To test the potential regulatory impact of miR165/166 on somatic embryogenesis (SE), the level of the candidate target transcripts, PHB and PHV, was evaluated during embryogenic transition induced in Arabidopsis explants. RT-qPCR analyses indicated a significant increase of PHB (up to 8-fold) and PHV (up to 4-fold) transcripts at the early (5 d) and advanced (10 days) SE stages (Figure 1). We analyzed an SE culture of previously characterized STTM165/166 line with an abolished miR165/166 function (Yan et al., 2012) and found a strong accumulation (up to 50-fold) of PHB and PHV transcripts (Figure 1), which suggests a negative impact of miR165/166 on PHB and PHV expression in SE.

The activity of PHB promoter in the embryogenic culture was monitored with the use of the pPHB::GUS line. The analysis showed that in freshly isolated explants (0 day) the GUS signal was dispersed across explant and the strongest signal was detected along the adaxial side of the cotyledons (Figure 2A) that is involved in SE induction (Kurczynska et al., 2007). Further intensification of the GUS signal in the cotyledons was observed in the explant undergoing SE induction (5 and 10 days culture) but the PHB expression signal was also detected in other explant parts including the root (5 days) and hypocotyl (10 days) that are not involved in SE (Figures 2B,C).

Whole mount RNA in situ hybridizations (WISH) with miR166-specific probes were used to examine the spatio-temporal localization of miR166 in explants induced toward SE. The freshly isolated explants (0 day) exhibited strong miR166 signal at the basal explant part including the hypocotyl and the root (Figure 2D). The pattern of miR166 was drastically changed in the explants cultured on SE induction medium and the intensive accumulation of miR166 at the shoot apical meristem (SAM) and its proximity was characteristic of the explants that were cultured for 5 days and at the SAM or cotyledons at the 10th day of SE (Figures 2E,F).

In order to verify the hypothesis that miR165/166 might negatively regulate the PHB transcripts during SE induction, we compared the pattern of the GFP signal in pPHB::PHB-GFP with the pPHB::muPHB-GFP culture expressing PHB transcripts that are resistant to the miR165/166 cleavage (Miyashima et al., 2011). Analysis of the GFP signal showed that the PHB expression undergoes extensive spatio-temporal changes in the explants of both analyzed lines, but the pattern of PHB signal localization was distinctly different in 5 and 10 days culture of pPHB::muPHB-GFP in comparison to pPHB::PHB-GFP (Figures 2G–I vs. Figures 2J–L) explants. In the pPHB::muPHB-GFP tissue induced toward SE the PHB signal was detected in various explant parts including the hypocotyl, SAM and cotyledons (Figures 2H,I), while in the pPHB::PHB-GFP culture, the signal was limited to the cotyledons (Figures 2K,L). The apparent differences in the PHB expression pattern in the pPHB::muPHB-GFP vs. pPHB::PHB-GFP culture include a lack of PHB expression in the hypocotyl and root part and less intensive GFP signals in the cotyledons of pPHB::PHB-GFP (Figures 2G–I vs. Figures 2J–L).

In order to investigate a potential relationship between miR165/166-PHB/PHV and LEC2 in SE we analyzed LEC2 expression levels in embryogenic cultures with disturbed expression of the miR165/166 and PHB/PHV transcripts. We found that accumulation of the PHB transcripts in the gain-of-function phb1-d mutant and a STTM165/166 line led to the significantly increased LEC2 transcription (Figure 3A). Further evidence of a regulatory relationship between miR165/166-PHB/PHV and LEC2 was provided by the analysis of the embryogenic culture overexpressing LEC2 (Figure 3B). We observed that the overexpression of LEC2 during SE resulted in a significantly increased level of the PHB and PHV transcripts, thus inferring a positive feedback loop between PHB/PHV and LEC2. In support of this postulate, we found that similar to PHB, also LEC2 is expressed at 5th and 10th day of culture in the cotyledons, i.e., the explant parts that are involved in SE induction (Figures 3C–E). Collectively, these observations support the hypothesis on a regulatory relation between miR165/166-PHB and LEC2 in the embryogenic transition.

Surprisingly, the up-regulation of LEC2 transcripts was also noticed in SE culture of the phb and phv insertional mutants with defected expression of PHB (knock-out) and PHV (knock-down) genes, respectively (Figure 3A). To explain this result, we found the PHB and PHV transcripts to be up-regulated in phv and phb cultures, respectively (Figure S3). In conclusion, the level of the PHB and PHV transcripts appears to be controlled by a compensative regulatory mechanism in which insufficient expression of one of these genes results in a significantly increased transcription of the other gene.

