Plants exhibit an astonishing variety of shapes and develop multicellular bodies able to live for hundreds of years and reach considerable size. They rely on continuous growth and are able to regenerate organs from undifferentiated meristematic cells populations. Plant growth and organ differentiation, as well as response to environmental stimuli, are regulated, among other factors, by endogenous compounds called phytohormones. They control the plant developmental program by regulating cell division and expansion, tissue differentiation, and senescence. Phytohormones can act within the cell of origin or move to other sites in the plant, where they are perceived as a signal by hormone receptors (Davies, 2004).

The plant hormone auxin was first isolated as Indol-3-acetic acid (IAA) by Went (1926), as he studied the tropic response of Avena sativa coleoptiles. Subsequently, during the first half of the twentieth century, other four phytohormones were identified, including abscisic acid, cytokinins, gibberellins, and ethylene (Kende and Zeevaart, 1997). More recently, several additional compounds have been recognized as hormones including brassinosteroids (BR), jasmonate (JA), salicylic acid (SA), nitric oxide (NO), and strigolactones (SLs) (Tarkowská et al., 2014). Auxin is a regulator of many aspects of plant development, including cell division, elongation, differentiation, embryonic development, root and stem tropisms, apical dominance, and flower formation (Young et al., 1990; Woodward and Bartel, 2005; Tanaka et al., 2006; Möller and Weijers, 2009; Leyser, 2010; Müller and Leyser, 2011; Christie and Murphy, 2013; Gallavotti, 2013; Geisler et al., 2014). Besides IAA, which is the most abundant natural form of auxin, several auxin-like molecules have been identified. While 4-chloroindole-3-acetic acid (4-Cl-IAA), indole-3-butyric acid (IBA), and phenylacetic acid (PAA) are all found in plants, 2,4-dichlorophenoxyacetic acid (2,4-D) and naphthalene-1-acetic acid (NAA) are synthetic compounds that have biological activity similar to IAA (Bertoni, 2011; Simon and Petrášek, 2011).

Local biosynthesis, degradation and conjugation contribute to the modulation of IAA homeostasis at the cellular level. Availability of free IAA inside the cell is also controlled by auxin transport, which occurs in two distinct pathways: a passive diffusion through the plasma membrane (PM) and an active cell-to-cell transport, depending on the protonation state of IAA. IAA is a weak acid with a dissociation constant of pK = 4.8. In a neutral or basic environment IAA^- will be the most abundant form (99.4% ionized at pH = 7.0), whereas in the acidic extracellular space IAAH is predominant (about 20% protonated at pH = 5.5) (Delbarre et al., 1996; Estelle, 1998; Kramer and Bennett, 2006). IAAH can enter into the cell through the PM by passive diffusion or active transport by PM importers. Once inside the cytoplasm, which has a neutral pH, IAA^- becomes the predominant form and it cannot freely move out of the cell unless actively transported by efflux carrier proteins (Figure 1). The differential localization of transporters at specific sites on the PM creates a directional auxin flow that eventually establishes a polar auxin transport (PAT) stream through adjacent cells. Four classes of auxin transporters have been identified: the PIN-FORMED (PIN) exporters, the ATP-binding cassette (ABC)-B/multi-drug resistance/P-glycoprotein (ABCB/MDR/PGP) subfamily of ABC transporters, the AUXIN1/LIKE-AUX1 (AUX/LAX) importers, and the newly described PIN-LIKES (PILS) proteins.

Despite the fact that auxin was first isolated and studied in the monocot A. sativa, characterization of auxin transport proteins derives mostly from forward genetic studies of mutants with defects in development, organ morphogenesis, and gravitropism in the dicot Arabidopsis thaliana. In recent years, the number of studies on the biological role of PAT in monocots has increased. This has been facilitated by the lower cost of deep sequencing of whole plant genomes and transcriptomes and by the availability of tools such as transgenic lines carrying proteins with fluorescent tags, which are used in subcellular localization studies and PAT fluxes modeling (Mohanty et al., 2009; Egan et al., 2012; Yu et al., 2012). In this work, we present a comprehensive description of monocots auxin transporters and provide, where possible, functional comparison between monocot and Arabidopsis proteins.

PIN and PILS protein sequences of Arabidopsis, rice, maize sorghum, and Brachypodium (gene accession numbers are listed in Table S1) were aligned using the CustalW 2.0 software (Larkin et al., 2007). The alignment file was used to generate an unrooted tree with MEGA 6.0 (Tamura et al., 2013), applying the Neighbor-joining method, the Poisson model and 500 bootstrap replications. Bootstrap analysis values >60 are indicated at each node.

