The mechanisms whereby leaf anlagen undergo proliferative growth and expansion to form wide, flat leaves are unclear. The maize gene NARROWSHEATH1 (NS1) is a WUSCHEL-related homeobox3 (WOX3) homolog expressed at the margins of leaf primordia, and is required for mediolateral outgrowth. To investigate the mechanisms of NS1 function, we used chromatin immunoprecipitation and laser-microdissection RNA-seq of leaf primordial margins to identify gene targets bound and modulated by NS1. Microscopic analyses of cell division and gene expression in expanding leaves, and reverse genetic analyses of homologous NS1 target genes in Arabidopsis, reveal that NS1 controls mediolateral outgrowth by repression of a growth inhibitor and promotion of cell division at primordial leaf margins. Intriguingly, homologous WOX gene products are expressed in stem cell-organizing centers and traffic to adjoining cells to activate stem-cell identity non-autonomously. In contrast, WOX3/NS1 does not traffic, and stimulates cell divisions in the same cells in which it is transcribed.
Plant leaves are typically dorsiventrally flattened and broad, to maximize light capture, gas exchange and photosynthetic efficiency. Leaves develop from the periphery of shoot apical meristems (SAMs), which comprise pools of pluripotent stem cells that give rise to all the above-ground organs of the plant. Leaf primordia are dorsiventrally asymmetrical from their inception (Kaplan, 2001; Caggiano et al., 2017); the dorsal side of the primordium develops adjacent to the SAM, and receives molecular signals that are distinct to those of the ventral side of the newly-emerged leaf. A mechanistic model for leaf outgrowth and flattening, inspired by molecular-genetic analyses of organ development in animals, proposed that the juxtaposition of dorsal and ventral leaf domains at the pre-primordial leaf margin organizes outgrowth along the mediolateral and proximodistal axes to generate wide leaves that project out from the stem (Waites and Hudson, 1995; Diaz-Benjumea and Cohen, 1993; Williams et al., 1994). Although several decades of molecular genetic analyses provide widespread support for this model, the detailed mechanisms whereby plant leaves grow wide remain unclear.
Duplicate mutations in the maize WUSCHEL-related homeobox3 (WOX3) genes NARROWSHEATH1 (NS1) and NARROWSHEATH2 (NS2) cause narrow leaves that fail to expand mediolaterally (Fig. 1A,B; Scanlon et al., 1996). Although the distal-most blade and leaf domains adjacent to the midrib are both intact in mature ns mutant (ns1-R ns2-R) leaves, lateral leaf domains are absent from the proximal blade and the entire length of the sheath. Predicted to encode transcription factors, NS1 and NS2 transcripts and protein accumulate in the margins of leaf primordia, and in the pre-primordial margins of leaf founder cells before they grow out from the SAM (Fig. 1C-F; Fig. S1; Nardmann et al., 2004; Shimizu et al., 2009). These phenotypes suggested a model wherein maize leaves comprise at least two distinct developmental compartments; the central compartment adjacent to the midrib contains blade and sheath domains that are present in both wild-type and ns mutant leaves (green regions in Fig. 1F), whereas the lateral domain requires NS function to grow out from the SAM and expand the leaf mediolaterally (yellow regions in Fig. 1F; Scanlon et al., 1996; Scanlon, 2000). Likewise, mutations in homologous WOX genes in Arabidopsis, Nicotiana, Medicago, Petunia and rice condition similar narrow leaf and lateral organ phenotypes, and their wild-type expression patterns overlap in the margins of incipient and emerged leaf primordia (Fig. S1; Matsumoto and Okada, 2001; Vandenbussche et al., 2009; Tadege et al., 2011; Nakata et al., 2012; Cho et al., 2013). These phenotypic and expression data suggest that specific plant homeobox genes are required for mediolateral and proximodistal outgrowth from the juxtaposed, dorsal and ventral domains at leaf primordial margins (Fig. 1F), as predicted by the Waites and Hudson (1995) model. However, the mechanisms whereby these leaf-specific WOX genes function during leaf initiation and expansion are unknown.