To verify the assumption of the existence of a regulatory relationship between miR160 and ARF10/ARF16/ARF17 in SE, the cultures derived from miR160b and miR160c insertional lines were analyzed. The results indicated that the expression of ARF10 and ARF16 was up-regulated in the mutant cultures while the ARF17 transcription was down-regulated (Figure 4A). In addition, a significantly increased accumulation of the ARF16 transcripts was observed in cultures from the mARF16 line that carry the miR160-resistant form of ARF16 (Figure 4B). A negative feedback loop between ARF10/ARF16 and miR160 in SE might be suggested as we found an increased miR160 level in the cultures of mARF16 and arf10arf16 mutants (Figure 4C).

WISH analysis with a miR160-specific probe indicated that miR160 localization patterns changed in explants during SE. More specifically, the explant at 0 day exhibited miR160 signal along a basal part of the explant while after 5 days of culture on the SE-induction medium, a strong accumulation of miR160 was observed in the SAM and its proximity (Figures 4D–E). At the advanced stage of SE (10 day) the miR160 signal was moved from the SAM area into the embryogenic tissue produced on cotyledons (Figure 4F) and a strong accumulation of miR160 was characteristic of somatic embryos (Figure 4f).

To investigate a relation of miR160 with the LEC2-mediated pathway of SE induction, the expression level of LEC2 in the embryogenic cultures with impaired expression of miR160 (miR160b, miR160c) and its target genes, ARF10 and ARF16 (mARF16, arf10arf16) was analyzed. LEC2 transcripts were accumulated in miR160b, miR160c, and mARF16 cultures and down-regulated in arf10arf16 culture (Figure 5A). Thus, a regulatory relationship between miR160-ARF10/ARF16 and LEC2 appears to be engaged in SE induction. In addition, a positive impact of LEC2 on ARF10/ARF16 cannot be ruled out as we noticed the increased expression of ARF10 and ARF16 in the embryogenic culture with LEC2 overexpression (Figure 5B). In contrast, ARF17 transcripts was not affected in the LEC2- overexpressing culture suggesting that LEC2 does not regulate ARF17 during SE.

The regulatory relationships of miR160 and miR165/166 with the LEC2 gene, a key regulator of SE through the control of auxin biosynthesis (Wójcikowska et al., 2013), suggested that the biological function of these miRNAs during SE might be related with auxin response. Consistent with this hypothesis, we observed that the embryogenic response that was displayed by the mutants with a disrupted expression and function of miR160 and miR165/166 on the auxin media was significantly different to the WT culture. That is, the miR160b, miR160c, mARF16, and STTM165/166 explants produced somatic embryos on the auxin-free medium and supplementation of the medium with 2,4-D drastically impaired their embryogenic response in a concentration-dependent manner (Figures 6A,B; Table S1). Thus, we hypothesized that miR160 and miR165/166 impact the auxin content in the SE-induced explants. In support of this hypothesis, we found that the levels of IAA-related indolic compounds (Bric et al., 1991) were significantly higher in the cultures of miR160b, miR160c, and STTM165/166 (Figure 6C). In addition, we found that the representative auxin-inducible AUX/IAA genes, IAA17, and IAA29 (Overvoorde et al., 2005), were up-regulated in the SE cultures with a reduced expression of miR160 (miR160b, miR160c) or an increased expression of the presumptive targets (STTM165/166, mARF16) (Figure S4). Because our results indicated that miR160 and miR165/166 control the LEC2 that has a regulatory role in the YUCCA (YUC)-mediated pathway of auxin biosynthesis during SE (Wójcikowska et al., 2013), we profiled the expression of YUC genes in STTM165/166, mARF16, miR160b, and miR160c cultures. We found that YUC1, YUC4, and YUC10 genes were highly up-regulated in these cultures (Figure 6D). Collectively, our results suggest that miR160 and miR165/166 may contribute to the embryogenic potential of Arabidopsis somatic tissues via the regulation of the LEC2-controlled pathway of auxin biosynthesis.