PINs are the most studied family of auxin transporters in plants. PIN genes are present in eight copies in Arabidopsis and encode integral membrane proteins with two conserved domains formed by transmembrane helices, typically five at both the N and C termini, and a less conserved central hydrophilic loop of variable length (Křeček et al., 2009; Ganguly et al., 2012). Their subcellular localization has been correlated with the length of the hydrophilic domain. In Arabidopsis, PIN1, -2, -3, -4, and -7 have a longer loop (ranging in size from 298 to 377 amino acid residues), PIN5 and -8 have a shorter loop (27–46 residues) and PIN6 contains an intermediate form (Křeček et al., 2009; Ganguly et al., 2010; Viaene et al., 2013). “Long” PINs are generally inserted into the PM while “short” PINs are located in the endoplasmic reticulum (ER) and they are thought to contribute to intracellular auxin homeostasis (Mravec et al., 2009; Ding et al., 2012; Cazzonelli et al., 2013). More recently, it has been demonstrated that PIN5 is also PM localized, depending on the cell type and developmental stage, and that PIN5, -6, and -8 function in polar cell-to-cell transport of auxin by regulating coordinated influx and efflux of IAA into and out of the ER (Bender et al., 2013; Sawchuk et al., 2013; Ganguly et al., 2014).

Several auxin transporters show polar localization in the cell, but it is only in the case of PIN proteins that polar targeting occurs more frequently (Figure 1). Shifting PIN polarity results in the alteration of PAT which leads to developmental defects in Arabidopsis (Löfke et al., 2013). The polar localization of PIN proteins is established by cycling between the PM and endosomal compartments such as the trans-Golgi network/early endosomes (TGN/EE). PIN recycling can take place via endocytosis of clathrin-coated vesicles and depends on phosphorylation and ubiquitination (Robert et al., 2010; Kleine-Vehn et al., 2011; Löfke et al., 2013). Unphosphorylated PINs, or those dephosphorylated by the PP2A/PP6 phosphatase, are recycled to the PM by the brefeldin A (BFA)-sensitive ADP-ribosylation factor-guanine nucleotide exchange factor (ARF-GEF) GNOM. Phosphorylation of PIN proteins by the protein kinase PINOID (PID) results in their GNOM-independent recycling to the PM on the opposite side of the cell (Friml et al., 2004; Dhonukshe et al., 2010). Monoubiquitination and subsequent polyubiquitination of PIN proteins induce their endocytosis, followed by trafficking from the TGN/EE to late endosomes, from where the SNX1/BLOC-1 complex mediates transfer to multivesicular bodies (MVBs) for vacuolar degradation (Habets and Offringa, 2014). Recently, another Arabidopsis kinase, D6 PROTEIN KINASE (D6PK), has been demonstrated to regulate PIN phosphorylation and, together with PID, D6PK promotes PINs-mediated auxin transport at the PM by maintaining their phosphorylation status. D6PK PM localization is essential to establish and maintain PIN phosphorylation, and d6pk mutants have defects in both negative gravitropism and phototropism due to impaired auxin transport (Zourelidou et al., 2009; Willige et al., 2013; Barbosa et al., 2014).

Phylogenetic studies on the origin and evolution of PIN proteins have demonstrated that their general structure is highly conserved across the plant kingdom and suggest that the last common ancestor of land plants had at least one “long” (canonical) PIN protein (Carraro et al., 2012; Bennett et al., 2014). Strong selective pressure maintained PINs function as auxin carriers while they underwent sub- and neo-functionalization with substantial modification to protein structure, possibly due to selective loss of phosphorylation sites in their central loop (Dhonukshe et al., 2010; Fozard et al., 2013; Bennett et al., 2014). This generated several clades of non-canonical proteins with shorter and divergent structure, leading to altered localization and biological function. Monocot PIN families are often enlarged due to whole genome duplications and the retention of multiple copies of similar proteins. Both Oryza sativa and Zea mays contain four PIN1 copies and present at least one monocot-specific gene, PIN9, which is divergent in sequence and expression pattern from the closest dicot PINs (Xu et al., 2005; Forestan et al., 2012; Bennett et al., 2014; Clouse and Carraro, 2014). The PIN9 protein profile prediction shows an intermediate hydropathic profile in between the “long” and the “short” PINs.