Here, we report the use of chromatin immunoprecipitation sequencing (ChIP-seq) and laser microdissection RNA sequencing (LM-RNA-seq) to identify genes bound and modulated by the maize homeodomain protein NS1/WOX3. Comparative reverse genetic analyses of homologous gene targets, combined with molecular genetic and microscopic examinations of leaf margin development suggest a model whereby plant homeobox genes make leaves grow wide.
NS1/ functions downstream of auxin
The phytohormone auxin is a conserved regulator of leaf initiation; transport-induced auxin maxima in the SAM epidermis correlate with the sites of new primordial outgrowth in diverse plant species (Reinhardt et al., 2000; O'Connor et al., 2014). Accumulation of NS1 transcripts is upregulated more than twofold after application of 0.1 µM auxin (indole acetic acid) to maize seedlings (Fig. 1G); equivalent upregulation is observed following treatment with 0.1 µM cytokinin (kinetin). The Arabidopsis WOX3 ortholog PRS1 is also upregulated ∼twofold by auxin treatment (Caggiano et al., 2017), suggesting that both of these orthologous leaf homeobox genes act downstream of auxin. Moreover, comparisons of wild-type and ns mutant seedlings revealed no changes in transcript accumulation for SPARSE INFLORESCENCE1 (SPI1), the maize homolog of the Arabidopsis auxin biosynthetic gene YUCCA 1 (Fig. S2A-D; Gallavotti et al., 2008; Zhao et al., 2001). In addition, localization of the DR5∼RFP auxin-response reporter and accumulation of PIN1-like auxin transport proteins are equivalent in the margins of ns mutant and wild-type sibling leaf primordia (Fig. S2E-J). Taken together, these data suggest that NS1 functions downstream of auxin biosynthesis, transport, and response.
Identification of gene targets bound and modulated by NS1
An NS1 polyclonal antibody described in Shimizu et al. (2009) identifies leaf homeodomain protein accumulation in approximately three cells at the margins of leaf primordia, and in pre-primordial margins of the incipient leaf primordium before it emerges from the SAM (Fig. 1E; Fig. S1A-F). Chromatin from two-week-old B73 seedlings dissected to contain meristematic and young leaf tissue were used in a ChIP-seq experiment. Comparisons of NS1-targeted genomic sequences versus those bound by the non-specific control antibody found that NS1 bound to a total of 2518 loci found with a q-value cut-off of 0.5, which corresponds to 793 nearest genes. We found that 80.4% of the peaks were within 10 kb of genes, with the furthest being 386,940 bps away (Table S1). Bound sequences were significantly enriched in 5′ untranslated regions (UTRs) (271 peaks or 10.8% FDR 0.0046) and 3′UTRs (311 peaks or 12.4% FDR 0.00017) (Fig. 2A).
To identify genes that are both bound and transcriptionally modulated by NS1, LM-RNA-seq was used to harvest tissue and extract RNA from the marginal tips of the second and third leaf primordia closest to the SAM (i.e. P2 and P3 staged leaves), containing the cells where NS1 transcripts accumulate (Fig. 2B,C). RNA-seq of these microdissected margin cells identified 1144 genes that are differentially expressed (DE) in ns mutant primordial margins (Fig. 2D; Table S2). Union of the 793 genes bound by NS1 and the 1144 transcripts DE in ns margins identified a total of 52 genes that are bound and modulated by the NS1 homeodomain transcription factor. The majority of NS1 bound-and-modulated genes (36/52) were transcriptionally repressed (Table S3); these data are consistent with previous reports that the WUSCHEL-class of WOX transcription factors function predominantly as transcriptional repressors (Lin et al., 2013), although some NS1 transcriptional target genes are indeed activated (Leibfried et al., 2005; Busch et al., 2010; Yadav et al., 2013; Pi et al., 2015). The top 11 genes with the most-enriched ChIP-seq peaks of transcriptionally-repressed NS1 target genes all have peaks within the transcriptional termination site or the last exon (Table S4), and include the predicted transcription factors ETHYLENE RESPONSE FACTOR 7 and ERF DOMAIN PROTEIN 9, HAIRY MERISTEM 1, JASMONATE-ZIM-DOMAIN PROTEIN 1, and two maize paralogs of the Arabidopsis AUXIN RESPONSE FACTOR 2 (ARF2) gene (ARF10 and ARF25; Galli et al., 2018). Intriguingly, ARF2 is previously described as a repressor of lateral organ growth in Arabidopsis; arf2 mutations condition enlarged leaf lamina (Okushima et al., 2005; Schruff et al., 2006).
