Ontogeny of the Maize Shoot Apical Meristem

Laser Microdissection and RNA-Seq of SAM Ontogeny

Laser microdissection enables the isolation of discrete domains within microscopic samples for use in transcriptomic analyses (Nelson et al., 2006; Scanlon et al., 2009). Six samples were microdissected from developing embryos and 14-d-old seedlings, including (1) the cells comprising the embryo proper of the proembryo (Figures 1M and 1S), (2) the organizing SAM and emerging scutellar hood of the transition-stage embryo (Figures 1N and 1T), (3) the SAM and the initiating coleoptile of the coleoptile-stage embryo (Figures 1O and 1U), (4) the SAM and leaf primordium of a stage 1 embryo (Figures 1P and 1V), (5) the SAM and plastochron 1 leaf of a 14-d-old seedling (L14; Figures 1Q and 1W), and (6) the lateral meristem and the newly initiated husk leaf from the 14-d-old seedling (Figures 1R and 1X). Two biological replicates were obtained per sample; replicates comprised cells from three to eight separate embryos or seedlings (see Supplemental Table 2 online). Total RNA isolated from the microdissected cells was subjected to two rounds of linear amplification to generate microgram quantities of RNA amenable to RNA-seq analyses (reviewed in Brooks et al., 2009). Amplified RNA was used to construct cDNA libraries, and Illumina-based RNA-seq generated a total of 130 million 44-bp sequence reads that were aligned to the maize genome (see Methods; Schnable et al., 2009). A summary of the total number of reads per biological replicate and their alignment to the maize gene space is presented in Supplemental Table 2 online. Read counts for each transcript were normalized to account for variation in library size across the 12 samples and are reported as reads mapped per million mapped sequences (RPM) (see Methods). Note that the RNA amplification step, which uses the polyA tail of the mRNA, excludes noncoding RNAs from the data set. Pairwise comparisons between each of the six samples and a comparison across the entire sample set were performed to identify differentially accumulated transcripts, taking into account false discovery rates (FDRs; see Methods). For all comparisons, an adjusted P value (q value ≤ 0.05) was used to identify upregulated transcripts.

Transcripts from a total of 20,610 genes are represented in the combined RNA-seq data sets (Figure 2A). There is little variation in the total number of genes represented among the six stages; the transition stage comprised the largest number of individual gene transcripts (16,454), whereas the coleoptile stage had the fewest (15,952). Oscillations in gene expression, wherein certain genes are used reiteratively during development, have been demonstrated to impose some limitations on the use of transcriptomic analyses of developmental ontogeny (Brady et al., 2007; Moreno-Risueno et al., 2010). Indeed, six-way comparisons of the transcriptional data sets identified 12,135 genes that are common to all samples (Figure 2A). Previous studies in maize and Arabidopsis thaliana guided the selection of candidate genes predicted to function during shoot development, which were surveyed for transcriptional patterns across the six samples. The 226 selected candidate genes were dispersed into seven developmental categories, including stem cell maintenance, lateral organ initiation, dorsiventral patterning, chromatin structure and remodeling, hormonal signaling, apical-basal patterning, and cell division and growth. A heat map was generated based on the relative transcript accumulation of each candidate gene during the six SAM developmental stages, revealing dynamic changes in transcriptional pattern before and after formation of the meristem, after the nascent meristem becomes functionally mature, during elaboration of three distinct lateral organ types (i.e., the scutellum, the coleoptile, and the leaf), and during initiation of juvenile, adult, and husk leaves (Figure 2B) (see Supplemental Data Set 1 online). These summary data confirm that the combined strategies of laser microdissection and RNA-seq can successfully detect transcriptional differences on a global scale.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Genes Represented in Each of the Six RNA-Seq Data Sets.

(A) All genes represented in each of the six samples (RPM ≥ 1).

(B) Heat map based on the relative transcript accumulation of 226 candidate genes sorted according to implicated function. 1, L1 stage; 14, L14 stage; C, coleoptile stage; L, lateral meristem; P, proembryo; T, transition phase.

