Targeted expression of SbMATE in the root distal transition zone is responsible for sorghum aluminum resistance


For correspondence (e-mail


Aluminum (Al) toxicity is one of the major limiting factors for crop production on acid soils that comprise significant portions of the world's lands. Aluminum resistance in the cereal crop Sorghum bicolor is mainly achieved by Al-activated root apical citrate exudation, which is mediated by the plasma membrane localized citrate efflux transporter encoded by SbMATE. Here we precisely localize tissue- and cell-specific Al toxicity responses as well as SbMATE gene and protein expression in root tips of an Al-resistant near-isogenic line (NIL). We found that Al induced the greatest cell damage and generation of reactive oxygen species specifically in the root distal transition zone (DTZ), a region 1–3 mm behind the root tip where transition from cell division to cell elongation occurs. These findings indicate that the root DTZ is the primary region of root Al stress. Furthermore, Al-induced SbMATE gene and protein expression were specifically localized to the epidermal and outer cortical cell layers of the DTZ in the Al-resistant NIL, and the process was precisely coincident with the time course of Al induction of SbMATE expression and the onset of the recovery of roots from Al-induced damage. These findings show that SbMATE gene and protein expression are induced when and where the root cells experience the greatest Al stress. Hence, Al-resistant sorghum plants have evolved an effective strategy to precisely localize root citrate exudation to the specific site of greatest Al-induced root damage, which minimizes plant carbon loss while maximizing protection of the root cells most susceptible to Al damage.


More than 40% of the world's arable soils are highly acidic (pH < 5), where the toxic aluminum (Al) species, Al3+, is released from soil clays and directly affects plant roots, resulting in stunted and damaged root systems. This leads to significant reductions in crop yields due to drought stress and nutrient deficiencies (Kochian et al., 2004). Crop plants have adopted different strategies to cope with Al stress (Kochian et al., 2004; Horst et al., 2010; Ryan et al., 2010). The best known Al resistance mechanism involves the Al-activated exudation of organic acids (OA) from the growing root tip, where the released OAs chelate Al+, forming non-toxic OA–Al complexes (Kochian et al., 2004; Horst et al., 2010; Ryan et al., 2010). Malate and citrate are the most common OAs released by roots of different plant species to confer Al resistance. For instance, wheat and Arabidopsis roots release malate, while sorghum and maize roots mediate a citrate efflux in response to Al stress. Hence, it is not surprising that the first two Al resistance genes identified in crops were those that encode an Al-activated root malate transporter (TaALMT1) in wheat and Al-activated root citrate transporters in sorghum and barley (SbMATE and HvMATE, respectively) (Furukawa et al., 2007; Magalhaes et al., 2007). Crop plants also respond differently to Al stress (Kochian et al., 2004; Magalhaes et al., 2007; Horst et al., 2010; Ryan et al., 2010). For example, the Al resistance responses are very rapidly induced by Al in maize and Arabidopsis (within 1–2 h) with regards to induction of the expression of genes encoding the OA transporters and the induction of OA efflux from the root tips, while Al induction of Al resistance genes is a much slower process in sorghum (Magalhaes et al., 2007). In sorghum, root growth is inhibited similarly in Al-resistant and Al-sensitive lines during the first 2 days of Al exposure. By the third day of Al treatment, root growth in resistant lines starts to recover from Al-induced damage, and by 5–6 days in Al, root growth is equal to root elongation under control (–Al) conditions. This induction of Al resistance in resistant genotypes mirrors the time course of Al induction of SbMATE gene expression as well as the increase in the rate of citrate exudation by roots. That is, it takes 6 days of Al exposure to elicit the full induction of SbMATE gene expression (Magalhaes et al., 2007). The Al-sensitive genotypes never recover from inhibition of root growth by Al and Al-induced damage to the root tip, which becomes progressively greater over the same 6-day period of Al exposure, and very low levels of expression of SbMATE and Al-activated root citrate exudation are seen over the same time period (Magalhaes et al., 2007).

The root apex is divided into the meristematic zone, which is the site of cell division, and the subsequent zone of root cell elongation (Baluska et al., 1990). Ryan et al. (1993) were the first to show that the wheat root apex is the primary site of Al toxicity, and only when the root apex is exposed to toxic levels of Al is inhibition of root growth observed. Localized application of Al in gel blocks either independently to each of the zones of the maize root tip or to the entire root apex suggested that the distal part of the transition zone (DTZ) was the root growth region most sensitive to Al (Sivaguru and Horst, 1998; Sivaguru et al., 1999). The current detailed microscopic analysis of Al stress in the root shows how localized this zone of Al stress is, and that Al resistance gene/protein expression is specifically localized to a few external cell layers of the same subregion of the root apex.

