Wild-type plants used were the Arabidopsis thaliana ecotype Columbia-0. The 35S:MIR408 (MIR408-OX) and amiR408 lines were previously described by Zhang and Li (2013). The T-DNA insertion lines of SALK_091945 for ARPN and SALK_023935 for LAC13 were from the Arabidopsis Biological Resource Center (ABRC). The sterilized seeds were sown on half-strength Murashige and Skoog (MS) medium after storage at 4°C for 4 days. The seedlings were grown in a growth chamber with the following settings: standard long-day (16 h of light/8 h of darkness) conditions at 23 ± 1°C, light intensity of 100–120 μmol m^-2 s^-1, and a relative humidity of approximately 50%. All phenotypic characterization experiments were conducted on multiple biological samples and repeated at least three times.

To obtain the 35S:Plantacyanin (ARPN-OX) transgenic plants, the full length cDNA of Plantacyanin gene was amplified by PCR using primer pairs with restriction endonuclease cleavage site, and inserted into the pJim19 binary vector (Basta resistance; Zhang et al., 2008), then transformed into the wild type background. Transgenic plants were selected with 20 mg/L Basta. T3 generation homozygous lines were used for all experiments.

To obtain the 35S:LAC13 (LAC13-OX) transgenic plants, the full length cDNA of LAC13 was amplified by PCR using primer pairs with restriction endonuclease cleavage site, and inserted into the pJim19 binary vector (Hygromycin resistance; Zhang et al., 2008), then transformed into the wild type background. Transgenic plants were selected with 40 mg/L hygromycin. T3 generation homozygous lines were used for all experiments.

The ARPN-OX/LAC13-OX plants were generated by crossing ARPN-OX and LAC13-OX transgenic plants. F2 or F3 homozygous progenies were identified by growing the seedlings on MS medium containing both hygromycin and basta.

All the PCR products were obtained using Phusion high-fidelity DNA polymerase (NEB¹) and the resultant constructs were sequenced by GENWIZE to ensure their integrity. The plasmids were electroporated into Agrobacterium tumefaciens GV3101, respectively. Arabidopsis transformation was performed by the floral dip method (Clough and Bent, 1998). All of the primers used to generate the above-mentioned constructs are listed in Supplementary Table S2.

Total RNA was extracted from Arabidopsis seedlings using the RNeasy plant mini kit (Qiagen) and quantified by NanoDrop 2000, and the integrity was examined by agarose gel electrophoresis. For real-time qRT-PCR, 5 μg of DNaseI treated RNA was reverse transcribed using the SuperScript II reverse transcriptase (Invitrogen), and the resultant cDNA was analyzed using the Power SYBR Green PCR Master Mix (Takara) with a Bio-Rad CFX96 real-time PCR detection system in triplicate. A total volume of 20 μl reaction system was set up according to the manufacture’s instruction, including 10 μl of 2x SYBR Green Master Mix, 0.2 μM of forward primer, 0.2 μM of reverse primer, 50 ng of cDNA template, ddH₂O up to 20 μl. The Actin2 amplicon was used for normalization. Determination of relative gene expression level was carried out using the standard 2^-ΔΔC(T) method. All experiments were performed on three independent biological samples with each including three technical replicates. Statistical significance was calculated using the two-tailed Student’s t-test. For RNA gel blot analysis, 20 μg of total RNA were loaded per lane and blotting was performed as described previously. Fragments of ARPN and LAC13 used for probe labeling were generated by PCR. Primer sequences are listed in Supplementary Table S2.

Measurement of glucose content was performed using the Glucose and Sucrose Assay Kit (Biovision) according to the manufacturer’s instructions as described previously (Zhang et al., 2014). Rosette leaves from the plants grown in soil under standard long-day (16 h of light/8 h of darkness) conditions for 2 weeks were homogenized in a glucose assay buffer, and the supernatant was collected after centrifuging at 12,000 rpm for 10 min. The reaction system was set up in a 100-μL total volume including 50 μL of sample in glucose assay buffer and 50 μL of glucose assay mix (46 μL of glucose assay buffer, 2 μL of glucose probe, and 2 μL of glucose enzyme mix) and incubated at 37°C for 30 min. The A₅₇₀ was collected, and glucose concentrations of the test samples were calculated based on the standard curve. The experiments were performed on three independent biological samples.