Establishing a Meristem

To identify SAM-specific gene transcripts, transcript accumulation in the premeristematic proembryo was compared with all samples microdissected after SAM formation (i.e., the transition stage, the coleoptile stage, the L1 stage, the L14 stage, and lateral meristems). A total of 4145 gene transcripts were either upregulated or present/absent in comparisons of the proembryo with meristem-containing samples (see Methods; Figure 3A; see Supplemental Data Set 2 online). Of the 4145 gene transcripts represented, 284 are present in the proembryo and absent from samples containing a meristem, whereas 742 are unique to samples containing a meristem (Figure 3A). These 742 meristem-specific genes were distributed into MapMan functional categories, and more than one-half of the predicted gene products fell into three categories: transcriptional regulation (123 genes), genes that function in transport (103 genes), and genes of unknown function (271 genes) (Figure 3C; see Supplemental Data Set 3 online) (Thimm et al., 2004). Nineteen genes implicated in transcriptional regulation were exclusively detected in the premeristematic proembryo, whereas 123 were found only in samples containing a meristem (Figure 3B). These 123 meristematic genes that function in transcriptional regulation were sorted into gene families of known predicted function (Figure 3D; see Supplemental Data Set 4 online).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Establishing a Meristem.

(A) Venn diagram of all 4145 transcripts upregulated or present/absent in the premeristematic proembryo compared with the five samples with meristems. Transcripts in the overlapping region of the Venn diagram are present before and after the meristem forms, whereas transcripts in the nonoverlapping regions are present before but not after meristem formation or vice versa.

(B) The 418 out of the 4145 transcripts that are implicated in transcriptional regulation are presented in a Venn diagram.

(C) MapMan bin distribution of the 742 transcripts that are differentially accumulated after the meristem forms. CHO, carbohydrate.

(D) The 123 transcripts implicated in transcriptional regulation present after the meristem forms sorted into their respective gene product families. In situ hybridizations using ZYB16 ([E] to [H], [J], and [L]) and KN1 ([I] and [K]) probes. Arrowheads point to transcript accumulation. col, coleoptile stage; pro, proembryo; tr, transition phase.

Bars = 100 μm.

Transcription factors implicated in stem cell maintenance include six CLASS I KNOTTED1-LIKE HOMEOBOX (KNOX) genes and three BEL-LIKE (BELL) HOMEOBOX genes; CLASS I KNOX and BELL proteins form heterodimers that regulate stem cell identity (see Supplemental Data Set 3 online) (Reiser et al., 2000; Bellaoui et al., 2001; Müller et al., 2001; Smith et al., 2002; Mukherjee et al., 2009; reviewed in Hay and Tsiantis, 2010). At least two gene transcripts encoding WUSCHEL-RELATED HOMEODOMAIN (WOX) proteins (WOX2A and WOX9B) were upregulated in the proembryo. In Arabidopsis, WOX2 and WOX9 (also known as STIMPY) are also expressed before SAM initiation and specify the apical-basal axis of the embryo (Haecker et al., 2004; Wu et al., 2007; Breuninger et al., 2008). Previous reports in maize revealed that WOX2A transcripts accumulate in the apical cap of the proembryo before localizing to the germinal face of the transition-stage embryo (Nardmann et al., 2007). The accumulation of WOX2A and WOX9B transcripts in the premeristematic maize proembryo is consistent with a conserved function of these WOX putative paralogs in establishing apical-basal domains in the maize embryo before SAM initiation.

