The conversion of SAM to ACC by ACS is the major regulatory step in ethylene biosynthesis, hence all conditions which ultimately lead to ethylene formation cause an accumulation of ACC (Yang and Hoffman, 1984). As mentioned previously, ACS is a member of a superfamily of proteins requiring pyridoxal-5′-phosphate (PLP) as a cofactor, known as PLP-dependent proteins. In most plant species, ACS is encoded by a multigene family, the members of which show distinct but overlapping expression patterns. The Arabidopsis genome contains eight ACS genes encoding functionally active ACS enzymes (ACS2, ACS4-9, and ACS11), and a ninth catalytically inactive member, ACS1 (Liang et al., 1992; Van Der Straeten et al., 1992; ACS1 in the latter reference corresponds to ACS2 in the former). ACS mutant complementation by Tarun and Theologis (1998) suggested that the enzyme functions as a homodimer whose active site is formed through interaction of shared residues from the monomeric subunits. Capitani et al. (1999) confirmed this mode of action by study of the quaternary structure of an apple ACS protein. Besides functioning as homodimers, it was suggested that ACSs are also capable of forming heterodimers (Tarun and Theologis, 1998), which was corroborated by Tsuchisaka and Theologis (2004a). The latter study revealed that the ACS proteins can only form enzymatically active heterodimers among members of the same phylogenetic branch, while all functional ACS are homodimers with shared active sites. The capability of acting as homo- or hetero-dimers is a characteristic also present in other PLP-dependent enzymes.

In most plant species, the members of the ACS gene family are differentially regulated at the transcriptional level, in an organ-specific, tissue-specific and/or cell-type specific manner (Kende, 1993; De Paepe and Van Der Straeten, 2005). To date, most gene expression studies have been performed on an organ or tissue specific basis. Rodrigues-Pousada et al. (1993) performed a complete analysis of ACS1 (termed ACS2 in Liang et al., 1992) gene expression during Arabidopsis development using a GUS-reporter construct. Ishiki et al. (2000) investigated the ACS gene expression of etiolated melon seedlings (root, hypocotyl, and cotyledons) and melon fruit, Peng et al. (2005) investigated the Arabidopsis ACS gene expression in the root and the shoot, and Xue et al. (2008) determined ACS expression patterns in different rose floral tissues (sepals, petals, stamens, gynoecia, and receptacles). However less abundant, a number of examples can be found of cell type-specific gene expression analyses. Geisler-Lee et al. (2010) studied the ACS gene expression in different cell types in maize roots. Tsuchisaka and Theologis (2004b) looked both at the organ specific and cell-type specific Arabidopsis ACS expression patterns. In seedlings, similar but not identical expression patterns were found in the light and in darkness. Interestingly, no expression of ACS9 is observed in both conditions, while in the light ACS8 is the only gene expressed in the root tip. In mature plants, ACS1 is mainly expressed in vascular tissue, ACS2, 4, 5, 6, and 8 are expressed in roots, inflorescence stem, siliques and younger leaves, ACS11 is expressed in roots, inflorescence stem, younger leaves, cauline leaves, and ACS9 is barely expressed. In their cell type-specific expression studies they analyzed tissues from cotyledons, the hypocotyl and the root. In cotyledons and the hypocotyl, expression of all genes, except ACS1 and ACS9, is restricted to the epidermal cell layer, the vascular bundles and the guard cells. In the roots, three different zones were examined: lateral root cap, cell division, and cell expansion. ACS8 is uniquely expressed in the lateral root cap zone. In the other two zones, most ACS expression is restricted to the endodermis, pericycle, and stele. However, IAA treatment enhanced the expression of ACS7 and ACS8 in the epidermis and of ACS11 in all cell types. Using data from GeneAtlas and the AREXdb, these expression patterns were confirmed by Dugardeyn et al. (2008); however, they also revealed expression in the root cap for ACS2, 4, 8, 10, and 12. This discrepancy with the analysis of Tsuchisaka and Theologis (2004b) probably results from limitations of the GUS reporter. Although little attention has been given to the exact mechanisms at the origin of these differences in expression levels, it is established that ACSs are differentially regulated in response to a variety of developmental or environmental signals and in response to plant hormones.