Given that both miR160 and miR165/166 impact LEC2 expression during SE we then investigated the regulatory relationships between presumptive miR160 (ARF10, ARF16) and miR165/166 (PHB/PHV) targets. We found that the level of ARF10 and ARF16 transcripts was significantly increased in the phb, phv, phb1-d and STTM165/166 lines (Figure 7A), thus suggesting that PHB might positively affect ARF10 and ARF16 expression. Moreover, PHB was up-regulated in the culture of arf10arf16 double mutant, and down-regulated in mARF16, miR160b and miR160c indicating that ARF10 and ARF16 might negatively control PHB expression (Figure 7B).

Altogether our results suggest that miR160 and miR165/166, possibly through the regulation of ARF10/ARF16 and PHB/PHV, respectively, contribute to the SE induction mechanism associated with LEC2-controlled auxin biosynthesis pathway (Figure 8). However, further experiments are necessary to determine the mode of interaction between the miR160 and miR165/166 pathways during SE induction and to further elaborate the gene regulatory networks that they are involved in.

The A. thaliana (L.) Heynh. Columbia (Col-0) and Landsberg erecta (Ler) parental genotypes and the transgenic lines miR160b, miR160c, phb, phv, phb1-d and pPHB::GUS were supplied by Nottingham Arabidopsis Stock Centre (NASC). The T-DNA insertional double mutants arf10-2 arf16-2 (hereafter noted as arf10arf16) and P_ARF16::mARF16 (hereafter noted mARF16) were kindly provided by Prof. Xiaoya Chen (Institute of Plant Physiology and Ecology, China). The seeds of 2x35S::STTM165/166 line (hereafter noted as STTM165/166) were kindly provided by Dr. Guiliang Tang (Michigan Technological University, USA). The pPHB::PHB-GFP and pPHB::muPHB-GFP line that carries the mutated, resistant to the miR165/166 cleavage version of the PHB transcript (Miyashima et al., 2011), were kindly provided by Prof. Keiji Nakajima (Nara Institute of Science and Technology, Japan). The 35S::LEC2-GR line overexpressing LEC2 upon DEX treatment was described previously (Ledwon and Gaj, 2009). To monitor expression of LEC2 the pLEC2::GFP line with nuclear localized GFP was generated by cloning the LEC2 promoter sequence (−2020 to +5 relative to ATG of LEC2) into pCR™8/GW/TOPO® (Invitrogen), and then recombining these plasmids with Gateway-compatible pCGTAG plasmids using LR clonase (Invitrogen). Col-0 plants were transformed with use of the floral dip method (Clough and Bent, 1998).

The characteristics of the transgenic lines that were used in the study are presented in Table S2A.

Seeds sterilized with 20% solution of commercial bleach were plated on 0.5x MS solid medium. The plates were kept at 4°C in darkness for 2 days and then transferred to a growth chamber at 21 ± 1°C under a 16/8h photoperiod of 40 μM m⁻²s⁻¹ white, fluorescent light. The young seedlings were transplanted into Jiffy-7 pots and grown in a walk-in type green room under the conditions described above until harvesting of siliques.

Immature zygotic embryos (IZEs) at green cotyledonary-stage were used as explants to induce SE. IZEs were cultured in E5 solid medium with 5 μM 2,4-dichlorophenoxyacetic acid (2,4-D) according to Gaj (2001). In some experiments (Figures 4A,B), other concentrations of 2,4-D were used including 0; 0.1; 1.0; 10.0; 30.0 μM. The explant capacity for SE was evaluated in 21 day-culture and two parameters were calculated: SE efficiency (the percentage of explants that formed somatic embryos) and SE productivity (the average number of somatic embryos per explant). All of the culture combinations were evaluated in three replicates and at least 30 explants (ten explants/Petri dish) were analyzed per one replicate.

The whole mount in situ hybridization (WISH) of miRNA molecules was performed according to modified protocol of Dastidar et al. (2016). Embryos/explants were dissected/collected in a drop of PBS (Phosphate-buffered saline buffer) and immediately transferred to 4% paraformaldehyde on ice. LNA Digoxin 5'-end labeled probes were used following to the Dastidar et al. (2016). The slides were scanned using Panoramic FLASH 250 II.