Proteins were extracted from the rosette leaves of plants grown in soil under standard long-day (16 h of light/8 h of darkness) conditions for 2 weeks, and immunoblotting was performed as described previously (Feng et al., 2004). After electroblotting on a nitrocellulose membrane, protein gel blot analysis was performed using antibodies against plastocyanin (Acris Antibodies, AS06141). Detection was performed using goat anti-rabbit IgG (H+L) horseradish peroxidase conjugate secondary antibodies (Sigma). RPT5 (Abcam, ab22676) was used as loading control.

Dry weight was measured by drying aerial parts of 2-week-old plants grown in soil at 80°C for 10 h. For cell size measurements, a fully expanded first rosette leaf was fixed in FAA [formalin:glacial acetic acid:ethanol (70%) = 1:1:18)]. The epidermal cells of the leaves or the petioles were peeled off by hand, and then photographed using a Leica DM5500 microscope. The epidermal cell area and mesophyll cell length were measured using ImageJ analysis software.

Total RNA from three biological replicates was extracted with an RNeasy plant mini kit (Qiagen) from the aerial parts of 10-day-old plants grown on MS plate under continuous light condition. Library construction and sequencing on the HiSeq 2000 platform were performed according to the manufacturer’s instructions (Illumina) by BioMarker cooperation (Beijing, China). Bowtie (Langmead et al., 2009) was used to map the sequence reads to the Arabidopsis genome (TAIR 10) and only uniquely mapped reads were used in subsequent analyses. Reads mapped to exonic regions of annotated gene models were normalized against the length of the transcript. On average, 74.4% of the total reads mapped to the Arabidopsis reference genome sequence (Supplementary Table S3), with most of the reads mapping to exons and only a small portion mapping to introns or intergenic regions. To compare gene expression between wild type and MIR408-OX genotypes, length-normalized read density was quantile-normalized. Relevant genes were those identified as differing in expression by at least 1.5-fold (p < 0.01) in the test and control samples. The raw sequence data reported in this paper have been deposited in the Genome Sequence Archive (Wang et al., 2017a) in BIG Data Center (Wang et al., 2017b), Beijing Institute of Genomics (BIG), Chinese Academy of Sciences, under accession number CRA000553 that are publicly accessible at http://bigd.big.ac.cn/gsa.

The net photosynthesis was measured using a portable photosynthesis system (CIRAS-1, PP Systems, United Kingdom) by an open system. Measurements were made by attaching a light source to the leaf chamber window under saturating photosynthetic photon flux densities (1500 μmol m^-2 s^-1), ambient CO2 concentration (Ca) at 290 μmol mol^-1, and relative humidity of 50–60%. Leaf temperature and vapor pressure deficit (VPD) were maintained at 28°C and 0.99 ± 0.2 kPa, respectively. Data were determined at least in six leaves from six different plants of each genotype grown in soil for 2 weeks in a growth chamber.

For cellular copper content analysis, 10-day-old seedlings grown on half MS medium under continuous light were harvested and weighted, and then washed twice with 1 mM EDTA and once with double-distilled water. The following treatment and measurement were performed as previously described (Zhang et al., 2014).

The endogenous hormone content was measured using the Agilent 1100 high-performance liquid chromatography (HPLC) system (Agilent, Palo Alto, CA, United States). The aerial parts from more than fifteen individual plants of wild type and MIR408-OX plants grown in soil for 2 weeks were cold-dried under vacuum. 0.2 g dried leaves were added to 10 mL 80% aqueous methanol and immediately homogenized on ice, then kept at 4°C overnight in darkness with continuous shaking. The homogenates were centrifuged for 10 min at 4500 g at 4°C, and the supernatant was collected and evaporated under vacuum. Dry residue was re-dissolved in 5 mL of ammonium acetic buffer (0.1 M, pH 9.0) and centrifuged at 14 000 rpm for 20 min, and the supernatant was collected and purified sequentially through Polyvinylpolypyrrolidone (PVPP) column and DEAE Sephadex A225 column. Before HPLC analysis, the elution with 50% aqueous methanol was concentrated by Sep-Pak C18 column (Waters Chromatography). Standard gibberellins were purchased from Fluka Co. (Switzerland), and all solvents and buffers were HPLC-quality. Two biologically independent replicates were performed.