Analyses of ARF2 homologs in maize and Arabidopsis
ARF10 and ARF25 comprise duplicated maize orthologs of Arabidopsis ARF2 that are significantly bound and modulated by NS1 (Fig. 2E). ARF10 and ARF25 have a peak fold enrichment of 7.77 (9.69E-11) and 8.32 (q-value 4.36E-10), respectively. LM-RNA-seq analyses of the marginal tips of P2-P3 leaf primordia reveal that ARF10 and ARF25 are 2.1- and 2.7-fold higher (adjusted P-value of 2.37E-2 and 1.43E-4, respectively) in ns mutant leaf margins when compared with wild-type siblings (Fig. 2F; Table S2). Owing to 90.7% nucleotide identity among their coding sequences, paralog-specific nucleic acid hybridization probes cannot be constructed. In agreement with previous transcriptomic analyses showing widespread expression of ARF10 and ARF25 during maize ontogeny (Harper et al., 2011), in situ hybridization analyses reveal that ARF10/ARF25 transcripts accumulate throughout the maize seedling shoot (Fig. 3; Knauer et al., 2019). ARF10/ARF25 transcripts are found throughout young leaf primordia, including the husk leaves of axillary meristems, and are enriched in the epidermis and margins of later primordia (Fig. 3A). We note that in situ hybridizations are not inherently quantitative assays; no differences in ARF10/ARF25 hybridization intensity are obvious in ns mutant and wild-type leaf primordia.
Two previous studies reported that arf2 null mutations in Arabidopsis condition enlarged seeds and lateral organs, including leaves, stems and carpels, which is attributed to increased cell division and expansion. These data suggested that ARF2 functions as a pleotropic inhibitor of lateral organ growth in Arabidopsis (Okushima et al., 2005; Schruff et al., 2006). We next exploited this plant model system to determine whether WOX3 and ARF2 interact genetically in Arabidopsis, as predicted by our ChIP-seq analyses of NS1 function in maize. Single mutations in PRS1/WOX3 cause either the complete deletion of the lateral sepals or extreme reductions in lateral sepal width (Fig. S3; Matsumoto and Okada, 2001). No defects in mediolateral development of the leaf lamina or petiole are described in prs1/wox3 mutants, although the lateral stipules are completely deleted from the very base of prs1/wox3 mutant leaves (Fig. 4A-G; Nardmann et al., 2004; Shimizu et al., 2009; Matsumoto and Okada, 2001). All floral and vegetative wild-type phenotypes are restored in prs1/wox3 mutant plants via the introduction of the PRS1-GFP reporter allele driven by the native PRS1 promoter (Fig. 4G; Fig. S3F; Shimizu et al., 2009). A novel arf2 null allele (arf2-12) was generated by CRISPR/Cas9 mutagenesis, which contained an 832 bp deletion at the beginning of the coding sequence (Fig. 4H). In the complemented prs1/wox3; pPRS1-PRS1∼GFP background, arf2-12 mutants show stereotypical arf2 mutant phenotypes including elongated carpels and overgrowth of leaf laminar tissues; lateral stipules are intact at the leaf base (Fig. 4I; Fig. S3G). Moreover, the lateral stipules are restored in arf2-12 prs1/wox3 double mutants (Fig. 4J-L), thereby suppressing the prs/wox3 mutant leaf phenotype. In contrast, no rescue of the pr1s/wox3 lateral sepal development phenotype is observed in arf2-12 prs1/wox3 double mutant flowers (Fig. S3C,H).