AtPIN1 is expressed early during embryonic development and later both in the primary root and in the inflorescence stem. Disruption of AtPIN1 expression leads to the formation of naked, pin-shaped inflorescences and abnormalities in the number, size, shape, and position of lateral organs (Okada et al., 1991; Gälweiler et al., 1998). As suggested by the pin1 phenotype, PIN1 plays an important role in establishing the plant developmental plan and is involved in floral bud formation, phyllotaxis (the arrangement of leaves and flowers around the stem), vascular development, vein formation, embryogenesis, lateral organ formation, anther development, and root negative phototropism (Gälweiler et al., 1998; Benková et al., 2003; Reinhardt et al., 2003; Weijers et al., 2005; Feng et al., 2006; Scarpella et al., 2006; Lampugnani et al., 2013; Zhang et al., 2014). Arabidopsis pin2 was the first pin mutant identified in a screen for agravitropic seedlings by Bell and Maher (1990). Initially, it was called agr1, and the gene responsible for the phenotype was cloned independently by four research groups and named AGR1/EIR1/PIN2/WAV6 (Chen et al., 1998; Luschnig et al., 1998; Müller et al., 1998; Utsuno et al., 1998). AtPIN2 functions in auxin-regulated root gravitropic response and its expression levels and polar cellular localization are altered by salt stress (Abas et al., 2006; Sun et al., 2008). AtPIN3 is expressed during embryo development and the Atpin3 mutant shows reduced growth, and decreased apical hook formation (Friml et al., 2002b; Zádníková et al., 2010). It has been shown that AtPIN3 plays an important role in root gravitropism, as in vertically grown seedlings AtPIN3 is positioned symmetrically at the PM in the columella cells, but rapidly re-localizes laterally to the lower PM of the statocytes following gravistimulation. AtPIN3 relocalization determines the direction of the auxin flux, which leads to asymmetric auxin accumulation and subsequent differential cell growth (Friml et al., 2002b). AtPIN3 is also involved in root negative phototropic response, as blue-light induced AtPIN3 polarization is needed for asymmetric auxin distribution (Zhang et al., 2013).

AtPIN4, as well as AtPIN7, are involved in auxin controlled embryo, primary root, and apical hook development (Friml et al., 2002a, 2003; Vieten et al., 2005; Kleine-vehn et al., 2010; Zádníková et al., 2010). AtPIN7 also undergoes relocalization similar to AtPIN3 in response to gravistimulation in a subgroup of the columella cells (Rosquete et al., 2013). AtPIN5 is involved in auxin homeostasis, and it has been demonstrated that it can export auxin from yeast cell (Mravec et al., 2009). AtPIN6 is involved in stamen development, and microarray and reporter assays have demonstrated that it is necessary for nectaries development (Bender et al., 2013). Expression analysis of pPIN6::PIN6-GFP lines during leaf development has demonstrated that AtPIN6 localizes to the ER and expression is initiated in broad sub-epidermal domains that later on narrow to sites of vein formation (Bender et al., 2013; Sawchuk et al., 2013). Moreover, AtPIN6 is implicated in processes such as shoot apical dominance, lateral root primordia development, adventitious root formation, root hair outgrowth, and root waving where it regulates auxin homeostasis (Cazzonelli et al., 2013). AtPIN8 is expressed in the male gametophyte, and has a crucial role in pollen development and functionality (Ding et al., 2012). AtPIN8, together with AtPIN5 and AtPIN6, also take part into leaf vein network patterning by regulating intracellular auxin transport between the cytoplasm and the ER lumen. Their action is exerted coordinately with AtPIN1, in order to modulate intracellular auxin levels in extending veins (Sawchuk et al., 2013).