A first example illustrates how ACS genes in tomato are differentially regulated during fruit ripening. Similar to Arabidopsis, tomato ACS genes are part of a multigene family with differential expression patterns regulated by developmental, biotic and abiotic signals. When ripening starts in tomato or other climacteric fruits, the regulation of ethylene biosynthesis switches from auto-inhibitory to auto-stimulatory. In these plant species, two systems of ethylene production have been proposed (Barry et al., 2000; Alexander and Grierson, 2002). System 1 operates during vegetative growth, during which ethylene inhibits its own biosynthesis (auto-inhibition), while system 2 operates upon fruit ripening, during which ethylene induces its own biosynthesis (auto-catalysis). Barry et al. (2000) observed that the transcripts corresponding to four ACS genes, LE-ACS1A, LE-ACS2, LE-ACS4, and LE-ACS6, were detected in tomato fruit. A detailed expression analysis using the wild type and ripening mutants revealed that each of these ACS genes is regulated in a unique way. System 1 ethylene biosynthesis is regulated by the expression of LE-ACS1A and LE-ACS6 in green fruit, followed by a transition period in which the MADS-Box Transcription Factor RIN (RIPENING INHIBITOR) plays an important role in increasing LE-ACS1A expression and inducing LE-ACS4 expression (Figure 2). The increased ethylene production triggers the expression of LE-ACS2. Subsequently, autocatalytic system 2 is initiated. The elevated ethylene production results in a negative feedback on the system 1 pathway, reducing LE-ACS1A and LE-ACS6 expression. System 2 is maintained by expression of LE-ACS2 and LE-ACS4 (Lincoln et al., 1993; Barry et al., 2000).

The next few examples illustrate that the expression of ACS genes show different susceptibilities to other plant hormones, and that these susceptibilities can differ between different cell types. Rodrigues-Pousada et al. (1999) provided the first evidence for transcriptional regulation of ACS1 by auxin and cytokinin. Yamagami et al. (2003) demonstrated that indole-3-acetic acid (IAA) induces six ACS genes (ACS2, 4, 5, 6, 8, 11) in 7 days old etiolated and light grown seedlings. These results were confirmed by Tsuchisaka and Theologis (2004b), who also observed an effect of IAA on the expression of ACS7. The effect of IAA on different ACS genes differed between cell types. For example, IAA induced expression of ACS8 and ACS11 in all cell layers of the cell expansion zone of the root, while it activated ACS5 expression only in the endodermis. Dugardeyn et al. (2008) revealed that, with the exception of ACS4, the same genes are upregulated by gibberellins (GAs). Nemhauser et al. (2006) showed that ACS6 is up-regulated by ABA and brassinolide (BL), and that ACS10 is down-regulated by IAA, gibberellic acid (GA), and ABA, as well as by methyl jasmonate (MeJA). Finally, Wang et al. (2005) showed that ACS4 and ACS5 are also up-regulated by ABA, and that ACS7 is upregulated by GA and ABA.

While the transcriptional regulation of ACS genes is of great importance, post-translational regulation of ACS by protein degradation also plays a major role in the regulation of ethylene biosynthesis (see Figure 3). The C-termini, however, unimportant for enzyme activity, were found to play a crucial regulatory role in enzyme stability for proteasomal degradation. The Arabidopsis ACSs are classified into three groups based on the presence or absence of phosphorylation sites in these C-termini (Liang et al., 1992; Chae and Kieber, 2005). The type I ACS enzymes, ACS2 and ACS6, have a C- terminus containing phosphorylation sites for both MPKs (mitogen-activated protein kinases; Liu and Zhang, 2004) and CDPKs (calcium-dependent protein kinases; Tatsuki and Mori, 2001). The type II ACS enzymes, ACS4, ACS5, ACS8, ACS9, and ACS11, have a C-terminus containing only phosphorylation sites for CDPKs. The type III ACS enzyme ACS7 has a C-terminus without any phosphorylation sites. Phosphorylation and/or dephosphorylation of ACS proteins has a severe impact on ethylene biosynthesis. An increased phosphorylation stabilizes the ACS protein, while dephosphorylation leads to an increased proteasomal degradation (Spanu et al., 1994; Liu and Zhang, 2004; Skottke et al., 2011; Xu and Zhang, 2014).

