167 regulates vegetative and reproductive growth by controlling the expression of auxin response factors 6 and 8 (ARF6/8) in plants.

Thus, the potential targets of miR167, involved in response to brassinosteroid stimulus and the auxin -mediated signaling pathway, may play a role in internode development.

While the potential targets of miRNA160 and miRNA167 regulate auxin response to control cell division.

The potential targets of miR159, miR166, miR167, miR171, miR172, miR393, miR858, and miR828 are involved in cell growth.

The juvenile to adult phase transition related microRNAs, mdm-miR156 and mdm-miR172, and the flowering related, mdm-miR535, mdm-miR168, and mdm-miR167, also showed high expression levels in apple shoot tips.

The expression of miRNA167 in YF was significantly higher than in CF at 65, 85, and 105 DABB.

The potential targets of miR166 and miR167 are involved in cell wall biogenesis and cell wall organization.

The expression of miRNA167 was significantly higher in YF shoot tips than in CF shoot tips.

Arabidopsis microRNA167 controls patterns of ARF6 and ARF8 expression, and regulates both female and male reproduction.

The highest expression level of miRNA167 in both CF and YF was observed during the slow growth period at 80 DABB.

The top five miRNA families, mdm-miR156, mdm-miR172, mdm-miR171, mdm-miR167, and mdm-miR399, had more than 10 members.

Hence, GA may promote cell elongation mechanisms in YF trees that involve miRNA167-ARF -mediated responses.

3

Gleave et al. [55] previously reported putative orthologues of SPL encoding genes as apple miR156 targets, an apple ARF16 as a miR167 target, orthologues of AP2 and TOE1 as miR172 targets and a copper superoxid dismutase, MdCSD as a miR398 target.

Differential accumulation of miRNA was detected among tissues for miR167, miR160 and miR398 and their respective predicted targets MdARF6, MdARF16 and MdCSD, but the target fold change was low (Figure 6B).

miR167 was highly expressed in the shoot apex and leaf, had low levels of expression in the stem, periderm and xylem and was not detected in the phloem tissue.

Conservation status miRNA family Arabidopsis Oryza(rice) Populus(poplar) Predicted target gene(s) miR156 √ √ √Squamosa promoter -binding proteins[57] miR159/319 √ √ √GAMYB transcription factors[57] miR160 √ √ √Auxin response factors (ARF) [57] miR162 √ √ √DICER-LIKE 1 (DCL1) [57] miR164 √ √ √NAC domain transcription factors[57] miR156/166 √ √ √HD-ZIP transcription factors[57] miR167 √ √ √Auxin response factors (ARF) [57] miR168 √ √ √ARGONAUTE 1 (AGO1) [57] miR169 √ √ √HAP2-like transcription factors[57] miR171 √ √ √Scarecrow-like transcription factors[57] miR172 √ √ √APETALA 2 transcription factors[58] miR390 √ √ √TAS3[59] miR393 √ √ √F-box transcription factors (TIR1) [60] miR394 √ √ √F-box transcription factors[60] miR396 √ √ √GRF, rhodenase[60] miR397 √ √ √laccase[60] miR398 √ √ √Copper superoxid dismutase, CytC oxidase[60] miR403 √ √ √ARGONAUTE 2 (AGO2)[20] miR408 √ √Peptide chain release factor, laccase[20] miR475 √PPR proteins[8] miR476 √PPR proteins[8] Figure 1 Differential expression of miRNAs in apple tissues.

Conservation status miRNA family Arabidopsis Oryza(rice) Populus(poplar) Predicted target gene(s) miR156 √ √ √Squamosa promoter -binding proteins[57] miR159/319 √ √ √GAMYB transcription factors[57] miR160 √ √ √Auxin response factors (ARF) [57] miR162 √ √ √DICER-LIKE 1 (DCL1) [57] miR164 √ √ √NAC domain transcription factors[57] miR156/166 √ √ √HD-ZIP transcription factors[57] miR167 √ √ √Auxin response factors (ARF) [57] miR168 √ √ √ARGONAUTE 1 (AGO1) [57] miR169 √ √ √HAP2-like transcription factors[57] miR171 √ √ √Scarecrow-like transcription factors[57] miR172 √ √ √APETALA 2 transcription factors[58] miR390 √ √ √TAS3[59] miR393 √ √ √F-box transcription factors (TIR1) [60] miR394 √ √ √F-box transcription factors[60] miR396 √ √ √GRF, rhodenase[60] miR397 √ √ √laccase[60] miR398 √ √ √Copper superoxid dismutase, CytC oxidase[60] miR403 √ √ √ARGONAUTE 2 (AGO2)[20] miR408 √ √Peptide chain release factor, laccase[20] miR475 √PPR proteins[8] miR476 √PPR proteins[8] Figure 1 Differential expression of miRNAs in apple tissues.