Transcripts encoding transcription factors that are known to function during organogenesis and that are not detected in the proembryo include members of the YABBY (six), GROWTH-REGULATING FACTOR (GRF, four), and NO APICAL MERISTEM/CUP-SHAPED COTYLEDON (four) gene families (the numbers in parentheses represent the number of transcripts in that gene family) as well as two additional WOX transcripts (Aida et al., 1999; Kim et al., 2003; Vroemen et al., 2003; Juarez et al., 2004; Zimmermann and Werr, 2005; Nardmann et al., 2007; Zhang et al., 2008; Sarojam et al., 2010; Santa-Catarina et al., 2012). Previous studies in Arabidopsis and maize demonstrated that YABBY genes are expressed in leaf and floral anlagen and in the marginal tips of leaf and floral primordia (Siegfried et al., 1999; Juarez et al., 2004). Consistent with these previous findings, transcript accumulation of Zea mays YABBY14 (ZYB14), Zea mays YABBY9 (ZYB9), and Zea mays YABBY10 (ZYB10) was upregulated in samples containing a meristem as compared with the proembryo stage (Juarez et al., 2004). Furthermore, transcripts of the Arabidopsis FILAMENTOUS FLOWER (FIL; also known as ABNORMAL FLORAL ORGAN and YABBY1) and YABBY3 (YAB3) accumulate in late globular-stage and heart-stage embryos, indicating that these putative YABBY paralogs also mark the initiating cotyledons (Siegfried et al., 1999). Transcripts of another YABBY homolog, which we have designated Zea mays YABBY16 (ZYB16), accumulate in the developing scutellum at the apex of the maize proembryo (Figure 3E). In addition, ZYB16 transcripts are retained in the scutellar tip, in the initiating coleoptile, and in the anlagen of initiating leaves at later stages in embryogenesis (Figures 3F to 3H). These data suggest that the scutellum displays a pattern of ZYB16 transcript equivalent to that of other lateral organs in the maize shoot, which is consistent with models ascribing leaf homology to the scutellum (Kaplan, 1996). Evidence from additional molecular markers, described below, lends further evidence to the hypothesis that the scutellum is a leaf-like lateral organ.

Moreover, in situ hybridizations performed using KN1 and ZYB16 probes on consecutive embryo sections of both proembryo (Figures 3I and 3J) and transition-stage samples (Figures 3K and 3L) confirm that the onset of ZYB16 transcript accumulation precedes that of KN1. By contrast, control hybridizations reveal that both KN1 and ZYB14 transcripts first accumulate during the transition stage (see Supplemental Figures 1A to 1D online). Taken together, these data suggest that the dual functions of the maize SAM are not simultaneously activated during SAM ontogeny; lateral organ initiation, as indicated by accumulation of ZYB16 transcripts, precedes stem cell maintenance in the maize SAM. These data extend and support the conclusions from previous analyses demonstrating that a maize mutant defective in meristem maintenance is capable of forming a scutellum and coleoptile, plus a few small leaves (Vollbrecht et al., 2000).

Morphological and Molecular Maturation of the Maize SAM

Significant morphological differences are recognized in a newly formed transition-stage shoot meristem and a mature, L14-stage SAM that has initiated multiple foliar leaves. At the transition stage, the meristem is flattened and composed of ∼150 cells organized in longitudinal files at least two cell layers deep, which correlates with the transcript accumulation of the stem cell marker KN1 in these embryos (Figures 1B and 1H) (Poethig et al., 1986; Smith et al., 1992; Jackson et al., 1994). Lateral organs are not initiated from the early transition-stage SAM; the scutellum forms during the proembryo stage (described previously), whereas the coleoptile does not initiate until the late transition/coleoptilar stage (Abbe and Stein, 1954). By contrast, the L14-stage SAM attains a postlike architecture comprising multiple cell layers organized into distinct functional zones and is actively initiating leaves (Figures 1E and 1K). To investigate the transcriptional changes that correlate with these morphological differences during SAM maturation, the transcriptome of the L14-stage SAM was compared with that of the newly formed transition-stage SAM. This comparison identified 1706 transcripts that were upregulated or present/absent during these developmental stages (Figure 4A; see Supplemental Data Set 5 online). Transcripts that fell into metabolic and regulatory MapMan functional categories are displayed in Supplemental Figures 2 and 3 online (see Supplemental Data Sets 6 and 7 online). A Fisher’s exact test revealed that transcripts implicated to function in protein and mitochondrial electron transport/ATP synthesis are enriched in newly initiated SAMs, whereas in the L14 stage, SAM transcripts that function in hormone metabolism, transport, and transcriptional regulation are upregulated (see Methods; Figure 4B) (Thimm et al., 2004).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Transcriptomic Comparisons of a Newly Formed Meristem to a Mature Meristem.

(A) Venn diagram of the 1706 transcripts upregulated or present/absent when comparing the transition stage to the L14 stage. Transcripts in the overlapping region of the Venn diagram are present in both newly formed and mature meristems, whereas transcripts in the nonoverlapping regions are present before but not after meristem maturation or vice versa.