1-aminocyclopropane-1-carboxylic acid synthase stability is controlled by two types of kinases, Ca²⁺-dependent protein kinase (CDPK) and mitogen-activated protein kinases. Types I and II ACS proteins are phosphorylated by CDPKs, as in both groups CDPK phosphorylation motifs were identified in their C-termini (Tatsuki and Mori, 2001; Sebastia et al., 2004). Type I ACS proteins are dephosphorylated by protein phosphatase 2A (PP2A) and Protein phosphatase 2C (PP2C), decreasing their protein stability (Skottke et al., 2011; Ludwikow et al., 2014). Additionally, the activity of the type I ACSs can be increased by phosphorylation by the MAP kinases MPK3 and MPK6, which have been shown to be pathogen/stress-activated (Liu and Zhang, 2004; Joo et al., 2008; Han et al., 2010). Although there are no CDPK phosphorylation sites found in the C-terminus of the type III ACS7 protein, it was recently shown that ACS7 can be phosphorylated in vitro in its catalytic domain (Huang et al., 2013). The E3 ubiquitin ligase targeting type-I ACSs for proteasomal degradation has not been identified yet. The degradation of the type-II ACS enzymes is mediated by ETHYLENE-OVERPRODUCER1 (ETO1) and ETO1-like 1/2 (EOL1/2), substrate-specific adaptor proteins of a Cullin4-based E3 ubiquitin ligase complex (Chae et al., 2003; Wang et al., 2004; Christians et al., 2009). The current model is that the phosphorylation of the type II proteins blocks the ability of the ETO1 and EOL1/2 proteins to bind, inhibiting the degradation of these ACSs. The degradation of the type-III ACS7 enzyme and the type-II ACS4 enzyme is mediated by XB3 ORTHOLOG 2 IN ARABIDOPSIS THALIANA (XBAT32), a RING-type E3 ligase (Prasad et al., 2010; Lyzenga et al., 2012).

Cytokinin is one of the hormones capable of regulating ethylene biosynthesis, as reflected by the elevated ethylene biosynthesis after treatment of etiolated and light grown seedlings with cytokinin (Vogel et al., 1998a,b; Woeste et al., 1999). Cytokinin does not elevate ACS transcript levels, but decreases the rapid degradation of ACS4 and ACS5 (Vogel et al., 1998b; Woeste et al., 1999; Chae et al., 2003). Brassinosteroids elevate ethylene biosynthesis in a similar fashion by increasing the stability of ACS5 and ACS9, and its effect is additive with that of cytokinin (Hansen et al., 2009). The stabilization of these ACSs by cytokinins and brassinosteroids includes the inhibition of the C-terminus-dependent targeting by ETO1 or EOL1/2; however, a C-terminus-independent regulation was also suggested. Hence, ACSs are regulated by other plant hormones through regulatory signals that can act together to continuously adjust ethylene biosynthesis in the different plant tissues and in response to signals from the environment.