An RNAqueous kit (Ambion) was used to isolate total RNA and mirVana^TM Kit was used to isolate miRNAs from the IZE explants. Depending on the age of a culture, 300 (0 day culture) to 100 (10 days culture) explants were collected for RNA/miRNA isolation. The concentration and purity of RNA was evaluated with a ND-1000 spectrophotometer (Nano-Drop). To avoid DNA contamination, RNA was treated with RQ1 RNase-free DNase I (Promega) following the manufacturer's instructions. First strand cDNA was produced using a RevertAid First Strand cDNA Synthesis Kit (Fermentas). RT-qPCR was carried out in a 10 L⁻¹ reaction volume using a LightCycler 480 SYBR Green I Master (Roche). The primers that were relevant to the genes being studied were used in the RT-qPCR analysis (Table S2B). The RT-qPCR reactions were performed as previously described in Wójcik and Gaj (2016). Primary data analysis was performed using LightCycler Software (Roche).

The oligonucleotides design, stem-loop RT and real time qPCR were performed according to Speth and Laubinger (2014). The primers sequences used in the study are listed in Table S2B. Primary data analysis was performed using LightCycler Software (Roche). Relative RNA levels were calculated and normalized to an internal control of the At4g27090 gene encoded 60S ribosomal protein (Thellin and Zorzi, 1999). In all of the analyzed tissue samples, the control gene displayed a constant expression pattern with Cp = 19 ± 1. The plant tissues for the analysis of gene expression were produced in three biological replicates and two technical replicates of each repetition were carried out. The relative expression level was calculated using 2^−ΔΔCt.

To detect GUS signal the pPHB::GUS explants were stained with a standard X-Gluc (Sigma Aldrich) solution at 37°C for 12 h according to Jefferson et al. (1987). The tissue was inspected under Delta-Optical SZ-630T microscope and images were saved as jpg files with a Canon EOS 60D camera.

Analysis of GFP signal was carried out using a Nikon Eclipse Ni-E/Ni-U fluorescent microscope system. GFP fluorescence was excited with halogen lamphouses with a 100–240 VAC (Prior Lumen200) and a wavelength of 488 nm. Photographic documentation was recorded by Nikon Digital Sight DS-Fi2 with DS-U3 camera, using the NIS-Elements F computer program version 4.0.

A colourimetric technique that enabled the detection of indolic compounds, including IAA, was applied (Bric et al., 1991; Wójcikowska et al., 2013). To evaluate the relationship between IAA content and miR160 and miR166 activity, IZE-derived cultures of Col-0 and miR160b, mARF16, STTM165/166 genotypes were analyzed. IZE explants were induced on a 5μM 2,4-D medium and tissues were sampled on the 5th or 10th day of culture. The procedure was performed as described previously (Wójcikowska et al., 2013). Each analysis was carried out in three biological replicates.

The statistical analyses were performed using either the Student t-test or a two-way ANOVA (p < 0.05) followed by Tukey's honestly significant difference test (Tukey HSD-test) (p < 0.05). The figures show the averages from at least three biological replicates with the standard error.

During ZE, miR165/166-mediated repression of the PHB and PHV genes enables the morphogenesis-to-maturation transition and correct specification of shoot apical meristem (SAM) and root apical meristem (RAM) in the developing embryo (Grigg et al., 2009; Miyashima et al., 2013). Relevant to zygotic development, our results suggest that miR165/166 restricts PHB/PHV expression to control embryonic development in Arabidopsis somatic cells cultured in vitro. In support of this postulate, we found that the expression profile of PHB/PHV is inverse to the one that was indicated for miR165/166 in the WT embryogenic culture (Szyrajew et al., 2017) (Figure S1A). In line with this finding, we showed the PHB and PHV transcripts to be accumulated in the culture of the STTM165/166 lines with an abolished miR165/166 function. In order to further test the regulatory relationship between miR165/166 and PHB, the spatio-temporal patterns of PHB and miR166 were investigated in SE-induced explants with use of reporter lines (pPHB::GUS; pPHB::PHB-GFP) and WISH, respectively. Our results indicated changes in the localization of PHB signal and miR166 during SE induction. Although PHB and miR166 both accumulate in the proximity of the shoot apical meristem (SAM), the area that is involved in SE induction (Kurczynska et al., 2007), their expression patterns appear to not overlap. Accordingly, cotyledons that are predominantly involved in somatic embryo formation displayed a strong expression of PHB while the WISH signal of miR166 was limited to the SAM and its proximity. Further evidence that miR166 might spatio-temporally repress the PHB transcripts in SE-induced tissue was supported by the observation of a more widespread and intense PHB expression in the pPHB::muPHB-GFP than in the pPHB::PHB-GFP culture (Figures 2G–I vs. Figures 2J–L). miR165/166 appears to spatiotemporally restrict the PHB transcripts in the explant parts/cells that are not responsive to SE induction. In contrast, the SE-responsive cells that are dispersed along the explant cotyledons (Kurczynska et al., 2007) seem to accumulate PHB transcripts (Figures 1, 2K,L), which in part, might result from the decreased content of miR166 (Szyrajew et al., 2017). However, more advanced cytohistological analysis is required to evaluate the relation between the miR165/166 content and the PHB transcript level in the explant cells that are undergoing SE induction.