(B) The 1706 upregulated transcripts sorted into modified MapMan categories. Total number of transcripts in each functional category is displayed in the bars. Darker bars in (B) and (D) designate the transition stage, lighter bars indicate the L14 stage. Asterisk after transcript number indicates enrichment of that functional category using Fisher’s exact test (P ≤ 0.01; FDR, 6%). CHO, carbohydrate.

(C) The 192 out of 1706 transcripts upregulated or present/absent that are implicated in transcriptional regulation in a newly formed meristem and a mature meristem.

(D) The 192 transcripts are distributed into their respective gene product family.

Enrichment of protein synthesis in transition-stage SAMs is consistent with previous findings that enhanced ribosome function is required to establish a shoot meristem (Weijers et al., 2001; Tzafrir et al., 2004; reviewed in Byrne, 2009). Transcripts encoding the maize RIBOSOMAL PROTEIN S5A and S5B (RPS5A, also known as Arabidopsis MINUTE-LIKE1 [AML1]; RPS5B) are both upregulated in transition-stage embryos compared with the L14-stage SAMs. Development of aml1 mutant embryos is arrested at the globular stage in Arabidopsis before the meristem is established (Weijers et al., 2001). In addition, as compared with the L14-stage SAM, the transition-stage SAM contains more upregulated gene transcripts that are implicated in protein targeting (14:5), amino acid synthesis (19:9), and DNA synthesis and repair (24:11). These data suggest that an increase in ATP production and in DNA and protein metabolic activity immediately precedes the proliferative growth displayed during later stages of SAM ontogeny.

The dual functions of stem cell maintenance and lateral organ initiation occur concomitantly in mature L14-stage SAMs, as is reflected in the predicted functions of 109 transcripts implicated in transcriptional regulation that are upregulated at this stage. Among these are four transcripts encoding known stem cell regulators, including three CLASS I KNOX genes and a GATA domain protein homologous to HANABA TARANU in Arabidopsis (Figures 4C and 4D; see Supplemental Data Set 8 online) (Reiser et al., 2000; Zhao et al., 2004; Mukherjee et al., 2009; reviewed in Hay and Tsiantis, 2010). Upregulated transcripts encoding proteins that function in organogenesis and cell proliferation of lateral organs include WUSHEL-RELATED HOMEOBOX3A (WOX3A), the YABBY gene DROOPING LEAF2 (DL2), the lateral organ boundary gene INDETERMINATE GAMETOPHYTE1 (IG1), and two GRF-LIKE transcripts (Kim et al., 2003; Nardmann et al., 2004; Yamaguchi et al., 2004; Evans, 2007; Nardmann et al., 2007; Zhang et al., 2007; Zhang et al., 2008). In addition to three AUXIN RESPONSE FACTOR (ARF) class transcription factors, several transcripts that encode hormone biosynthetic proteins are also upregulated in stage 14 SAMs, including seven gibberellin metabolic transcripts implicated in lateral organ growth and development. Transcripts involved in ethylene and jasmonate signaling and regulation also accumulate differentially in mature, stage 14 SAMs (see Supplemental Figure 3 and Supplemental Data Set 7 online). These data suggest that, compared with the transition stage, the stage 14 meristem contains a wider diversity of transcripts implicated in both lateral organ initiation and stem cell maintenance.

Elaboration of the First Three Lateral Organs

Three different types of lateral organs are elaborated in maize embryos: the scutellum, the coleoptile, and five to six foliar leaves. A three-way comparison was made between the transition stage, the coleoptile stage, and the L1-staged shoot apex to determine the transcriptional differences correlated with the development of these morphologically distinct organs. A total of 532 transcripts were upregulated in the three-way comparison. Of these upregulated transcripts, 517 were sorted into three main clusters based on relative transcript abundance during the three embryonic stages. These include transcripts upregulated in the transition stage versus both the coleoptile and L1 stages (Figure 5A, cluster 1), transcripts upregulated in the transition stage versus the L1 stage (Figure 5A, cluster 2), and transcripts upregulated in the L1 stage versus the transition stage (Figure 5A, cluster 3) (see Supplemental Figure 4 and Supplemental Data Set 9 online). The 517 genes from the respective clusters were distributed into 25 MapMan annotated functional categories and were displayed in metabolic and regulatory pathways (Figure 5A, see Supplemental Figures 5 and 6 and Supplemental Data Sets 10 and 11 online) (Thimm et al., 2004). A total of 63 transcripts implicated in transcriptional regulation were present in the three main clusters, many of which have specific functions in embryo development, including but not limited to apical-basal patterning and cotyledon development (Figure 5B; see Supplemental Data Set 12 online).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Elaboration of the Scutellum, the Coleoptile, and the First Foliar Leaf.