While the conversion of SAM into ACC by ACS is known to be the major rate limiting step in ethylene biosynthesis, the enzyme converting ACC into ethylene (ACO) can become rate limiting under conditions of high ethylene production such as fruit ripening (Barry et al., 1996; Nakatsuka et al., 1997; Van de Poel et al., 2012; Rudus et al., 2013). In Arabidopsis, ACO is encoded by a small gene family of five members (ACO1, ACO2, ACO3, EFE, and ACO5). Dugardeyn et al. (2008) investigated the expression patterns of the Arabidopsis ACO genes using Geneatlas and AREXdb. On an organ-specific basis, ACO1 is present at low levels in all plant organs, while ACO2 is highly expressed allover, and EFE shows a strong expression in leaves. With respect to cell type- specific expression, all three genes are strongly expressed in the columella, while ACO1 is also strongly represented in the lateral root cap, and the expression of ACO2 is the highest in the vascular tissue of the fast elongation and differentiation zone. Though appearing in almost all vegetative and reproductive tissues, there are differences in accumulation of the members of the ACO gene family during different developmental and physiological processes (Barry et al., 1996; Rudus et al., 2013). In contrast to ACS, much less is known about the transcriptional and post-translational regulation of ACO genes. Similar to the ACSs, ACO expression can also be regulated by plant hormones, as reported for salicylic acid, auxins, ABA, and gibberellins (Leslie and Romani, 1988; Huang et al., 1993; Ogawa et al., 2003; Zhang et al., 2009).

Additional to the regulation of ethylene biosynthesis through ACS and ACO, ethylene biosynthesis is also regulated by the capability of forming ACC derivates (see Figure 4). Three types of conjugates have been identified so far; however, knowledge on the importance and function of these conjugates remains poor. The first ACC conjugate, N-malonyl-ACC (MACC), was isolated and identified independently by two research teams (Amrhein et al., 1981; Hoffman et al., 1982; Peiser and Yang, 1998). The conjugation of ACC into MACC is catalyzed by the enzyme ACC-N-malonyl transferase (AMT), a reaction requiring malonyl-coenzyme-A (Martin et al., 1995; Peiser and Yang, 1998). MACC can be translocated between the cytosol and the vacuole by ATP-dependent tonoplast carriers (Bouzayen et al., 1988, 1989; Tophof et al., 1989) suggesting that MACC formation and storage in the vacuoles might be important to control the pool of available ACC. This hypothesis was further strengthened by the observation from Liu et al. (1985) and Martin et al. (1995) that the production of MACC by AMT can be induced by exogenous ethylene treatments during the ripening of preclimacteric tomato (Liu et al., 1985; Martin et al., 1995). Furthermore, Jiao et al. (1986) and Hanley et al. (1989) demonstrated that MACC can be reconverted into ACC. A second conjugate that can be formed from ACC is γ-glutamyl-ACC (GACC; Martin et al., 1995). The formation of GACC is catalyzed by the enzyme γ-glutamyl transpeptidase (GGT), a reaction requiring glutathione (GSH; Martin et al., 1995; Peiser and Yang, 1998). A third ACC-conjugate is jasmonyl-ACC (JA-ACC), which is formed through the activity of the enzyme jasmonic acid resistance 1 (JAR1; Staswick and Tiryaki, 2004). This linkage between the precursors of ethylene and the active form of JA, isoleucinoyl-JA, might be a mechanism to control both hormones.

The pool of available ACC can also be reduced by the irreversible deamination of ACC (See Figure 4). Importantly, this reaction is not only plant-borne but also carried out by certain plant growth-promoting rhizobacteria (PGPR). Honma and Shimomura (1978) identified an ACC deaminase (ACD) in Pseudomonas sp. strain ACP. Plants are capable of releasing ACC into the rhizosphere to attract these PGPR which use ACC as a carbon and nitrogen source (Glick et al., 1998; Penrose and Glick, 2001; Penrose et al., 2001). As a consequence, the available ACC pool, and thus also plant ethylene production; is reduced; hence, plant growth is promoted through the plant–bacterium interaction (Glick et al., 2007a,b). In a similar manner, the reduction of the ACC pool by PGPR helps the plant to cope with stresses such as flooding (Grichko and Glick, 2001a), pathogen attack (Robison et al., 2001), or salinity (Ali et al., 2014). McDonnell et al. (2009) identified the first plant-encoded ACC deaminase and demonstrated its importance in regulating the ethylene balance.

For many hormones, the site of synthesis does not always coincide with the site of action. The same has been observed for ethylene. Because ethylene is a gaseous molecule, it diffuses rapidly through the plant tissues inducing mainly local responses, with the exception of aerenchyma through which long distance transport can be conducted, as observed in conditions of waterlogging. In contrast, long distance ethylene signaling between different plant tissues mostly occurs by the transport of ACC.