Constitutive expression of miR408 in Arabidopsis implies that it has the potential to act throughout the growth and development, and its importance at the seedling stage is well known (Zhang and Li, 2013). To determine the phenotypic impact of increased miR408 expression during the later development, transgenic plants overexpressing miR408 (MIR408-OX; Zhang and Li, 2013) were examined. The MIR408-OX transgenic plants grew rapidly and were morphologically larger than wild type (Figure 1A). In plants grown in soil for 2 weeks after stratification, the areas of rosette leaves of MIR408-OX transgenic plants were significantly increased compared with the ones of wild type (Figure 1B). The average petiole length from the cotyledon to the tenth leaves of the MIR408-OX transgenic plants also becomes much longer than that of wild type (Figure 1C and Supplementary Figures S1a,b). In addition, both fresh weight and dry weight of the MIR408-OX plants were dramatically increased compared with wild type (Supplementary Figure S1c and Figure 1D). While grown in soil for 6 weeks, the transgenic plants were much taller than wild type (Supplementary Figure S1d). In the plants at the reproductive stage, a comparison of flower size and silique length indicated significant increase for the MIR408-OX transgenic plants (Figure 1E and Supplementary Figure S1e). Furthermore, wild type plants exhibited a darker seed coat color than the transgenic ones (Figure 1F), and the seed yield of the MIR408-OX plants significantly increased (Figure 1G). In summary, the data show that constitutive expression of miR408 in wild type background leads to changes in morphology as well as increases in biomass and seed yield.

Either increased cell number or cell length could lead to larger leaf size (Tsukaya et al., 2002). To investigate whether enhanced leaf area in the MIR408-OX plants was primarily a result of cell division or cell expansion, the lengths of cells in the middle of the first rosette leaf were measured. Microscopic analysis revealed that the epidermal cell size of the transgenic plants was about 165% that of wild type (Figures 2A,B). However, the total number of epidermal cells showed no obvious changes in these two genotypes (Figure 2C). Consistent with the markedly enhanced leaf petiole elongation, the dramatic increase in the epidermal cell length was observed in the transgenic plants (Figures 2D,E) meanwhile the cell number showed no difference between wild type and MIR408-OX transgenic plants (Figure 2F). Together, increased leaf size was in good agreement with the increase in the epidermal cell length, suggesting that enhanced leaf size or leaf petiole elongation in the MIR408-OX plants primarily resulted from an increase in cell expansion rather than cell proliferation.

Previous reports indicated that cytoplasmic streaming, a key determinant of plant size, is generated by organelle associated myosin XI moving (Shimmen and Yokota, 2004; Ojangu et al., 2012; Tominaga et al., 2013). Gene knockout analyses of several myosin XI members lead to growth defects concomitant with reduction in cell size (Prokhnevsky et al., 2008; Peremyslov et al., 2010; Ojangu et al., 2012). Therefore, we examined the expression of nine myosin XI genes, including XI-B, XI-1, XI-G, XI-F, XI-J, XI-H, XI-I, XI-K, and XI-2. Compared to wild type, eight of the tested genes except for XI-J exhibited dramatically increased transcript levels in the MIR408-OX transgenic plants (Figure 3A). Among these myosin members, XI-K and XI-2 transcripts were the most abundant, which is consistent with the conclusion that XI-2 and XI-K are considered the major myosins providing the motive force for cytoplasmic streaming (Peremyslov et al., 2008; Ueda et al., 2010). Thus, it is likely that high level accumulation of myosin in the MIR408-OX transgenic plants might drive cytoplasmic growth, thereby promoting cell expansion and thus leading to the larger morphology.