(A) Three main clusters that comprise 517 of 532 genes upregulated in the three-way comparison of the transition stage (T), coleoptile stage (C), and L1 stage (L1) embryos. Distribution of genes into functional categories according to MapMan. CHO, carbohydrate; TCA, tricarboxylic acid.

(B) The 63 genes that are implicated in transcriptional regulation from the three main clusters distributed into gene families. The transition stage is shaded gray, the coleoptile stage is shaded black, and the L1 stage is shaded white. In situ hybridizations illustrate the accumulation of LEC1 transcripts during embryogenesis ([C] to [G]) and in 14-d-old seedlings ([H] and [I]). col, coleoptile stage; L1, leaf 1 stage; L14; 14-d-old seedling SAM; LM, lateral meristem; pro, proembryo; tr, transition phase.

Bars = 100 μm.

Several transcripts encoding transcription factors implicated in apical-basal patterning of the embryo were upregulated during the transition stage, including WOX2A, WOX5B, and NUTCRACKER (NUC; Haecker et al., 2004; Levesque et al., 2006; Nardmann et al., 2007). Whereas WOX2 functions in the apical embryo domain, WOX5 and NUC homologs accumulate in the basal regions of developing embryos, where they eventually specify distinct root domains (Haecker et al., 2004; Levesque et al., 2006; Nardmann et al., 2007). These root-specific transcription factors were probably detected in our transition-stage samples because these microdissected domains included both the apical and basal regions of the embryo (Figures 1N and 1T), whereas only shoot domains were isolated from later stages. Upregulated during development of both the scutellum and the coleoptile, the HEME ACTIVATOR PROTEIN3 (HAP3) class transcription factor Zea mays LEAFY COTYLDEON1 (LEC1) functions during lipid metabolism in maize, whereas homologs in Arabidopsis also specify cotyledon development (Mu et al., 2008; Shen et al., 2010). Fatty acid biosynthesis is especially proliferative during embryogenesis, when oil bodies are stored as energy reserves that will subsequently be used during germination. Zm-LEC1 is expressed throughout the apical cap in the proembryo and early transition stage (Figures 5C and 5D) (Zhang et al., 2002) but is restricted to the scutellar tip and base during the late transition stage (Figure 5E) before eventually localizing to the emerging coleoptile (Figure 5F). At the L1 stage, LEC1 transcripts accumulate at the germinal side of the scutellar hood and at the boundary between the apical embryo and the basal suspensor (Figure 5G). As in Arabidopsis, no Zm-LEC1 transcripts are detected in foliar leaves (Figures 5H and 5L) (Lotan et al., 1998; Suzuki et al., 2008; Shen et al., 2010). Accumulation of Zm-LEC1 in both the scutellum and coleoptile, but not in maize embryonic foliar leaves, is likewise consistent with models in which the scutellum and the coleoptile comprise the apical and basal zones of the single maize cotyledon. Interestingly, only four transcripts are exclusively upregulated in the coleoptile stage versus the transition and/or L1 stages. One such transcript encodes the maize BASIC HELIX-LOOP-HELIX transcription factor PHYTOCHROME-INTERACTING FACTOR3-LIKE5 (PIL5); the PIL5 homolog in Arabidopsis functions to inhibit seed germination (cluster 5; see Supplemental Figure 4 online) (Oh et al., 2007). The relatively few transcripts exclusively upregulated during the coleoptile stage may indicate that the coleoptile transcriptome is inherently similar to that of both the scutellum and the foliar leaf. These data are consistent with models ascribing the dual identity of the coleoptile as a component of the cotyledon that also functions as a photosynthetic leaf-like organ.

