The single primary root is preformed in the embryo and emerges from the basal pole of the seed 2–3 days after germination (Hochholdinger et al., 2004a). It has been demonstrated that the majority of the soluble proteome of maize primary roots of the inbred line B73 significantly changes between 5 and 9 days after germination. Only 28% of proteins identified in 5-day-old primary roots were also observed in 9-day-old primary roots (Hochholdinger et al., 2005).

In longitudinal orientation, maize roots are organized in several partially overlapping zones of development comprising the meristematic, elongation and differentiation zones (Ishikawa and Evans, 1995). To study the molecular function of radial zones of development, the differentiation zone of 2.5-day-old primary roots was manually separated into cortical parenchyma and stele (Saleem et al., 2010). The stele comprises the central vascular cylinder and the pericycle, while the cortical parenchyma includes all tissues surrounding the central vascular cylinder. The cortical parenchyma comprises a single endodermal cell layer, several layers of cortex cells and a single epidermal cell file. The stele transports water, nutrients and photosynthates, while the cortical parenchyma represents the ground tissue of the root, which has metabolic functions but is also involved in the transport of material into the stele. Phytohormone profiling revealed enhanced levels of auxin in the stele and of cytokinin in the cortical parenchyma (Saleem et al., 2010). Several cytokinin-dependent proteins were upregulated in the cortical parenchyma. This localization is consistent with the preferential accumulation of cytokinin in this tissue. Some of these proteins are enzymes, including a β-glucosidase that hydrolyze the cytokinin cis-zeatin O-glucoside and four enzymes involved in ammonium assimilation (Saleem et al., 2010). The antagonistic levels of auxin and cytokinin in the stele and cortical parenchyma and the cortical parenchyma-specific accumulation of cytokinin-regulated proteins highlight a molecular framework that regulates the function of these distinct root tissues (Saleem et al., 2010). Recently, a high-resolution map of the proteome and phosphoproteome of different tissues of the maize primary root (Marcon et al., 2015) complemented this initial tissue specific proteomic dissection of the maize primary root. In this analysis cortical parenchyma, stele, meristematic zone, and elongation zone were separately subjected to high-performance liquid chromatography coupled with tandem mass spectrometry. As a result, 11,552 distinct non-modified and 2,852 phosphorylated proteins were identified in these four root tissues (Marcon et al., 2015). Along the longitudinal axis, the abundance of functional protein classes was highlighted by two gradients. Along these gradients, the functional classes “RNA”, “DNA”, and “protein” peaked in the meristematic zone, whereas the categories “cell wall”, “lipid metabolism”, “stress”, “transport”, and “secondary metabolism” displayed maximum accumulation in the differentiation zone (Marcon et al., 2015). Comparison of this primary root tissue data set with high-resolution datasets of maize seed (Walley et al., 2013) and leaf (Facette et al., 2013) proteomes revealed that 13% of the identified proteins were exclusively detected in the primary root. These root-specific proteins displayed a high degree of tissue specific functionalization and underscored the plasticity of tissue specific proteomes in maize primary roots (Marcon et al., 2015).

On the subcellular level, the maize primary root was also a model to study the proteomes associated with cell walls (Zhu et al., 2007) and plasma membranes (Zhang et al., 2013). Cell wall proteins have been studied in the elongation zone of maize primary roots (Zhu et al., 2007). Among the identified cell wall proteins, many were involved in cell elongation and cell wall metabolism. Maize forms type II cell walls (Carpita, 1996). Therefore, several of the cell wall proteins identified in the elongation zone of maize primary roots have not been identified in proteomic studies that have focused on type I walls (Zhu et al., 2007). Similarly, the distribution of plasma membrane proteins in the growth zone of the maize primary root was studied during development (Zhang et al., 2013). The plasma membrane is the interface between the plant cell and the extracellular space or the apoplast and thus involved in the exchange of cellular molecules and in signal integration. This study demonstrated that cellulose synthases became more abundant with increasing distance from the root apex. This is consistent with the expected localization of cell wall deposition (Zhang et al., 2013).