Previous studies showed that alteration of Gibberellin (GA) metabolism or disruption of the GA signal transduction plays a critical role in regulation of plant dwarfism (Hedden and Phillips, 2000; Monna et al., 2002; Sasaki et al., 2002). Overexpression or repression of genes for enzymes in GA biosynthesis could lead to the alteration of GA levels and thereby result in dwarf or tall phenotypes. Thus, we determined the expression levels of genes involved in GA biosynthesis. As shown in Figure 3B, the transcript levels of biosynthetic genes for GA20-oxdases (GA20OX) and GA₃-oxidases (GA3OX) exhibited significantly induction in the MIR408-OX transgenic plants compared with wild type. On the contrary, transcripts levels of the biosynthetic negative regulators of PIL5 and DAG1 were dramatically decreased in the transgenic plants. Expression level of GA2-oxidase (GA2OX) for GA deactivation did not show obvious difference between the wild type and the MIR408-OX transgenic plants. Moreover, GA contents were examined in the rosette leaves of wild type and miR408 over-expressing plants grown in soil for 2 weeks. The levels of GA₄, the major endogenous bioactive GA in Arabidopsis, were significantly increased in the MIR408-OX plants although that of GA₉, the immediate precursor of GA₄, showed a slight increase (Figure 3C). In addition, the contents of GA₁ as well as its immediate precursor GA₂₀ were dramatically increased in the MIR408-OX plants. Taken together, these observations suggest that accumulation of miR408 might directly or indirectly modulate the cytoplasmic growth and/or GA biosynthesis to promote cell expansion and thus leading to the larger phenotypes.

A key element to increased growth and reproduction is photosynthesis. To investigate the impact of increased expression of miR408 on photosynthesis, the rates were measured using a portable photosynthesis system (CIRAS-1, PP Systems) in the fifth rosette leaves of wild type and MIR408-OX plants grown in soil for 2 weeks. The net photosynthetic rate was significantly increased in the MIR408-OX transgenic plants compared to that of wild type (Figures 4A,B). Consequently, the contents of glucose were found to be significantly increased in the rosette leaves of the transgenic plants (Figure 4C). In the final step of the linear electron transport of photosynthesis, ferredoxin-NADP⁺-oxidoreductase (FNR) catalyzes the reduction of NADP⁺ by ferredoxin (Fd) and provides the reducing power for CO₂ fixation in the Calvin cycle (Carrillo and Ceccarelli, 2003). FNR is supposed to be one of the limiting factors in photosynthetic electron transport, and the amount of FNR has been shown to correlate with photosynthetic activity (Hajirezaei et al., 2002). We found that the transcript levels of genes for leaf-type FNRs (ATLFNR1 and ATLFNR2) showed significant increase in the MIR408-OX plants (Supplementary Figure S2a).

Plastocyanin (PC), which is encoded by two paralogous genes, PETE1 (less abundant) and PETE2 (more abundant) is a copper-binding protein that functions as electron carrier in the thylakoid lumen of the chloroplast (Weigel et al., 2003; Abdel-Ghany, 2009; Pesaresi et al., 2009). Previous studies have indicated that miR408 repress several genes encoding for copper-binding proteins with non-photosynthetic usage (Yamasaki et al., 2007; Abdel-Ghany and Pilon, 2008). It is, thus, plausible that high level of miR408 repress non-photosynthesis related copper-binding proteins and increases copper’s availability for proteins such as PC. To address this question, we first measured copper content in the whole seedlings, and found that there were no differences between wild type and the MIR408-OX plants in overall copper content. The MIR408-OX plants, however, had higher chloroplastic copper than wild type plants (Table 1). Consistent with this observation, transcript level of PAA1 which specifically delivers copper from cytosol to chloroplast (Shikanai et al., 2003; Abdel-Ghany et al., 2005) was significantly increased in the MIR408-OX plants (Supplementary Figure S2b). Thus, elevated miR408 promotes the copper allocation to chloroplast from the cytosol. We next compared the PC levels between wild type and the MIR408-OX transgenic plants because PC is one of the major destinations of chloroplastic copper (Ramshaw et al., 1973). PETE2 was significantly increased in the MIR408-OX plants at both the mRNA and protein levels, though PETE1 was unchanged (Figure 4D and Supplementary Figure S2c). Taken together, these results suggest that over-accumulation of miR408 promotes photosynthesis by modulating the distribution of copper in plant cells.