Because of its role as a stress hormone, ethylene signaling in response to biotic and abiotic stresses has been investigated thoroughly, often including a role for ACC transport. Root to shoot transport of an ethylene signal in response to waterlogging or submergence is a prime example of the importance of ACC transport and has been studies multiple times (Bradford and Yang, 1980; Metraux and Kende, 1983; Zarembinski and Theologis, 1993; Else et al., 1995; Vartapetian and Jackson, 1997; Else and Jackson, 1998; Grichko and Glick, 2001b; Almeida et al., 2003; Vriezen et al., 2003; Jackson, 2008; Dawood et al., 2016).

When plants are waterlogged or submerged, the available oxygen levels in roots drop rapidly, triggering a myriad of effects on plant metabolism. First of all, it induces a transcriptional cascade targeting genes involved in anaerobic metabolism and survival (Gibbs et al., 2011). Secondly, it induces the expression of several ACSs in the roots (Olson et al., 1993, 1995; Zarembinski and Theologis, 1993; Van Der Straeten et al., 1997; Shiu et al., 1998; Rieu et al., 2005). At the same time the conversion of ACC into ethylene is suppressed, because ACO requires oxygen. After root to shoot transport, ACC promotes the expression of ACO genes, stabilizes ACO mRNA, and increases the activity of ACOs already present, resulting in an increased ethylene production in the shoots (English et al., 1995). Differential ACS and ACO expression patterns in response flooding or hypoxia have been observed in Arabidopsis (Peng et al., 2005) and maize (Geisler-Lee et al., 2010). Furthermore, waterlogging and submergence leads to an overall accumulation of MACC and GACC in both roots and shoots (Amrhein et al., 1981).

The importance of ACC transport in response to waterlogging or submergence can be illustrated by research conducted in tomato (Bradford and Yang, 1980; Amrhein et al., 1981; Else et al., 1995; English et al., 1995; Else and Jackson, 1998) and in rice plants (Metraux and Kende, 1983; Zarembinski and Theologis, 1993; Kende et al., 1998; Almeida et al., 2003; Vriezen et al., 2003). Bradford and Yang (1980) showed that ACC is synthesized in tomato plant roots and then transported from hypoxic roots through the xylem to the shoots, where it is rapidly converted to ethylene inducing leaf epinasty. Furthermore, they observed that drainage results in a simultaneous decrease in ACC flux and ethylene production, and that the petioles are dependent upon the ACC flux for the high rates of ethylene synthesis. Besides epinasty, some of the other responses of plants to waterlogging and submergence, such as reduced root permeability, closure of stomata, development of aerenchyma and adventitious roots, and premature fruit drop might be directly or indirectly related to the enhancement of ACC or derivatives, besides ethylene (Vartapetian and Jackson, 1997; Grichko and Glick, 2001b; Sasidharan and Voesenek, 2015). Amrhein et al. (1982) observed that ACC can also be transported through the phloem. Based on these findings, Morris and Larcombe (1995) assessed phloem transport by foliar application of radioactive ACC to the leaves of 21 day-old cotton plants. Total radioactivity was measured over time in the stem, the hypocotyl and the apex and the patterns were consistent with phloem-mediated transport of the radioactive ACC. The authors suggested that part of the ACC which accumulates in leaves after transport through the xylem from stressed roots (Bradford and Yang, 1980) may be re-exported to other organs through the phloem. Despite the observed overall increase in the formation of the ACC conjugates MACC and GACC (Amrhein et al., 1981), leaves are apparently not capable of exporting MACC to the phloem (Morris and Larcombe, 1995). This could be a consequence of translocation of MACC to the vacuoles of leaf cells (Bouzayen et al., 1988; Tophof et al., 1989).