A, Gel blot analyses of miR156, miR159, miR166, miR167 and miR172 expression.

RNA gel-blot analysis was used to examine the expression of miR156, miR159, miR166, miR167 and miR172 in shoot apex, leaf and stem tissues.

D-F, Spatial expression of miR156, miR167 and miR171 in Arabidopsis inflorescence stem; miR156 and miR167, but not miR171, were detected in Arabidopsis phloem by in situ hybridization using appropriate antisense oligonucleotide probes.

A-C, Spatial expression of miR156, miR167 and miR171 in apple seedling stem; miR156 and miR167, but not miR171, were detected in apple vascular tissue by in situ hybridization using appropriate antisense oligonucleotide probes.

B, Stem-loop RT-PCR analyses of miR156, miR159, miR166, miR167 and miR172 expression.

The relative levels of expression were higher for miR159, miR166 and miR167 than for miR156 and especially miR172, which was barely detectable.

In particular, accumulation of miR156, miR167, miR169, miR390, and miR398 was more than ten-fold higher in the phloem sap than in the vascular tissue, suggesting a possibility of an active mechanism regulating their presence in the phloem sap.

Antisense oligonucleotides corresponding to miR156 and miR167 but not miR171 produced a hybridization signal in the phloem of apple seedling stems (Figure 2A-C).

Both miR156 and miR167 antisense probes produced a strong hybridization signal in the phloem (Figure 2D-E), while no signal was detected with miR171 antisense oligonucleotide probe (Figure 2F), consistent with the RT-PCR results.

Using this approach miR156, miR159, miR160, miR162, miR167, miR169, miR396 and miR398 were clearly detectable; miR172, miR390 and miR393 produced a weak amplification signal; miR166 and miR397 amplification did not produce the expected product, but resulted in a smear not detected in the minus-RT control; miR164, miR168, miR171, miR394, miR403, miR408 and the miRNAs specific to poplar (miR475 and miR476) were not detected (Figure 4).

4

miR167 targets auxin response factors, which are the transcription factors that regulate the expression of auxin-responsive genes and play critical roles in plant development [38], [39], [40].

Targeted gene families were mostly involved in developmental processes and auxin response factors were targeted by two miRNA families - miR160 and miR167.

However, target genes for miR167 were detected only in predicted gene mo dels, and miR171 had predicted target genes only in mesocarp cDNA sequences.

In Ptc2, the miR167 locus was tandemly duplicated.

Relationship between date palm contigPDK_30s943301 containing an miR167 locus and orthologous segments from seven other plant species.

However, only one of two orthologous segments in Arabidopsis thaliana, Oryza sativa and Citrus sinensis had an miR167 locus.

However, one contig containing the miR167 locus showed collinearity between both monocots and dicots (Figure 1).

Evolution of an miR167 Locus within a Conserved Contig between Plant Species.

Our analysis of orthologous contigs containing miR167 between remotely related plant species indicated that genomic duplications significantly influenced the conservation and expansion of miR167 locus (Figure 1).

In Osa7, the miR167 locus was shuffled to a nearby region.

Of 18 orthologous regions, 14 (78%) had an miR167 locus in the collinear region, and four orthologous regions - Ath4, Osa3, Osa10 and Csi5 - lost the miR167 locus.

Phoenix dactylifera and Solanum lycopersicum had one orthologous miR167 locus, indicative of the ancient state of this unduplicated region.

Adding this monocot species into a comparative analysis between different land plants, we analyzed the evolution of an miR167 locus in an orthologous DNA segment shared between eight species from different families.

These observations imply that copies of miR167 can be lost or retained due to different species evolutionary histories, with genome-wide duplication as a major factor, but that conservation of a single copy is universally selected for.

The miR167 locus along with its flanking region was duplicated in seven species from different families, with the miRNA loci maintained in this region (Figure 1 and 4).

Detailed alignments and comparison of orthologous regions in date palm contigPDK_30s943301 (containing one miR167 locus) were conducted to highlight the variation and divergence between date palm and the other reference genomes (Figure 4).

Plants belonging to the Fabids (Populus trichocarpa, Malus domestica and Glycine max) had the highest number of conserved segments, and all miR167 loci were preserved.