NS1 function activates cell division and growth from leaf margins
Genetic and molecular evidence suggests that NS1/WOX3 functions to promote cell proliferation at developing leaf margins (Scanlon et al., 1996; Scanlon, 2000; Nardmann et al., 2004). In a test of this model, incorporation of the thymidine analog 5-ethynyl-2′-deoxyuridine (EdU) was quantified in marginal cells of P3 wild-type and ns mutant leaf primordia, as a molecular marker for entry into the DNA synthesis phase (S-phase) of the cell cycle (Kaiser et al., 2009). As shown in Fig. 5A-C, the three cells at the tips of wild-type P3 leaf margins, where NS1 transcripts and protein accumulate (Fig. 1D,E), enter the cell cycle approximately twice as frequently as cells in ns mutant P3 margins [counting wild-type margin pairs (n=46) and ns margin pairs (n=43) resulted in 1.8-fold higher frequency with a P-value of 0.0001]. Likewise, HISTONE H4 (H4) expression comprises an additional marker for entry into S-phase in plant cells (Bilgin et al., 1999). Of the four ZmH4 paralogs that are differentially expressed in ns mutant margins, our RNA-seq data reveal that they are all significantly downregulated in ns mutant P2/P3 leaf margins (i.e. GRMZM2G073275 1.75-fold, P-value=0.024; GRMZM2G084195 2.02-fold, P-value=0.033; GRMZM2G421279 1.76-fold, P-value=0.022; GRMZM2G072855 2.46-fold, P-value=0.028; Fig. 5D; Table S2). These data support the hypothesis that NS1 function promotes entry into S-phase in the marginal tip cells of maize leaf primordia.
Moreover, transgenic maize plants overexpressing NS1 from the constitutive 35S cauliflower mosaic virus (35S CaMV) promoter can exhibit abnormal proliferative outgrowths at leaf blade margins (Fig. 6A,B). Notably, these ruffled margin phenotypes are not found on all leaves of 35S:NS1 transgenic plants and, when present, form only in the distal domains of the leaf blade that are unaffected by mutations in NS1 and NS2 (Scanlon et al., 1996). Surprisingly, in situ hybridizations of 35S:NS1 transgenic leaves and wild-type siblings did not reveal constitutive accumulation of NS1 transcripts throughout the leaf primordia of NS1 overexpressing plants. In contrast, small patches of ectopic NS1 transcript accumulation are observed in the distal regions of some transgenic leaf primordia; in all observed cases, these ectopic patches of maize NS1/WOX3 homeobox gene expression correlated with abnormal thickening growth and/or elaborative outgrowth (i.e. ruffling) at or near the leaf margins (Fig. 6C-E′).
Lastly, in situ hybridizations were performed using NS1 or ARF25 probes on adjacent 10 µm histological sections of non-phenotypic plants harboring the 35S:NS1 overexpression construct (Fig. 6F,G″). Thus, these samples had wild type-like leaf margin phenotypes and showed no evidence of ectopic expression of NS1, in spite of the fact that they contained the 35S:NS transgene. Comparisons of adjacent sections predominately showed that the extreme marginal tip cells of these leaf primordia exhibit complementary expression of NS1 and ARF10/ARF25. That is, the leaf marginal tips accumulated NS1 transcripts (Fig. 6F′,F″) but were typically free of detectable ARF10/ARF25 mRNA (Fig. 6G′,G″), although ARF10/ARF25 expression is detected in the submarginal regions of these same leaves (Fig. 6G-G″).
An abundance of evidence supports the hypothesis that NS1/WOX3 promotes mediolateral outgrowth of leaves by the activation of cell proliferation at developing leaf margins. This evidence includes: the ns1 mutant phenotype (Fig. 1A,B; Scanlon et al., 1996); accumulation of NS1 at leaf primordial margins (Fig. 1D,E; Nardmann et al., 2004); NS1 repression of the putative growth-repressing Arabidopsis ARF2 homologs ARF10 and ARF25 (Fig. 2G); suppression of the Arabidopsis prs1/wox3 lateral stipule deletion phenotype by the arf2-12 mutation (Fig. 4J-L); elevated entry into S-phase by NS1-expressing leaf marginal cells (Fig. 5); and proliferative outgrowth of maize leaf margins overexpressing NS1 (Fig. 6C-E). Our ChIP-seq analysis suggests that NS1 activates leaf outgrowth indirectly, at least in part via the direct transcriptional repression of maize orthologs of the Arabidopsis growth repressor ARF2. Although activation via ‘repression of a repressor function’ may seem like a non-intuitive mechanistic strategy, this phenomenon is replete with examples from animal and plant development including previously described WOX gene functions (Leibfried et al., 2005; Boyer et al., 2005; Loh et al., 2006; Yadav et al., 2013; Pi et al., 2015). For example, the Arabidopsis homeodomain proteins WOX5 and WUS1 maintain stem cell identity in shoot and root meristems, respectively, by actively repressing transcription of target genes that promote differentiation programs. Our data reveal that, like WUS1 (Busch et al., 2010; Yadav et al., 2013), NS1/WOX3 functions as both a repressor and activator of target gene expression, in contrast with previous reports that WUS-class WOX genes function solely as transcriptional repressors (Lin et al., 2013). The rescue of the prs1/wox3 lateral stipule deletion phenotype in arf2 prs1 double mutants suggests that a genetic interaction between WOX3 and ARF2 identified herein is conserved in Arabidopsis and maize, although we currently have no evidence that the PRS1/WOX3 transcription factor directly represses ARF2 transcription. However, the prs1/wox3 lateral sepal deletion phenotype is not suppressed by the arf2 mutation (Fig. S3), suggesting that WOX3 function in Arabidopsis involves direct regulation of some additional gene target(s) aside from Arabidopsis ARF2. Reverse genetic analyses, in both maize and Arabidopsis, of additional NS1/WOX3 targets identified in this study (Table S3) will enable in-depth analyses of conserved and non-conserved WOX3 function in these model angiosperms.