The ABC superfamily of membrane proteins includes more than a hundred different members in plants (Kang et al., 2011). The subfamily B (ABCB) includes homologs of the mammalian MDRs/PGPs, several of which are involved in auxin transport (Geisler and Murphy, 2006; Cho and Cho, 2013). ABCB transporters are integral membrane proteins that actively transport chemically diverse substrates across the lipid bilayers of cellular membranes (Figure 1). The core unit of a functional ABC transporter consists of four domains: two nucleotide-binding domains (NBDs) and two transmembrane domains (TMDs). The two NBDs unite to bind and hydrolyze ATP, providing the driving force for transport, while the TMDs are involved in substrate recognition and translocation across the membrane (Jasinski et al., 2003; Higgins and Linton, 2004; Bailly et al., 2011). Arabidopsis has 22 ABCBs and the first ABCBs characterized as functioning in IAA traslocation were identified in seedlings (Sidler et al., 1998; Noh et al., 2001). ABCB1, ABCB4, ABCB14, ABCB15, ABCB19, and ABCB21 are associated with auxin transport, although not exclusively (Geisler and Murphy, 2006; Titapiwatanakun and Murphy, 2009; Kaneda et al., 2011; Kamimoto et al., 2012; Cho and Cho, 2013). To date, the best-characterized ABCBs are AtABCB1, AtABCB4, and AtABCB19. They all function in auxin driven root development and require the activity of the immunophilin TWISTED DWARF1 (TWD1)/FKBP42 to be correctly inserted at the PM (Wu et al., 2010). AtABCB1/PGP1 was the first plant MDR-like gene cloned from Arabidopsis and it is localized at the PM in the root and the shoot apex of seedlings (Dudlers and Hertig, 1992; Noh et al., 2001). The Atpgp1 original mutant exhibits only a subtle phenotype compared to wild type plants, but a new allele designated as atpgp1-2, shows a shorter hypocotyl and dwarf phenotype under long-day conditions (Geisler et al., 2005; Ye et al., 2013). Disruption of AtABCB19/AtMDR1 expression results in partial dwarfism and reduced PAT in hypocotyls and inflorescences (Noh et al., 2001). AtABCB19 functions together with AtABCB1 in long distance transport of auxin along the plant main axis in coordination with AtPIN1, and regulates root and cotyledon development and tropic bending response (Lin and Wang, 2005; Bandyopadhyay et al., 2007; Rojas-Pierce et al., 2007; Nagashima et al., 2008; Lewis et al., 2009; Christie et al., 2011). AtABCB4 is a root-specific transporter involved in auxin transport during root gravitropic bending, root elongation, and lateral root formation (Santelia et al., 2005; Terasaka et al., 2005; Kubeš et al., 2011; Cho et al., 2012). This transporter is substrate-activated and functions as an auxin importer at low substrate concentration, switching to auxin export as the availability of auxin increases (Yang and Murphy, 2009; Kubeš et al., 2011). AtABCB21 encodes a protein that is the closest homolog to AtABCB4 and is expressed in the aerial parts of the seedling and in the root pericycle cells. Just like AtABCB4, AtABCB21 functions as a facultative importer/exporter that controls cellular auxin levels (Kamimoto et al., 2012). AtABCB14 was first described as a malate importer that functions in the control of stomata aperture according to CO₂ levels (Lee et al., 2008). More recently, AtABCB14 and 15 have been shown to be active in the vascular tissue of the primary stem, which shows anatomical alterations in abcb14 and abcb15 mutants. Since IAA transport along the inflorescence is reduced in both mutants, it was proposed that AtABCB14 as well as AtABCB15 participate in auxin transport (Kaneda et al., 2011).

The existence of auxin importers in plants was first demonstrated studying the Arabidopsis auxin insensitive 1 (aux1) mutant, which carries defects in roots gravitropic response. AtAUX1 belongs to a small gene family composed of four highly conserved proteins that share similarities with amino acid transporters: AtAUX1, AtLAX1, -2, and -3 (Péret et al., 2012). AtAUX1 encodes a protein similar to fungal amino acid permeases and is expressed in columella, lateral root cap, epidermis, and stele tissues of the primary root where it acts as an auxin importer (Bennett et al., 1996; Swarup et al., 2001, 2004; Carrier et al., 2008). AtAUX1 is involved in auxin-regulated root gravitropic response together with the auxin exporter AtPIN2. The coordinated action of these two proteins forms a lateral auxin gradient which inhibits the expansion of epidermal cells on the lower side of the root relative to the upper side, eventually causing the downward root curvature (Swarup et al., 2005). aux1, as well as pin2 Arabidopsis mutants are agravitropic and aux1 also presents a decreased number of lateral roots due to defects in lateral root initiation (Marchant et al., 2002). AtAUX1 and AtLAX1, act redundantly in regulating the phyllotactic pattern in Arabidopsis although AtLAX2 is not expressed in the SAM L1 layer. Since AtLAX2 is expressed in the forming primordium vasculature, one hypothesis is that AtLAX2 enhances the strength of the primordium as an auxin sink by pulling IAA from the L1 layer of the SAM (Bainbridge et al., 2008; Kierzkowski et al., 2013). AtLAX2 is involved in vascular development in cotyledons and it is also expressed in young vascular tissues, the quiescent center and columella cells in the primary root (Péret et al., 2012). AtLAX3 is expressed in the columella and stele of the primary root and it is involved in lateral root development, as Arabidopsis lax3 mutants show delayed lateral root emergence (Swarup et al., 2008). No root growth–related defects or lateral root–related defects are observed in either lax1 or lax2 single mutants while aux1lax3 double mutant shows a severe reduction in the number of emerged lateral roots (Swarup et al., 2008). Auxin binding and import activity of AUX/LAX proteins has been demonstrated using an oocyte expression system for AtAUX1 (Yang et al., 2006), AtLAX3, and AtLAX1 or a yeast-based heterologous expression system in the case of AtLAX2 (Yang et al., 2006; Carrier et al., 2008; Péret et al., 2012).