Red arrows represent the miR167 loci.

Previous studies have observed that there are two miR167 loci [12], and our analysis showed that one of the two loci was preserved in genomic duplication.

miR167 was predicted to be involved in auxin response transcription factors, which are important for plant architecture.

The results suggest that this miR167 locus may be crucial for both monocots and dicots, as it is far more highly preserved in the process of genomic duplication and new species formation than other miRNAs.

The gain and loss of the conserved miR167 loci implies that conserved miRNAs are maintained despite sequence divergence between different plants as a result of genomic duplication.

Although miR167 was conserved in most orthologous regions from different plants, the flanking genes varied.

Our analysis of duplicated miRNA-containing segments indicated that two miR167 segments in poplar and soybean were duplicated very recently.

The black bars represent the conserved segment, and the triangles indicate the presence of an miR167 locus in the extant plant genomes.

Our analysis indicated that these ancient and conserved segments varied in their maintenance of miR167.

5

Of the DE-miRNAs detected, miR164 targets NAC (NAM, ATAF, CUC) genes, miR167 targets AUXIN RESPONSIVE FACTOR6/8 (ARF6/8), miR390 targets trans-acting small interfering RNA3 (TAS3) transcripts to produce ta-siRNAs, which in turn regulates plant development by repressing ARF2/3/4, and miR393 targets TIR1 genes.

In Arabidopsis, miR167 -targeted ARF6/8 are activators of auxin-responsive genes and could accelerate flowering timing (Nagpal et al., 2005); while the miR390-TAS3-repressed ARF2/3/4 are known to be repressors of auxin-responsive genes and flowering (Fahlgren et al., 2006).

Interestingly, the majority of DE-miRNAs identified in our study (miR162, miR167, miR390, miR393, miR396, miR398, miR408, miR535, miR1511, miR2118, miR3627, and miR7124) showed opposite expression patterns compared to the ones in the previous study on phase transition in apple trees (Xing et al., 2014).

miR164, miR167, miR390, and miR393 were found to involve in auxin signaling.

Among these families, major malus miRNAs, including miRNA156 (9 members), miRNA171 (14), miRNA172 (14), miRNA167 (10), and miRNA395 (9), were detected (Figure 2A).

6

For example, the expression of gma-miR167 is induced by far-red light (Jones-Rhoades and Bartel, 2004; Zhang et al., 2014a), the expression levels of ath-miR408 is induced by light (Li et al., 2014; Zhang et al., 2014a), and the expression levels of ath-miR396 (Zhou et al., 2007), ptc-miR168a-b, ptc-miR169, and ptc-miR398 (Jia et al., 2009) are induced by UV-light.

While differentially-expressed miR172 and miR398 were only present in the T1/N1 group, miR167, miR391, miR7121, and miR858 were only in the T2/N2 group, and miR535 and miR7127 only in the T3/N3 group.

For example, ARFs (auxin response factors) were predicted to be modulated by miR160 and miR167.

7

For example, miR156 could target nine members of the squamosa promoter -binding-like protein family, and miR167 targeted six members of the auxin response factor (ARF) family (Table 2; Table S6 in Additional file 1).

The alignment score threshold was set to 4.5 for conserved and less conserved miRNAs (except for two ARF targets of miR167 and two MYB targets of miR858, for which the score was 5) and to 5 for novel and candidate miRNAs.

The highest read abundance (166,000 RPM) was detected for miR156 and was 5 to 16 times more than other relatively abundant miRNA families, including miR165/166, miR167, miR396, and miR159, whose total abundance ranged from 10,000 to 30,000 RPM (Figure 1a; Table S3 in Additional file 1).

8

In addition, several members of ARF family predicted to be targets of miR160, and in particular miR167 (whose target is the ARF8) (Mallory et al., 2005).

9

The expression of miR167-resistant ARF6 (Figure  5A) leads to arrested ovule development and indehiscent anthers [62].

’, ‘Marcia’, ‘Sympathy’, and ‘Vital’, respectively) and its sequencing frequencies were 10 to 100 times more than other relatively abundant miRNA families, including miR156, miR157, and miR167 (Table  4).

10

This result differs from a study by Xia et al. [5], who found miR167 to be the most abundant in leaf sequencing data, closely followed by miR165/166.

11

A majority of the 42 known miRNA families had several members, and five families, mdm-miR156, mdm-miR171, mdm-miR172, mdm-miR167 and mdm-miR399, had 31, 15, 15, 10 and 10 members, respectively.