Previous work in Arabidopsis reported that PRS1/WOX3 (as well as the leaf homeobox gene WOX1, which has no ortholog in maize) is transcriptionally activated by the adaxially-localized protein ARF5, and is repressed by the abaxial transcription factors ARF3, ARF4 and ARF2 (Guan et al., 2017). In this way, PRS1/WOX3 and WOX1 accumulation is localized to the leaf primordial margin, at the juxtaposition of adaxial and abaxial leaf domains. Although it is not known whether PRS1/WOX3 repression of ARF2 expression is conserved in Arabidopsis as well as in maize, arf2 mutations suppress the prs1 mutant leaf phenotype (Fig. 4J-L), revealing a genetic interaction. Future ChIP-seq analyses of PRS1/WOX3 will determine whether ARF2 and PRS1 are indeed mutually-repressive.
NS1 is transcriptionally induced by both indole acetic acid (IAA) and kinetin (Fig. 1G), suggesting that NS1 function is downstream of auxin and cytokinin. We propose that NS1 is involved in lateral organ outgrowth, which is likewise associated with auxin (Reinhardt et al., 2000), and that this leaf outgrowth is mediated by cytokinin-activated cell divisions of organ initial cells at primordial leaf margins (Fig. 5). Furthermore, auxin accumulation during leaf initiation causes downregulation of KNOTTED1-like HOMEOBOX (KNOX) genes in leaf founder cells (Scanlon, 2003; Hay et al., 2006), although the detailed mechanism is unknown. Intriguingly, KNOX gene downregulation is incomplete in ns mutant SAMs, which correlates with the failure to elaborate lateral leaf domains (Scanlon et al., 1996). Although KNOX downregulation is disrupted in ns mutants, our data reveal that transcript accumulation of the maize auxin biosynthetic gene SPI1, accumulation of PIN1c auxin transport protein and localization of the DR5 auxin response reporter are not disrupted in ns mutant shoot apices (Fig. S2). These data implicate loss of NS1 function, and not defects in auxin biology per se, as responsible for the altered KNOX downregulation in ns mutant SAMs. Moreover, transcription of NS1 is activated by auxin (Fig. 1G), although our ChIP-seq data suggest that KNOX genes are not targeted by the NS1 transcription factor (Table S1). Taken together, these data suggest that NS1 acts downstream of auxin in a network to downregulate KNOX accumulation in maize founder cells, although the role of NS1 during KNOX downregulation is indirect.