Moreover, consistent with the finding that the pPHB::PHBmu-GFP plants phenocopy the miR165/166-resistant phb1-d mutant (Miyashima et al., 2013), we found that the pPHB:muPHB-GFP line showed a reduced SE response, which was similar to the phb1-d mutant (Figure S5). The impaired SE response of pPHB:muPHB-GFP and phb1-d explants on an auxin medium that is standard for SE-induction is possibly caused by an increased IAA accumulation, which might result from the upregulation of LEC2 (Figure 3A) and the activation of auxin biosynthesis YUC genes (Figure 6D). A relation between the endogenous auxin accumulation and the impaired embryogenic potential of tissue on a standard auxin medium was indicated for different genotypes in the present (Figure 6B) and other studies (Wójcikowska et al., 2013; Wójcik and Gaj, 2016). Altogether, the analysis of the reporter lines suggests that in SE-induced explants miR166 might restrict PHB expression to the cotyledon tissue and the resulting spatial distribution of PHB expression colocalizes with explant sites developing somatic embryos.

The rapid and intense accumulation of PHB transcripts in SE-involved explant parts that we observed in response to SE induction medium suggests that PHB might be involved in the very early events associated with the embryogenic transition of somatic cells. The genetic regulation of early events associated with embryogenic transition in the in vitro cultured somatic cells are poorly known and identification of the genes acting up- and down-stream from PHB in the very early SE culture might provide new insights into molecular mechanism of SE induction.

Recently, the impact of miR160 on developmental processes induced in vitro was reported. Accordingly, miR160-mediated repression of ARF10/ARF16/ARF17 was postulated to control the embryogenic response in culture of D. longan (Lin et al., 2015) and miR160-ARF10 was shown to control cellular reprogramming and callus formation in Arabidopsis (Liu et al., 2016). Our results indicate that miR160 might also contribute to SE induction in Arabidopsis through the repression of ARF10, ARF16, and ARF17, in light of (i) the opposite expression patterns of ARF10/ARF16/ARF17 and miR160 in the Arabidopsis embryogenic culture (Szyrajew et al., 2017; Wójcikowska and Gaj, 2017) (Figure S1B); (ii) the increased expression of ARF10 and ARF16 in the miR160 mutant cultures and the accumulation of ARF16 in the cultures expressing the miRNA-resistant version of ARF16. Differences in the accumulation of ARF10/ARF16 at distinct time points of the miR160 mutants SE culture may be caused by a redundancy between the members of the MIR160 family, all of which are able to cleave transcripts of ARF10 and ARF16 (http://plantgrn.noble.org/psRNATarget/).

The unimpaired SE response in arf10 and arf16 single mutants compared to the impaired SE response of double arf10arf16 mutants (Figure S5) suggests that ARF10 and ARF16 seem to function redundantly during SE. In support of this postulate, the redundant function of ARF10 and ARF16 was indicated in root cap formation (Wang et al., 2005) and seed dormancy (Liu et al., 2013) and recently, ARF10/ARF16 together with IAA17 were found to act as a protein complex (Ye et al., 2016).

Interestingly, we observed the significant accumulation of miR160 in mARF16 and arf10arf16 cultures that suggests a role of ARF16 in the control of miR160 during SE. The feedback regulation of miRNA production by the associated target genes was reported for mammalian and plant miRNAs (Wu et al., 2009; Lai et al., 2016). Importantly for the present result, miR160 was found to be negatively regulated by the targeted ARF17 to control auxin homeostasis and adventitious rooting in Arabidopsis (Gutierrez et al., 2009). Thus, it cannot be excluded that during SE in Arabidopsis miR160 is controlled by the targeted ARFs but this assumption needs further experimental verification.