Among other compounds, maize root cap cells secrete proteins into the rhizosphere via an amorphous gel structure called mucilage. From the mucilage secreted by 3 to 4-day-old primary roots 2,848 distinct extracellular proteins were identified by nanoLC–MS/MS (Ma et al., 2010). Among those, metabolic proteins represented the largest class. A comparison with the mucilage proteins previously identified in several dicot species suggested a considerable overlap between monocot and dicot mucilage proteomes (Ma et al., 2010).

Seminal roots are the second embryonic root type of maize formed after the primary root. They emerge from the scutellar node and their number is variable between different genotypes but also within a given genotype (Hochholdinger et al., 2004a). The maize mutants rtcs (rootless concerning crown and seminal roots; Hetz et al., 1996) and rum1 (rootless with undetectable meristems 1; Woll et al., 2005) do not initiate seminal roots. Both genes encode key components of auxin signal transduction. Auxin triggers transcriptional responses via the regulation of the activity of AUXIN RESPONSE FACTOR (ARF) proteins (reviewed in, Lavy and Estelle, 2016). At low auxin levels, Auxin/INDOLE-3-ACETIC ACID (Aux/IAA) transcriptional repressors interact with ARFs and repress their activity. In contrast, at high auxin levels TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX PROTEIN (TIR1/AFB) proteins bind to Aux/IAA transcriptional repressors and mediate their degradation via the proteasome. As a consequence, ARF transcription factors activate direct downstream target genes such as LOB domain transcription factors by binding to auxin response elements in their promoter (Xu et al., 2016). While rtcs encodes a LOB domain transcription factor (Taramino et al., 2007; Xu et al., 2015, 2016), rum1 encodes an Aux/IAA transcriptional regulator (von Behrens et al., 2011). It was demonstrated that rtcs and rum1 are not only critical for seminal root initiation but also for seminal root number because they likely underlie the two major QTLs controlling this trait (Salvi et al., 2016). Immature wild-type and rtcs embryos were subjected to a comparative proteome analysis 25 days after pollination (Muthreich et al., 2010). At this stage wild-type embryos started the initiation of seminal roots while mutant embryos did not. Among the differentially accumulated proteins, two phosphoglycerate kinases and a malate dehydrogenase, which are crucial checkpoints of cellular energetics, were preferentially accumulated in wild-type versus mutant embryos (Muthreich et al., 2010). Similarly, the proteomes of wild-type and rum1 embryos were compared 30 days after pollination (Saleem et al., 2009). In maize, GLOBULIN 1 and GLOBULIN 2 are the most prevalent storage proteins in embryos (Kriz, 1989; Belanger and Kriz, 1991). These proteins are exclusively expressed during embryo development (Belanger and Kriz, 1991; Kriz, 1999). In total, seven GLOBULIN 1 isoforms and four GLOBULIN 2 isoforms were differentially accumulated between wild-type and rum1 embryos 30 days after pollination suggesting significant regulation of these proteins by RUM1 (Saleem et al., 2009). It was pointed out that the regulation of GLOBULIN1 by the phytohormone abscisic acid (ABA), which stimulates root elongation and branching, might link this class of proteins to root formation (Saleem et al., 2009). Only little overlap was observed between embryonic proteins differentially accumulated between wild-type and rtcs embryos 25 days after pollination (Muthreich et al., 2010) and wild-type and rum1 embryos 30 days after pollination (Saleem et al., 2009). This highlights the distinct regulation of the maize embryo proteome by RTCS and RUM1, but this difference could also be in part due to the analysis of the distinct stages of embryo development.

After germination, whorls of shoot-borne roots are formed at successive nodal structures of the shoot (Hochholdinger et al., 2004b). The maize rootstock develops ∼70 shoot-borne roots in the course of development (Hoppe et al., 1986). Therefore, shoot-borne roots make up the major proportion of the adult maize root system. The previously described maize mutant rtcs (Hetz et al., 1996) is impaired in the initiation of all shoot-borne roots. In wild-type maize, the first node from which shoot-borne roots emerge is the coleoptilar node. In a proteome study, the soluble proteomes of wild-type versus mutant rtcs coleoptilar nodes were compared 5 and 10 days after germination (Sauer et al., 2006). These stages coincided with the initiation and emergence of shoot-borne roots in wild-type coleoptilar nodes, respectively. Several of the differentially accumulated proteins detected in this study are involved in the regulation of plant development (Sauer et al., 2006). The differentially accumulated proteins at the two developmental stages displayed only little overlap, indicating that distinct sets of proteins control initiation and emergence of lateral roots. RNA gel blot experiments of a subset of differentially accumulated proteins demonstrated that these expression differences are already manifested on the RNA level (Sauer et al., 2006).