Several gene transcripts upregulated at the L1 stage versus the transition stage are believed to have specific functions in foliar leaves, such as transcripts predicted to encode transcription factors regulating leaf size and leaf cell proliferation (i.e., SPATULA, AINTEGUMENTA, and ENHANCED DOWNY MILDEW2) (Mizukami and Fischer, 2000; Ichihashi et al., 2010; Tsuchiya and Eulgem, 2010). Additional examples include the YABBY-like transcription factor, which is implicated in leaf midrib formation (DROOPING LEAF1 [DL1]), and homologs of the microRNA biogenesis gene (SERRATE; Clarke et al., 1999; Yamaguchi et al., 2004; Laubinger et al., 2008). Interestingly, DL transcripts accumulate in young leaf primordia but not in the coleoptile or scutellum of rice (Oryza sativa) embryos (Yamaguchi et al., 2004), which further indicates that the grass coleoptile attains some, but not all, foliar-leaf attributes.

Among the differentially accumulated transcripts identified in embryonic lateral organs are several that exhibit tissue/organ-specific expression and are useful markers of embryo development, including (1) the scutellum-specific TAPETUM DETERMINANT1-LIKE (TDL1), (2) the vasculature-specific RAN BINDING PROTEIN2 (RANBP2), (3) the coleorhiza-specific POLLEN OLE E I-LIKE (OLE), and (4) two epidermis-specific maize LIPID TRANSFER PROTEINS (LTP*1 and LTP*2). TDL1 transcripts are first detected in the tip of the developing scutellar hood at the transition stage (Figure 6B) and in the outer cell layer of the scutellum in the coleoptile-stage and L1-stage embryos (Figure 6C). No TDL1 transcripts accumulate in the proembryo or in the 14-d-old seedling SAM and lateral meristem (Figures 6A, 6D, and 6E). RANBP2 transcripts accumulate in the developing vasculature of all lateral organs elaborated in the embryo and in 14-d-old seedlings (Figures 6F to 6J). OLE transcripts accumulate in the basal region of transition-stage embryos, opposite that of the apical expression described for the scutellum marker ZYB16 (Figures 3F, 6K, and 6I). As development proceeds, OLE transcripts are restricted to the basal region of the embryo, which eventually defines the coleorhiza, a sheathing structure that surrounds the maize primary root (Figures 6M to 6O). The two protodermal markers, LTP*1 and LTP*2, are expressed in the apical regions of the proembryo and early transition-stage embryo (Figures 6P and 6Q; see Supplemental Figures 7A and 7B online), whereupon these LTP transcripts accumulate in the emerging scutellum of the late transition-stage embryo (Figure 6R). Subsequently, LTP*1 and LTP*2 localize to the epidermal cell layer of embryonic lateral organs (Figures 6S and 6T; see Supplemental Figures 7C and 7D online).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Differentially Accumulated Transcripts during the Elaboration of Embryonic Organs.

In situ hybridizations illustrate transcript accumulation of TDL1 ([A] to [E]), RANBP2 ([F] to [J]), OLE ([K] to [O]), and LTP*1 ([P] to [T]). 24 DAP, embryo harvested 24 d after pollination; col, coleoptile stage; L1, leaf 1 stage; L14; 14-d-old seedling SAM; LM, lateral meristem; pro, proembryo; tr, transition phase.

Bars = 100 μm.

Other useful developmental markers identified in our RNA-seq analyses include the homolog of an Arabidopsis LIGHT-DEPENDENT SHORT HYPOCOTYLS1 and ORYZA LONG STERILE LEMMA1 (ALOG) transcript ALOG*1, which is implicated during specification of organ boundaries, and two gene transcripts of unknown predicted function, designated as UNKNOWN*1 (UNK*1) and UNKNOWN*2 (UNK*2). In maize embryos, ALOG*1 transcripts demarcate the boundary between all lateral organs and the developing SAM (Figures 7A to 7E). UNK*1 transcripts first accumulate in a few cells at the site of organ initiation and then become localized to the adaxial side of the emerging lateral organ (Figures 7F to 7J). By contrast, UNK*2 transcripts are detected in the growing edges of the sheath-like coleoptile (Figures 7K to 7O) and in the margins of all subsequent embryonic lateral organs. The similar accumulation patterns of these three transcripts during elaboration of the scutellum, coleoptile, and first foliar leaf further supports the model ascribing leaf homology to the scutellum and the coleoptile.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Developmental Markers Identified in RNA-Seq Analyses.