The distinct differences in the level of the primary transcripts of MIR160a, MIR160b, and MIR160c in Arabidopsis embryogenic culture (Szyrajew et al., 2017) suggest that MIR160 genes contribute differently to the regulation of SE. In line with this assumption we found ARF10 and ARF16 but not ARF17 to be accumulated in miR160b and miR160c mutant cultures. This result implies that during SE ARF17 might be under the control of MIR160a. In support for this postulate, exclusively MIR160a was indicated to control ARF17 during early development of zygotic embryo (Liu et al., 2010). The analysis of ARF17 expression in the SE culture with defected MIR160a expression would verify this hypothesis but the severely reduced fertility of the foc mutant (miR160a) (Liu et al., 2010) makes isolation of immature zygotic embryos and thus the establishment of SE culture difficult.

The assumption that a role of ARF17 in SE might be different to ARF10/ARF16 supports also the finding that in embryogenic culture expression of ARF17, in contrast to ARF10/ARF16, is not affected by LEC2 overexpression. Moreover, the expression pattern of ARF17 in SE differs from ARF10/ARF16 and accordingly, a level of ARF17 mRNAs increases in late SE and is auxin-independent (Wójcikowska and Gaj, 2017). Thus, it cannot be ruled out that miR160-ARF17 operates in the advanced SE culture associated with the formation of somatic embryos. Suggestive for this postulate is formation of the defective zygotic embryos in plants with suppressed miR160-directed regulation of ARF17 (Mallory et al., 2005).

Importantly for a role of miR160- and miR165/166-regulated pathways in SE, we found that transgenic forms with defective expression and function of miR160 (miR160b, miR160c, mARF16) and miR165/166 (STTM165/166) were capable to produce somatic embryos on auxin-free medium and auxin treatment severely impaired their embryogenic response. Similar capacity for SE induction on auxin-free medium displayed also the culture overexpressing LEC2 and accumulated IAA (Ledwon and Gaj, 2011; Wójcikowska et al., 2013). Similar to the culture overexpressing LEC2, in embryogenic culture of the miR160 and STTM165/166 we found increased accumulation of the indolic compounds and enhanced expression of LEC2 that was coupled with activation of the YUC (YUC1, YUC4, and YUC10) genes encoding the auxin biosynthesis enzymes involved in SE induction in Arabidopsis (Wójcikowska et al., 2013).

The results imply that miR165/166 might control SE induction by impacting the PHB and PHV genes that encode the direct activators of LEC2 (Tang et al., 2012). Although the PHB and PHV are closely related and demonstrate the high degree of functional interchangeability (McConnell et al., 2001; Prigge et al., 2005) the mechanism adjusting their individual contribution to the controlled processes, including SE, has not been yet revealed.

In support for the LEC2-related function of PHB in SE we observed the up-regulated expression of LEC2 in embryogenic culture of phb-1d mutant that was reported to spontaneously produce somatic embryos (Tang et al., 2012). Moreover, we found LEC2 overexpression to enhance PHB/PHV transcription level suggesting that a positive feedback regulation exists between PHB/PHV and LEC2 during SE. In support of this assumption, we have identified the RY- motif recognized by LEC2 (Braybrook et al., 2006) in the PHB promoter (AGRIS Atcis DB) but the experimental verification for the direct binding of LEC2 to the PHB promoter during SE would be necessary to test this further. Interestingly, the LEC2-binding RY-motif is also present in the promoters of the MIR165/166 genes (Wang and Perry, 2013) and thus it is conceivable that LEC2 might also control expression of MIR165/166 genes during SE. So far, few TFs directly regulating MIRNA genes have been implicated in plants including activation of MIR165a and MIR166b by SHR (SHORT ROOT) and SCR (SCARECROW) during post-embryonic development (Carlsbecker et al., 2010; Miyashima et al., 2013). LEC2 regulation of MIRNA genes has not been reported yet but it cannot be excluded as FUS3, a TF structurally and functionally related with LEC2 (Harada, 2001), was suggested to control MIR156, MIR160, MIR166, MIR396 genes in the embryogenic culture of Arabidopsis (Wang and Perry, 2013).

Altogether, several lines of evidence infer that the miR165/166-PHB/PHV regulatory node controls induction of the embryogenic program in somatic cells of Arabidopsis through targeting LEC2. The possible role of miR165/166 in the regulation of HD-ZIP III TFs during SE was also postulated in sweet orange and Larix leptolepis but the targeted effectors and molecular pathways controlled were not identified (Wu et al., 2011; Li et al., 2013).