We note that the restriction of NS1 ectopic overexpression to relatively infrequent small patches of transcript accumulation when driven by the constitutive 35S CMV promoter (Fig. 6D,E) reveals that maize has evolved an extraordinarily robust mechanism to confine NS1 gene expression to the marginal tips of leaf primordia. AUXIN RESPONSE FACTORS and WOX3 homologs in maize and Arabidopsis are upregulated by auxin (Fig. 1G; Caggiano et al., 2017; Galli et al., 2018; Ori, 2019). However, whereas ARF10 and ARF25 transcripts accumulate broadly in maize primordia (Fig. 3) but are reduced within the edges of leaf margins (Fig. 6G), NS1 expression is limited to a few cells at the marginal leaf tips (Fig. 1D; Fig. 6C,F; Nardmann et al., 2004). These data suggest a model wherein auxin induces expression of ARF10, ARF25 and NS1 in maize leaf primordia. Thereafter, NS1 accumulation is restricted to the marginal tip cells by some unknown factor(s), and represses the expression of growth-inhibitory ARF2 homologs in these same NS1-expressing cells at the leaf tip. In this way, auxin-induced NS1/WOX3 function promotes mediolateral expansion from the leaf margin. The well-studied WOX homeodomain proteins WUS and WOX5 traffic from the stem-cell organizing centers that express their corresponding mRNAs, to specify stem-cell identity in neighboring cells of the shoot and root meristem, respectively (Pi et al., 2015; Yadav et al., 2011). In contrast, our previous studies showed that PRS1/WOX3 does not traffic (Shimizu et al., 2009); this current study suggests that the leaf homeobox gene NS1/WOX3 has evolved to activate cell division and proliferative growth in the same leaf margin cells in which it is expressed.
MATERIALS AND METHODS
Genetic stocks and plant growth
Maize stocks segregating for the narrow sheath mutant phenotype were obtained from the ns 1:1 line as previously described (Scanlon et al., 1996); phenotypically wild-type plants from this line are heterozygous for the ns1 mutation and homozygous for the ns2 mutation (genotype NS1/ns1-R ns2-R/ns2-R), whereas ns mutant plants are homozygous for both ns1-R and ns2-R. To generate the 35S:NS1 overexpression lines, the NS1 (GRMZM2G069028) coding sequence was cloned into the entry vector pENTR/D (Thermo Fisher Scientific), and then integrated into the binary vector pB7FG2 (Karimi et al., 2002) behind the 35S CaMV promotor via the Gateway® System. The pB7FG2 binary vector also harbors the bar gene, which allowed for selection of transgenics using the herbicide Basta. The transformation into maize hybrid Hi-II was performed at the Plant Transformation Facility at Iowa State University (Ames, IA, USA). 35S:NS1 plants were outcrossed to inbred line B73 three times before use. All maize plants were grown in the Cornell Guterman Greenhouse (conditions: 29.4°C day/23.9°C night; 16 h light/8 h dark; soil type: 1:1 Turface MVP; PROFILE Products).
Arabidopsis seeds segregating for the prs1-1 mutation in the Landsberg erecta (Ler) ecotype were kindly supplied by K. Okada (Matsumoto and Okada, 2001). The prs1 PRS1-GFP rescue line was created by transforming prs1-1 plants with the GFP vector pMDC107 carrying the PRS1 coding region (AT2G28610) and 3 kb upstream sequences cloned from Arabidopsis genomic DNA. The arf2-12 mutant allele was produced via CRISPR/Cas9 mutagenesis as previously described (Pauwels et al., 2018). WT Ler, prs1-1 and prs1-1 PRS1-GFP plants were transformed via Agrobacteria (GV3101)-mediated floral dip with the dual sgRNA/Cas9 vector pMR333 obtained from M. Ron (University of California, Davis, CA, USA). Two sgRNAs were designed on CRISPOR (Haeussler et al., 2016) toward the 5′ end of the ARF2 coding sequence (AT5G62000). The specific sequences were: Protospacer 1 F/R: 5′-ATTGTTTCAATGAAAGGTAATCG/AAACCGATTACCTTTCATTGAAA-3′; Protospacer 2 F/R: 5′-ATTGAATGCACCTGGAACCTCGG/AAACCCGAGGTTCCAGGTGCATT-3′.
BASTA-selected T1 plants were PCR screened using the following primer sequences to identify an 832 bp deletion between the two sgRNA sites, visible via gel electrophoresis. Gene-specific primers used were: 5′-TGGACTACCGAAGCGAGTTT-3′; 5′-TGTGTCGGATGCAGTCAAGG-3′.