PIN-LIKES represent the most recently characterized family of plant auxin transport proteins and include seven members in Arabidopsis. PILS show low (10–18%) sequence identity with PINs and they are all capable of transporting auxin across the PM in heterologous systems (Barbez et al., 2012). PILS regulate intracellular auxin accumulation at the ER and thus reduce the availability for free auxin that can reach the nucleus, possibly exerting a role in auxin signaling that is comparable to that of AtPIN5 (Barbez et al., 2012; Barbez and Kleine-Vehn, 2013). The PILS family is conserved throughout the plant lineage, having representatives in several taxa including unicellular algae, where PINs have not been found yet. This indicates that PILS could be evolutionarily older than PINs (Feraru et al., 2012; Viaene et al., 2013). Six PILS have been identified in rice, 10 in maize, 7 in sorghum, and 8 in Brachypodium (Figure 3; Feraru et al., 2012). Forestan et al. (2012) identified two maize proteins that in sequence comparison analysis do not cluster with Arabidopsis PINs: ZmPINX and ZmPINY. Our sequence comparison verified that the two proteins are more similar to PILS rather than PINs, as previously hypothesized (Figure 3). Expression analysis for these genes shows that they are ubiquitously expressed and differentially up-regulated in maize organs. In detail, ZmPINX is up-regulated in root apex and male and female inflorescences, while ZmPINY is highly expressed during kernel development (Forestan et al., 2012).

Auxin has a fundamental role in plant organs formation and its polar transport across cellular membranes is crucial for the correct development and response to external stimuli. Master regulators of PAT are auxin transport proteins, which have been extensively studied in Arabidopsis but not in other species, mainly due to the difficulty to obtain loss-of-function mutants. In monocots, only a few of these transporters have been characterized, mainly in rice and maize and most of the information available has been obtained by expression analyses without functional characterization. There are substantial divergences in development and plant structure between monocots and dicots. Differences are present in seed, vascular system, and leaf developmental programs (Tsiantis, 1999; Scarpella and Meijer, 2004; Coudert et al., 2010; Sreenivasulu and Wobus, 2013). The monocot root system architecture and cellular organization also differ considerably from those of dicots (Hochholdinger et al., 2004; Smith and De Smet, 2012). In addition, monocots have a segmented stem as opposed to the unsegmented stem of dicots. Auxin transporter families are larger in monocots allowing for the possibility of functional redundancy, but also for neo- and sub-functionalization of specific proteins. Monocot-specific and organ-specific proteins exist and they have a distinct role in regulating auxin driven organ development (PIN9). In some cases, alterations in PAT result in interesting new traits, such as dwarfism in maize and sorghum br2 and dw3 mutants respectively, which can be exploited to generate more productive lines through breeding programs. Moreover, many more short-statured mutants exist in maize that may have defects in auxin transport, although none of these mutants have been characterized in any detail. Interestingly, quite a few of these mutants exhibit dominant inheritance (Johal, unpublished) that makes them interesting in at least two ways. First, they can help side step gene redundancy problems and allow the functional exploration of additional genes. Second, they can be used in MAGIC (mutant-assisted gene identification and characterization)-based enhancer suppressor screens to unveil natural variation in a trait of interest (Johal et al., 2008). Even transgenic reporters for auxin activity can be used in lieu of bona fide dominant mutants for such MAGIC screens. The traditional enhancer/suppressor screens based on mutagenesis can also be employed to identify additional genes that encode auxin transporters. One such resource already exists in sorghum, where a line carrying a dw3 mutation in SbABCB1 was EMS mutagenized to produce and sequence M2 populations for both forward and reverse genetics (Krothapalli et al., 2013). These M2 populations can be screened to identify other genes in the network with the ability to suppress or enhance the dw3 phenotype. Finally, there is the exciting possibility of using new genome editing and reverse genetics tools such as CRISPR/Cas9, which has been shown to work in rice and maize (Miao et al., 2013; Liang et al., 2014). Technologies like this can be used to alter the expression and function of genes encoding auxin transporters in monocots and this may lead to important new breakthroughs in our understanding of their roles in development and response to the environment.

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