Our results suggest that miR165/166-PHB/PHV and miR160-ARF10/ARF16 regulatory modules might regulate SE induction through LEC2. Accordingly, the significant changes in LEC2 expression levels in SE cultures with a disturbed expression and function of the ARF10 and ARF16 genes (mARF16 and arf10arf16) suggest that these ARFs positively regulate LEC2. In addition, ARF10 and ARF16 seem to contribute to LEC2 regulation in SE due to the auxin-stimulated expression of LEC2 and the similarity of spatio-temporal expression pattern of LEC2 and ARF10/ARF16 in embryogenic culture (Kurczynska et al., 2007; Ledwon and Gaj, 2009; Wójcikowska and Gaj, 2017). Considering that ARF10 and ARF16 repress auxin-regulated genes (Guilfoyle and Hagen, 2007) they are unlikely to control LEC2 directly and the intermediary genetic elements remain to be identified.

In conclusion, both the miR165/166- PHB/PHV and miR160- ARF10/ARF16 nodes might control the embryogenic transition in Arabidopsis somatic cells via regulating LEC2, which is a key regulator of SE induction. The convergent functions of miR160 and miR165/166 in regulation of a common TF, the WOX5 (WUSCHEL-RELATED HOMEOBOX 5) gene, was reported in control of distal stem cell differentiation and embryonic root development (Grigg et al., 2009; Ding and Friml, 2010). Given that a role of WOX5 in formation of RAM in somatic embryos of Arabidopsis was reported (Su et al., 2015; Wang and Chong, 2016) the contribution of miR160 and miR165/166 to SE through regulation of the embryonic root development might be considered.

In addition to controlling the auxin biosynthesis-related LEC2 gene, ARF10/ARF16 might also impact SE induction via the regulation of the genes that are involved in the signaling of other hormones including ABA. Accordingly, miR160-ARF10 was found to play an important role in ABA-auxin crosstalk in seed germination and post-embryonic developmental programs (Liu et al., 2007). ARF16 was indicated to be required for the ABI3 expression (Liu et al., 2013) that encodes a transcription factor that is involved in ABA signaling during seed development (Finkelstein et al., 2002). It is worth noting that an abi3-1 mutant that was insensitive to ABA was found to be significantly impaired in its SE response (Gaj et al., 2006). These findings together with the extensive interactions between auxin and ABA signaling during plant development (Rock and Sun, 2005; Teale et al., 2008; Thole et al., 2013) including the induction of SE in Arabidopsis (Braybrook and Harada, 2008) infer that the disturbed ABA sensitivity that is expected in mARF16 might enhance auxin perception/signaling and as a result, an embryogenic response is triggered. Recently, the role of the ARF10-miR160 module in the regulation of cytokinin-auxin crosstalk was indicated (Liu et al., 2016). Important for the SE-induction mechanism, overexpression of miR160 was shown to enhance tissue sensitivity to cytokines (Turner et al., 2013), which were reported to play a key role in auxin-induced SE in carrot (Tokuji and Kuriyama, 2003) and Arabidopsis (Su et al., 2015; Wang and Chong, 2016).

The regulatory interaction between miR165/166 and miR160 in SE seems to include the negative feedback loop between ARF10/ARF16 and PHB/PHV, the targets of miR160 and miR165/166 pathways, respectively. Experimental supports for this notion include: (i) a positive impact of PHB on the ARF10/ARF16 expression level (increased ARF10/ARF16 transcription in phb, phv, phb1-d, STTM165/166 (ii) a negative relation of ARF10/ARF16 on PHB transcription in the mARF16, miR160b and miR160c cultures. The higher expression of ARF10 and ARF16 in the STTM165/166 than in the phb1-d culture may be caused by an increased level of miR160, which is able to cleave ARF10/ARF16 transcripts (Wang et al., 2005) in phb1-d (Figure S6A). In support of a possible role of PHB in the direct activation of ARF10, ARF16 in SE, a binding of PHB to ARF5 promoter during vascular patterning in Arabidopsis was documented (Müller et al., 2015). However, it is also possible that ARF16 might repress PHB via up-regulation of miR166 as we found a decreased level of miR166 in arf10arf16 culture (Figure S6B).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.