1-aminocyclopropane-1-carboxylic acid transport is not only important during stress conditions. Several studies evaluated the gene expression patterns of ACS and/or ACO during different developmental processes and in different organs or cell types. Woltering and associates investigated the transport of ACC in Cymbidium orchid flowers for the coordination of senescence (Woltering, 1990). They revealed that both endogenous and exogenously applied ACC is rapidly transported from the site of production/application to other flower parts. Jones and associates analyzed ethylene production and ACS and ACO expression patterns in different floral organs of carnation flower (Dianthus caryophyllus L.) in response to pollination. Ethylene production was observed in the styles, petals, and ovaries. High ACS expression levels were found in petals and styles, while ACO expression was highest in the ovaries (Jones and Woodson, 1997, 1999). In Arabidopsis, Dugardeyn et al. (2008) observed little ACO expression in the root meristematic zone, while several ACS genes showed high expression in this zone (ACS3, 4, 10, and 12). The epidermis and endodermis of the elongation/differentiation zone also showed little ACO expression, while different ACS genes were induced. Gallie et al. (2009) investigated the cell-type specific gene expression patterns of maize root cells and found that ACS was only expressed in the root cortex, while ACO was mainly expressed in the protophloem sieve elements, suggesting ACC transport between these cell types. In all of the above-mentioned studies, the spatial differentiality in ACS and ACO expression within a given organ suggests that ACC is transported from the site of synthesis to the site of conversion to ethylene.

Shortly after the identification of ACC, the occurrence of intercellular ACC transport over short distances and intracellular compartmentalization of ACC in the vacuole was proven. Lurssen (1981) investigated ACC uptake in soybean leaf disks by measuring ethylene production after incubation with ACC. Pretreatment with ACC, followed by treatment with non-polar amino acids resulted in a reduced ethylene production, suggesting that ACC is transported by an amino acid transport system. Saftner and Baker (1987) observed the uptake of radiolabeled ACC and α-amino-iso-butyric acid (αAIB, structural analog of ACC) in tomato pericarp slices. Finally, both Tophof et al. (1989) and Saftner and Martin (1993) demonstrated the transport of ACC across the tonoplast in mesophyll cells in three different plant species (barley, wheat, and maize).

In a recent study, a first plant ACC transporter was found. Shin et al. (2015) identified an Arabidopsis thaliana mutant, are2 (ACC-resistant2), showing a dose-dependent resistance to exogenously applied ACC. In the light, the wild type shows an elongated hypocotyl and shortened root growth in the presence of 1 μM exogenous ACC (Smalle et al., 1997), while the are2 mutant exhibited a reversed phenotype, with a shorter hypocotyl and more elongated roots. When grown in darkness in the presence of 1 μM ACC the hypocotyls and the roots of the are2 mutant were not significantly inhibited as compared to the wild type. However, at elevated concentrations of ACC (10 μM or above) a triple response was observed, indicating sensitivity of the are2 mutant (Figure 5). Interestingly, the mutant displayed normal sensitivity to ethylene, and the synthesis of ethylene was not impaired, neither did the insensitivity to ACC affect sensitivity of the seedlings to other plant hormones. The are2 mutant corresponded to a null allele of LHT1 (LYSINE HISTIDINE TRANSPORTER 1), a plasma membrane (PM)-localized amino acid transporter. LHT was first identified by Chen and Bush (1997), and further characterized by Hirner et al. (2006). To confirm that are2 and LHT1 are allelic, Shin et al. (2015) performed an allelism test with the lht1-5 mutant. The double mutant and the single mutants showed the same ACC resistance and early leaf senescence phenotypes. An uptake experiment using radiolabeled ACC and protoplasts from wild type and are2/lht1-5 mutant leaf mesophyll cells, showed a reduction of 40% in ACC uptake in both mutants, supporting a function of LHT1 in ACC uptake transport. In Arabidopsis, LHT1 is part of a family of amino acid transporters with 10 members (AtLHT1-10), all of which are localized to the plasma membrane and transport a broad spectrum of amino acids from the cell wall space inward (Hirner et al., 2006). The LHT genes show distinct and overlapping expression patterns, indicating that LHT1 homologs may function in ACC uptake under certain developmental or environmental conditions (Foster et al., 2008). Based on unpublished results, Shin et al. (2015) suggest that at least one LHT1 homolog could complement the lht1 mutant.