The T-DNA insertion in pMR333 also contains a pOLE-OLE∼GFP (AT4G25140) marker to identify plants carrying the T-DNA insertion by fluorescent seed coat. Seeds from selected T1 plants were thereby screened for the absence of GFP-fluorescence in the seed coat, as a way to select against lines harboring the CAS9 construct and avoid additional CAS9-mediated mutational activity in the T2 generations and beyond. Lines were progressed until at least the T3 generation and homozygosity was confirmed through PCR and sequencing. All Arabidopsis plants were grown in LM111 media (Lambert Peat Moss) under standard long-day conditions (light: 16 h day, 100 µmol; Temperature: 22°C; Humidity: 50%) at the Cornell Agricultural Experiment Station in prototype 45-square foot step-in growth chambers.
Scanning electron microscopy
Arabidopsis cauline leaf axes and flowers were dissected fresh and then flash frozen in slushed liquid nitrogen for scanning electron cryomicroscopy (CryoSEM). Samples were run on a FEI Strata 400S DualBeam Focused Ion Beam scanning electron microscope (FIB/SEM) fitted with a Quorum PP3010T CryoSEM/FIB preparation system. Frozen samples were loaded into a vacuum and briefly sublimated (∼2 min at −80°C followed by 2 min at −70°C) to remove crystalline ice contamination from the transfer process before being sputter coated with gold palladium at 20 mA for 30 s.
ChIP was carried out as previously described (Song et al., 2016) with the following modifications. Approximately 50 meristems, including the P2 and P3 from two-week-old B73 seedlings, were used for each of two biological replicates per antibody; 1 µg anti-NARROW SHEATH 1 rabbit polyclonal antibody (Shimizu et al., 2009, 1/350) was used in the treatment and 1 µg non-specific rabbit IgG was used as a negative control (MAGnify Chromatin Immunoprecipitation System, 49-2024, 1/350). Chromatin was sonicated using the Covaris focused-ultrasonicator. Chromatin was precleared by incubating with anti-rabbit antibody Dynabeads™ before precipitating with Dynabeads incubated with anti-NS1 treatment antibody or nonspecific rabbit IgG negative control. Libraries were prepared using KAPA Hyper Prep Kit (Hoffmann-LaRoche, KK8501) and sequenced on HiSEQ 2500 Rapid Run 2×100 RR Paired End system (Illumina). Reads were aligned with BWA mem settings -M -t aligned to AGPv3. Peaks were called with Macs2 with the following settings: macs2 call peak -t -f BAMPE -g 2060056721 -n -B --call-summits. PAVIS was used to calculate peak distribution across the genome.
RNA in situ hybridizations, immunohistolocalizations and qRT-PCR
Shoot apices from greenhouse-grown two-week-old seedlings were fixed overnight at 4°C in FAA (3.7% formalin, 5% glacial acetic acid and 50% ethanol in water). Tissues were dehydrated at 4°C through a graded ethanol series (50%, 70%, 85%, 95%, 100%) for 1 h each, with three changes in 100% ethanol, and kept in 100% ethanol at 4°C overnight. Tissues were then passed through a graded Histo-Clear (National Diagnostics) series (3:1, 1:1, 1:3 ethanol: Histo-Clear) with three changes in 100% Histo-Clear; all changes were 1 h each at room temperature. Samples were then embedded in Paraplast®Plus (McCormick Scientific), sectioned and hybridized using antisense digoxygenin-labeled RNA probes as previously described (Johnston et al., 2014a).
Hybridization probes for NS1 (GRMZM2G069028) were prepared as previously described (Nardmann et al., 2004). Gene-specific primers were used to prepare 761 bp-long in situ hybridization probes for ARF25 (GRMZM2G116557): 5′-GATGACAGTCGTCACCGTCT-3′; 5′-TTAGGAACCAAACCACCAGG-3′.
Immunolocalizations were carried out as previously described (Boutte et al., 2006; Lee et al., 2009) using an Arabidopsis PIN1 (gift from J. Traas, ENS de Lyon, France) antiserum diluted 1:300 or a 1:350 dilution of affinity-purified rabbit anti-NS1 antiserum (Shimizu et al., 2009), and the Alexa Fluor 488-conjugated secondary antibody (Life-Technologies, 1/500).