In situ hybridizations illustrate transcript accumulation of ALOG*1 ([A] to [E]), UNK*1 ([F] to [J]), and UNK*2 ([K] to [O]). col, coleoptile stage; L1, leaf 1 stage; L14; 14-d-old seedling SAM; LM, lateral meristem; pro, proembryo; tr, transition phase.

Bars = 100 μm.

Comparative Transcriptomic Analyses of Juvenile, Adult, and Husk Foliar Leaves

Juvenile and adult foliar leaves of maize exhibit significant morphological differences. Juvenile leaves are short and narrow and accumulate epicuticular waxes but no epidermal hairs, and lateral meristems of juvenile nodes form long vegetative shoots tipped by a male inflorescence (tassel) (Poethig, 1988; Poethig, 1990; Moose and Sisco, 1994). By contrast, adult leaves are long and broad and have epidermal hairs but accumulate little epicuticular waxes, and adult lateral meristems form short branches that initiate husk leaves and form a female inflorescence (ear) (Poethig, 1988; Poethig, 1990; Moose and Sisco, 1994). Furthermore, leaves initiated from both the SAM and juvenile lateral meristems differ in many ways from husk leaves formed from adult-staged lateral meristems. Compared with foliar leaves, husk leaves have a lower vein density (Langdale et al., 1988), accumulate ribulose-1,5-bis-phosphate carboxylase/oxygenase in mesophyll as well as bundle sheath cells (Pengelly et al., 2011), and make a greater contribution to grain filling (Fujita et al., 1994; Sawada et al., 1995).

A complex interactive network of phase-specific transcription factors and regulatory small RNAs regulate vegetative and reproductive phase change in plants (Moose and Sisco, 1996; Lauter et al., 2005; Chuck et al., 2007; Hultquist and Dorweiler, 2008). To examine the transcriptional differences among juvenile-stage SAMs, adult-stage SAMs, and lateral shoot meristems from adult nodes on a genomic scale, the SAM and youngest leaf primordium were microdissected from L1-stage embryos and L14-stage seedlings, and lateral meristems were isolated from L14-stage seedlings. RNA-seq analyses and pairwise comparisons of these three microdissected samples identified a total of 5103 upregulated transcripts. A total of 12 major patterns (clusters) of transcript accumulation were identified in the three-way comparisons (Figure 8A). Cluster six is the largest and comprises 1121 transcripts that were upregulated at the L14 stage compared with the L1 stage. Other large clusters include (1) cluster 4 (886 transcripts), which comprises transcripts upregulated at the L14 stage compared with both the L1 stage and lateral meristem; (2) cluster 9 (596 transcripts), comprising transcripts upregulated in the lateral meristem compared with the L1 stage; and (3) cluster 2 (536 transcripts), which comprises transcripts upregulated in the L1 stage as compared with the L14 stage. A Fisher’s exact test was performed to identify enriched functional groups within the 12 gene clusters (Figure 8B; see Supplemental Data Set 13 online). Functions enriched during the initiation of a juvenile leaf include RNA processing and binding (cluster 2); DNA synthesis, repair, and chromatin remodeling (cluster 3); protein synthesis and targeting (clusters 1 and 2); and the tricarboxylic acid cycle (cluster 1). The only category exclusively enriched in the lateral meristem is unknown function (cluster 9). Consistent with differences in cell wall composition between juvenile and adult foliar leaves, transcripts implicated in cell wall synthesis and degradation are enriched in both the L14-stage SAM and lateral meristems (cluster 12) (Abedon et al., 2006).

• Download figure
• Open in new tab
• Download powerpoint

Figure 8.

Initiation of Juvenile, Adult, and Husk Leaves.