The identification of LHT1 as an ACC transporter is a cornerstone in unraveling the mechanism and regulation of ACC transport in plants. Giving the complexity of ethylene biosynthesis and signaling, and the evidence that ACC can be transported both over long distances between plant organs and over short distances within or between cells, one can assume that ACC transport involves several proteins, presumably even different transport systems. It is thus quite possible that many more ACC transporters are yet to be identified. Performing uptake experiments of the are2 mutant using radiolabeled ACC, Shin et al. (2015) observed an incomplete reduction of ACC uptake, supporting the presence of additional ACC transporters.

Aiming at a more profound understanding of ACC transport, it is tempting to look at the organ and cell-type specific gene expression patterns of the ethylene biosynthesis genes ACS and ACO as a main determinant of ACC accumulation. In tissues with high ACS expression, where, however, little or no ACO expression is found, ACC export might be important. In contrast, in tissues with low ACS expression, though high ACO expression, ACC import might be important. For an overview of the detailed expression patterns for both enzyme families, we extracted cell-type specific and developmental stage specific absolute expression values from the Toronto Arabidopsis eFP browser (BAR). The relative expression percentages, as presented in Figures 6 and 7, were calculated for each gene, relative to the expression value of the cell-type with the highest absolute value.

Figure 6 presents an overview of the cell type-specific expression levels in the roots, the hypocotyl and the leaves. Little ACO expression is found in the root epidermal atrichoblasts, hypocotyl xylem, and leaf guard cells, while some ACSs are highly expressed in these cell-types. This suggests that the ACC produced is most likely transported to other tissues (with higher ACO expression) where it is converted into ethylene. It has to be acknowledged, however, that these expression patterns do not give much information about long distance or intracellular transport of ACC. Figure 7 presents an overview of the relative expression levels for certain tissues of different developmental stages. These expression patterns show in which developmental tissues ethylene production is important, but they also indicate between which tissues the translocation of ACC might occur. For example, in young flowers the ACOs are barely expressed, while in mature flowers increased expression levels are observed for both ACO2 and EFE. This means that ethylene production is most likely higher in mature flowers. In addition, there is little difference in the expression levels of the ACSs between these two developmental stages. This might suggest that the increased ethylene production in mature flowers is dependent on the supply of ACC from other tissues. In pollen, high levels of ACS expression can be found, while there is almost no ACO expression. This discrepancy corresponds with the observation that high levels of ACC are present in mature pollen, and that this pollen-ACC may be an important mediator of the early response of flowers to pollination (Whitehead et al., 1983; Reid et al., 1984). This was confirmed by (Reid et al., 1984), with the observation that application of ACC to the stigmas of carnation flowers (Dianthus caryophyllus), causes an initial increase in gynoecium and petal ethylene production similar to that reported for pollinated flowers, and that application of [2-¹⁴C]ACC to the stigmas resulted in radioactive ethylene production both by the gynoecia and the petals. This report provides additional evidence that during pollination (and post-pollination) ACC is exported from the pollen, and transported toward different flower tissues, where it is converted into ethylene.

The ACS and ACO gene expression patterns as described above can be compared with those of known transporters to find novel ACC transporters. There are two major paths that can be followed here. On the one hand, one can look at members of general transporter families such as the family of ATP binding cassette (ABC) transporters, of which some members are important for the transport of IAA, ABA, CK, and SL (Xu et al., 2014; Borghi et al., 2015). One the other hand one can look at known amino acid transporters. The Arabidopsis genome contains 60 or more predicted amino acid transporters, belonging to at least two large families.