For qRT-PCR analyses of gene-specific auxin and cytokinin responses, root excised 14-day-old B73 maize seedlings were incubated with IAA or kinetin, which was first dissolved in 1 M KOH then diluted to a working concentration of 0.1 μM at pH 5.8. Control samples were cultured in soil treated with water containing an equimolar concentration of KOH. Gene-specific primers (below) were designed for use with SYBR-Green (Quanta) in qRT-PCR as previously described (Zhang et al., 2007). Three biological replicates were examined; data are presented using the 2−ΔΔct method (Livak and Schmittgen, 2001) with threshold values normalized to accumulation of each transcript after control treatment as described using Bio-Rad iQ5 Version 1.0 software (Zhang et al., 2007). The gene-specific primers used were: ZmNS1-GRMZM2G069028–5′-ATGGAGGTGGAGCTGGGTTA-3′, 5′-CACAGATCAGTGCTCCATTGCATCTGTG-3′; ZmHK1-GRMZM2G069028–5′-GGCTCGACAACTGCCGAGTAC-3′, 5′-GTCGTTCCCACTACCAATCTGGAG-3′; ZmARF5-GRMZM2G035405–5′-GCTATCACGAGCTCCGTAGG-3′, 5′-CGGTCGACGAATACAAGCTG-3′; ZmRR7-GRMZM2G096171–5′-CTCGCACTACTTCCAGTTCCTCCTC-3′, 5′-GACGGAGCCATTGGACCATCTG-3′.
Laser microdissection, library preparation and sequencing
Two-week-old seedlings of ns1 mutants and wild-type seedlings from the ns 1:1 line (described above) were dissected and fixed and embedded for LM-RNA-seq) as previously described (Scanlon et al., 2009). The marginal tips of P2 and P3 leaves were targeted for microdissection from 10 µm transverse serial sections (Fig. 2D,E) based on the localization of NARROWSHEATH1 protein in immunohistological sections (Fig. 1E; Fig. S1) using the Positioning and Ablation with Laser Microbeams system (PALM; Microlaser Technologies). RNA extraction was performed according to the manufacturer's instructions using Arcturus™ PicoPure™ RNA Isolation Kit and RNA amplification using the Arcturus™ RiboAmp™ HS PLUS Kit (Thermo Fisher Scientific). Libraries were prepared using the NEBNext UltraTM RNA Library Prep Kit for Illumina, and sequenced using NextSeq 500 75 Single End.
RNA-seq alignment, counting and normalization
Illumina adapter sequences were trimmed using Trimmomatic v0.39. Reads were aligned to B73 genome RefGen V3 with HiSAT2 (Kim et al., 2015) and counted with HTSeq (Anders et al., 2015). Raw counts were normalized using R package edgeR v3.20.9. Differential expression was calculated using R package DESeq2. The raw ChIP-seq data and RNA-seq data are available at the NCBI Bioproject number PRJNA633509.
We thank K. A. Spoth for expert technical assistance with CryoSEM, and S. Leiboff and D. Henderson for the images shown in Fig. 1C,D, respectively. We thank J. Cammarata and J. Satterlee for comments on the manuscript. We thank J. Traas for the use of the Arabidopsis PIN1 antibody. Thanks to A. Corella and E. Palmer for care of our plants in growth chambers. The CRYO-FIB/SEM work made use of the Cornell Center for Materials Research Shared Facilities which are supported through the National Science Foundation MRSEC program (DMR-1719875). Additional support for the FIB/SEM cryo-stage and transfer system was provided by the Kavli Institute at Cornell and the Energy Materials Center at Cornell, DOE EFRC BES (DE-SC0001086).
Conceptualization: M.J.S., P.A.C., R.J., B.R.C.; Methodology: P.A.C., R.J., R.S., B.R.C.; Validation: P.A.C.; Formal analysis: P.A.C., R.J., R.S.; Investigation: M.J.S., P.A.C., R.J., R.S., B.R.C.; Data curation: P.A.C.; Writing - original draft: M.J.S.; Writing - review & editing: P.A.C., R.J., R.S., B.R.C.; Visualization: P.A.C.; Supervision: M.J.S.; Project administration: M.J.S.; Funding acquisition: M.J.S., P.A.C.
This work was supported by National Science Foundation grants 1238142 to M.J.S. and 1612235 to P.A.C. Deposited in PMC for immediate release.
The ChIP-seq data and raw RNA-seq data are available at the NCBI Bioproject under accession number PRJNA633509.
The authors declare no competing or financial interests.