To gain more insight into the molecular changes associated with miR408 overexpression, total RNA was extracted from whole seedlings of wild type and the MIR408-OX seedlings grown on MS plate for 10 days under continuous light, and subjected to high throughput sequencing. We identified 3,591 differentially expressed genes, of which 2,343 were induced and 1,248 repressed in the MIR408-OX transgenic plants, respectively (Figure 5A and Supplementary Table S1). To further confirm the RNA sequencing data, we performed RT-qPCR to monitor the transcripts levels on a handful of copper-responsive as well as randomly selected genes. As shown in Figure 5B, the results generated by the two methods agreed well. Gene Ontology (GO) analysis revealed that the genes induced in MIR408-OX transgenic plants preferentially associated with GO terms such as response to stimulus and stress, microtubule-based movement, chlorophyll metabolic process, pigment biosynthetic process, and metal ion transport (Figure 5C). Biological pathways responsive to ribosome biogenesis, lipid location, response to oxidative stress, and RNA processing were greatly enriched among the downregulated genes (Figure 5C). Regarding annotated pathways, ribosome, photosynthesis, carbon fixation, pigments biosynthetic processes are among the most significantly enriched terms (Figure 5D). Thus, we hypothesize that these transcriptomic changes may contribute to the photosynthesis, myosin-mediated cytoplasmic growth, abiotic stress responses, and better protection of cells against reactive oxygen species (ROS) especially from the higher photosynthesis of the MIR408-OX transgenic plants. Overall, these results suggest that overexpression of miR408 influences multiple metabolic processes directly or indirectly and thus impacts on multiple aspects of plant growth and development.

Identification and functional analysis of the targets of miRNA is crucial for understanding the biological function of miRNA. To further dissect the function of miR408, two of its target genes, Plastacyanin (ARPN) and LAC13 (Figure 6A), were analyzed. We obtained two T-DNA insertion lines of SALK_091945 for ARPN and SALK_023935 for LAC13 from the Arabidopsis Biological Resource Center (ABRC). Unfortunately, we did not see any visible phenotypes for both mutants under normal growth conditions, which are most likely due to the functional redundancy of the miR408 target genes, especially for those laccase members. Thus, we resorted to the overexpression approach to determine their functions. In order to create ARPN or LAC13 overexpressors, we used the CaMV35S promoter to drive protein expression in Arabidopsis. As shown in Figure 6B, homozygous transgenic plants contained higher transcript levels of ARPN or LAC13 compared with the wild type. Compared to wild type, the ARPN-OX transgenic lines showed shorter siliques and increased silique number of inflorescence (Figures 6C,D). Seed formation was dramatically decreased when siliques were randomly picked for seed counts (Figure 6E) and thus led to less seed yield although the fresh and dry weight showed no obvious differences from the wild type (Table 2). For the LAC13-OX transgenic plants, we did not observe any obvious phenotypic changes at either vegetative or reproductive stage (Table 2). However, the F1 progenies of ARPN-OX and LAC13-OX plants showed much shorter siliques and significantly reduction of seeds number and seed yield although the silique number of inflorescence increased (Figures 6C–E). In addition, the F1 progenies exhibited significant reduction of fresh and dry weight compared with the wild type (Table 2). Meanwhile, the transgenic plants (amiR408) with decreased level of endogenous miR408 generated by the artificial miRNA approach reported in the previous study (Zhang and Li, 2013) showed similar phenotypes to the ARPN-OX/LAC13-OX plants (Supplementary Figure S3 and Table 2), suggesting that miR408 may play a role in control of reproductive development by regulating its target genes. Taken together, these results demonstrate that miR408 appears to have pleotropic effects on plant growth and development.