After germination, lateral roots emerge by division of pericycle cells in the differentiation zone of all root types (Hochholdinger et al., 2004b). The maize mutants lrt1 (lateral rootless 1; Hochholdinger and Feix, 1998) and rum1 (Woll et al., 2005) are impaired in the initiation of lateral roots. Differentially accumulated proteins were determined in a proteome survey of 9-day-old wild-type primary roots with lateral roots and lrt1 primary roots without lateral roots (Hochholdinger et al., 2004c). Remarkably, 10% of proteins identified in this study were predominantly accumulated in lrt1 primary roots that lack lateral roots suggesting that the presence of lateral roots can significantly influence the proteome of the primary root (Hochholdinger et al., 2004c). To study the molecular processes prior to lateral root formation, 2.5-day-old wild-type and mutant rum1 primary roots were subjected to a proteome analysis (Liu et al., 2006). At this early stage of development no morphological differences were observed between the primary roots of the two genotypes. Nevertheless, differentially accumulated proteins involved in defense, lignin biosynthesis and in the citrate cycle were identified (Liu et al., 2006). In a follow up study the differentiation zone of 2.5-day-old wild-type and mutant rum1 primary roots were manually separated in cortical parenchyma and stele and subjected to a proteome analysis (Saleem et al., 2009). Differentially accumulated proteins between the mutant rum1 and wild-type demonstrated that RUM1 regulates the proteome of cortical parenchyma and stele tissues. Among the biochemical pathways regulated by RUM1, enzymes related to glycolysis were differentially expressed between these tissues (Saleem et al., 2009). It has been demonstrated that key enzymes of glycolysis such as hexokinase (Halford and Paul, 2003) and pyruvate kinase (Moore et al., 2003) control root elongation via sugar sensing and signaling. It was therefore suggested that the reduced primary root length of the rum1 mutant could be conferred by tissue specific imbalances of glycolytic enzymes during root development (Saleem et al., 2009). Finally, in a pioneering work, pericycle cells which give rise to lateral roots were captured via laser microdissection (Schnable et al., 2004) and subjected to a proteome analysis (Dembinsky et al., 2007). As a starting point for future cell-type specific analyses, this initial study identified the twenty most abundant soluble proteins of the maize pericycle cell proteome (Dembinsky et al., 2007).

Root hairs are tubular extensions of trichoblasts in the epidermis, which significantly increase the absorbing surface of the root and thus the uptake of nutrients (Gilroy and Jones, 2000). In maize, root hair patterning is random and it is not predictable which epidermal cells form root hairs (Dolan, 1996). Several genes control root hair elongation in maize. The root hair defective genes rth3 (Hochholdinger et al., 2008), rth5 (Nestler et al., 2014), and rth6 (Li et al., 2015) are functionally linked in the processes of cell wall loosening, cellulose synthesis, and organization of synthesized cellulose, respectively. In contrast, the gene rth1 (Wen et al., 2005) encodes a SEC3 subunit of the exocyst complex (Hala et al., 2008). To gain a better insight into the root hair proteome, a reference set of 2,573 soluble proteins of maize root hairs of 4-day-old primary roots of the inbred line B73 was identified (Nestler et al., 2011). Root hairs are an ideal model for single cell proteome analyses in maize because they can be easily separated from the primary root after freezing the roots in liquid nitrogen. Among these root hair proteins, homologs of 252 proteins have been associated with root hair development in other species (Nestler et al., 2011). A comparison of the maize root hair reference proteome with those of soybean revealed conserved, but also unique protein functions related to root hairs in these species (Nestler et al., 2011).