One of these amino acid transporter families is the LHT family. Previous research already showed that LHT1 is expressed in roots and mesophyll cells, and that LHT2 expression is localized to the tapetum of anthers (Chen and Bush, 1997; Lee and Tegeder, 2004). Figures 6 and 7 present an overview of the cell type-specific and development specific expression patterns for the Arabidopsis LHTs. LHT1 expression is high in epidermal atrichoblasts and guard cells, two cell types in which ACC transport was suggested to be important (see above). This is in part a confirmation that in these tissues ACC transport could be supported by LHT1 action. LHT2, LHT5, and LHT7 also show high expression in guard cells, while LHT8 and LHT10 show high expression in the hypocotyl xylem. In the pollen, where ACC export might be essential for the induction of pollination, LHT2, 5, 8, and 9 are highly expressed. These results reveal that certain other members of the LHT family might also be linked with ACC transport. However, as they are also amino acid transporters, these expression patterns might not necessarily mean that this is the case.

Because ACC is a non-protein amino acid, it seems evident that ACC transporter(s) would have the characteristics of an amino acid transporter, or be transported by a protein that serves in transport of other amino acids as well. In an assay measuring ethylene upon incubation of leaf disks with ACC, Lurssen (1981) tested competition of other amino acids with ACC. Uptake of ACC was inhibited by L-amino acids with non-polar side chains such as L-methionine, L-tyrosine, and L-phenylalanine, but not by polar amino acids. These were the first indications that ACC might be transported by an amino acid transporter. Saftner and Baker (1987) investigated ACC transport in tomato pericarp slices, and observed that the uptake was inhibited by neutral amino acids, with the highest inhibition from methionine. Given that ACC uptake can be inhibited by the competition of other amino acids, it might be of interest to investigate whether other known amino acid transporters are also capable of transporting ACC, for instance by a systematic analysis using heterologous expression in yeast or in protoplasts. ACC transport could be assayed using radiolabeled ACC or by measuring ethylene production.

Besides comparing gene expression patterns of amino acid transporters, as was suggested in the previous section, another approach is to look at the amino acid substrates of the different transporters. As discussed earlier, ACC transport can be suppressed by neutral and/or non-polar amino acids (Lurssen, 1981; Saftner and Baker, 1987). Amino acid transporters having a high affinity for one or more of these amino acids might be considered potential candidates for ACC transport.

The studies discussed in the previous two sections assume that novel ACC transporters can be found in families of known transporters. It is, however, quite possible that some ACC transporters remain uncharacterized to date. For identification of novel transporters several approaches could be taken. Most of our current knowledge on plant transporters has been obtained by use of the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe (Dreyer et al., 1999). Using one of these heterologous expression systems, cDNA libraries can be screened to identify novel ACC transporters. An alternative approach is a forward genetic screen using the chemical mutagen ethyl methanesulfonate (EMS). For this forward genetic screen, the phenotypic analysis might include the screening for ACC resistance (dose-dependent or not) as seen by Shin et al. (2015) for are2/lht1-5.

Several studies explored the possibility of transport of ACC conjugates. Fuhrer and Fuhrer-Fries (1985) investigated the formation and the transport of ACC and MACC in pea plants after wounding, which normally induces an increase of ethylene biosynthesis. They concluded that MACC could be transported from the shoot to the root and that the roots act as an MACC sink. However, further investigation is needed to rule out the possibility that the MACC increase at the roots may result from de novo MACC production after ACC transport from the shoot to the roots. Furthermore, the MACC in the shoot could be reconverted into ACC. Finlayson et al. (1991) investigated the transport and metabolism of ACC in sunflower seedlings. This study presents evidence for the presence of MACC in xylem sap of seedlings treated with radiolabeled ACC. However, the presence of MACC in the xylem sap could not be confirmed by GC-MS analysis of untreated seedlings. These results suggest that MACC can be transported through the xylem, but that this transport might only occur during certain stress conditions. Morris and Larcombe (1995) investigated the possibility of MACC transport through the phloem in cotton. In contrast to the previously discussed studies, they found no evidence for MACC transport.

Overall it is clear that additional evidence is needed to confirm a more general nature of MACC transport, and add to our understanding of its biological function in plants. Moreover, it is also important to further investigate whether the other ACC conjugates can be transported, or whether they rather function as a storage forms of ACC to regulate ethylene production within the organ or tissue where they are formed.