(A) A total of 12 clusters were generated based on the transcript accumulation of the 5103 upregulated genes when making the three-way comparison of the L1 stage, L14 stage, and lateral meristem.

(B) Functional category enrichment of the 12 clusters. The −log₁₀P value and FDR 5% are included for significant enrichment. White indicates nonsignificant enrichment, whereas red indicates significant enrichment. Gray indicates that no genes are present in the category for that cluster. In situ hybridizations using the following probes: ALOG*2 ([C] to [E]), LTP*3 ([F] to [H]). L1, leaf 1; L14, leaf 14; LM, lateral meristem; TCA, tricarboxylic acid.

Bars = 100 μm.

Transcription factors implicated in vegetative phase change were differentially expressed between juvenile and adult leaves. As previously described, eight SQUAMOSA-PROMOTER BINDING PROTEIN-LIKE (SPL) transcripts targeted by the juvenile-stage regulatory RNA miR156 were upregulated in the L14 stage and lateral meristems (see Supplemental Data Set 14 online) (Chuck et al., 2007; Hultquist and Dorweiler, 2008; Strable et al., 2008). In addition, GLOSSY15 (GL15), an AP2-like transcription factor that promotes the adult leaf identity, is upregulated in 14-d-old seedlings (Evans et al., 1994; Moose and Sisco, 1996; Lauter et al., 2005). Transcriptional profiles of husk leaves are more similar to L14-stage than L1-stage leaves, as illustrated by the 417 gene transcripts that were upregulated in both the L14-stage SAM and the lateral meristem versus the L1-stage SAM and are displayed in metabolic and regulatory pathways (cluster 12) (see Supplemental Figures 8 and 9 and Supplemental Data Sets 15 and 16 online). This is consistent with previous studies showing that, once initiated, vegetative phase change is comprehensive and affects all subsequently elaborated organs of the maize shoot (Poethig, 1990).

Two transcripts upregulated during husk leaf initiation (ALOG*2 and LTP*3) were analyzed by in situ hybridization. ALOG*2 transcripts accumulate in the lateral organ boundaries of all three samples, suggesting a conserved function in defining foliar and husk leaves (Figures 8C to 8E). However, LTP*3 transcripts accumulate in the outer cell layer of young leaf primordia and in the tunica of the lateral meristems but are excluded from the L1-stage and L14-stage SAMs (Figure 8B). Although the lateral meristem and the L14-stage SAM are both initiating vegetative leaves, a total of 1505 transcripts were upregulated (clusters 4, 5, 7, 8, 10, and 11), which reflects distinct developmental genetic mechanisms within these shoot meristems. The transcripts that are found in metabolic or regulatory pathways are displayed in Supplemental Figures 10 and 11 and Supplemental Data Sets 17 and 18 online.

In summary, the SAM transcriptome undergoes dynamic changes throughout the course of vegetative plant development. Analysis of embryonic transcript accumulation before meristem formation identified ZYB16, a gene implicated in lateral organ initiation. Expression of ZYB16 before KN1 suggests that SAM function during organogenesis precedes that of stem cell maintenance. Transcriptomic analysis of a newly formed SAM indicates that, at this early developmental stage, the embryonic shoot is still establishing apical/basal domains and accumulating storage products to be used during the proliferative organogenesis and growth that defines later stages of maize embryo development. By contrast, the mature maize SAM is enriched for transcripts that function in transcriptional regulation, hormone signaling and metabolism, and transport. Transcriptomic analyses of the three distinct types of lateral organs elaborated during embryogenesis provide support for models ascribing the maize cotyledon as a leaf-like, bimodal structure that comprises both the scutellum and the coleoptile. Finally, three-way comparisons between the initiation of juvenile, adult, and husk leaves reveal phase-specific and meristem-specific differential transcript accumulation. Novel molecular markers are identified that specify the cotyledon, the scutellum, the vasculature, the coleorhiza, initiating lateral organs, lateral organ boundaries, (Sen et al., 2009) and the protoderm or epidermis. Finally, all these transcriptomic data are publicly available at the Maize Genetics and Genomics Database (MaizeGDB; www.maizegdb.org) (Sen et al., 2009).