The objective of this study was to investigate whether the plant mechanisms underlying Al resistance are a more general response that occurs throughout the root tip or if these responses are targeted to very specific root growth zones and specialized cells in this region. We approached this task by determining the spatial localization of Al toxicity responses in the root tip, including root cell damage, generation of reactive oxygen species (ROS), Al accumulation, callose formation (a marker of Al toxicity), localized the expression of SbMATE protein as well as gene expression in root tips of a pair of Al-resistant and Al-sensitive sorghum near-isogenic lines (NIL). Our results demonstrated that sorghum plants have evolved a strategy that enhances the effectiveness of Al resistance based on root citrate exudation, while maximizing the carbon use efficiency of these processes by limiting Al-activated root citrate exudation to the specific root tissues that are most vulnerable to Al stress.


The external cell layers of the DTZ are the most Al-sensitive target for Al toxicity

Aluminum resistance in sorghum is gradually induced by Al stress. It is interesting to observe that the Al-resistant (ATF10B) and Al-sensitive (ATF8B) sorghum NILs show comparable root growth inhibition after 1 day of Al treatment. Subsequently, only the roots of the resistant NIL show improved root growth after 3 days of Al treatment (Figure S1). This response correlates well with the time courses for Al induction of SbMATE gene expression and the rate of Al-activated root citrate release (Magalhaes et al., 2007). To better understand the spatial dynamics of Al toxicity and resistance in sorghum roots, we first investigated Al-induced ROS production in root tips, as this is a well-documented component of Al toxicity (Jones et al., 2006). Figure 1 summarizes these responses in low-magnification images of ROS production and damaged cell/plasma membrane integrity in root tips in response to Al treatment. After 1 day of Al treatment, the highest level of ROS production and propidium iodide (PI) staining was restricted to a band in the root DTZ in both Al-resistant and sensitive NILs (Figure 1b–b″″ and e-e″, respectively). The intensity of the Al-induced ROS signal and the degree of membrane damage was substantially higher in roots from the Al-sensitive line compared with their Al-resistant counterparts. After 3 days of Al treatment, the conspicuous ROS- and PI-stained bands had nearly disappeared from root tips on the Al-resistant plants (Figure 1c–c″). Interestingly, the pattern of Al accumulation by root cells visualized by the Al fluorescent dye lumogallion, yielded a similar banding pattern in roots (Figure 1h,h′), with high Al accumulation in the same root DTZ region 1 day after Al treatment in both the Al-resistant and Al-sensitive lines. However, this band of increased Al accumulation completely disappeared after the 3-day period of Al exposure in roots of the Al-resistant line (Figure 1i). One day after Al treatment, the Al-sensitive line also displayed a similar, but somewhat obscured, Al banding pattern, indicating extensive accumulation of Al in the DTZ (Figure 1k,k′). Damaged cell layers that had accumulated Al were sloughed off from the root apex of the sensitive line after 3 days of Al treatment (Figure 1f–f″), resulting in an absence of the fluorescent signal for Al damage (Figure 1l).

Figure 1.

Aluminum-induced production of reactive oxygen species (ROS), root cell damage and Al-accumulation are localized to the distal part of the root transition zone. Stereomicroscope images depict ROS accumulation (a–f), cell damage (a′–f′), merged images of ROS and cell damage (a″–f″) and Al localization (g–l) along the intact roots of the near-isogenic lines after the indicated Al treatments. b′′′ and b′′′′ are magnified images of b and b″, respectively. Scale bars in a and g represent 1 mm.

Confocal three-dimensional (3D) images of ROS fluorescence in the root apices from the resistant and sensitive NILs showed that while there was no significant production of ROS or membrane damage in the Al-resistant line grown under control (–Al) conditions (Figure S2a–a″), ROS levels were quite high in the swollen root epidermal cells that formed a clear band of ROS fluorescence in the DTZ 1 day after Al treatment (Figure S2b–b″ and Movies S1, S3). Within the DTZ, ROS production was restricted to the epidermal and outer cortical cells (see cross-sectional orthogonal views, Figure S3f). Root cells in the more mature root region beyond the DTZ (see arrow in Figure S2b) exhibited much lower levels of ROS production and there was considerably less cell swelling in this root region compared with the cells in the DTZ (Figure S2b-b″ and Movie S1). No new ROS-stained cells were observed in the root DTZ after 3 days of Al treatment in the resistant NIL (Figure S2c–c″).

Note that root cells that showed more severe membrane damage as visualized by PI staining (Figure S2b′) were found in the root transition zone (TZ) just adjacent to the band of ROS fluorescence in the DTZ (Figure S2b), suggesting that PI staining (breaching of the plasma membrane) occurred after the bursts of ROS production. Figure S3 shows the same region at even higher resolution and better depicts the step-by-step evidence for the sequence of events leading to membrane damage. Reactive oxygen species were produced in both the cytoplasm and nuclei of the external cell layers of the root DTZ as indicated by the green fluorescence labeling (Figure 3a). However, the plasma membrane of the same cells remained intact as no PI staining was observed. In the TZ just behind the DTZ, the level of ROS was lower and breaching of the plasma membrane had occurred as indicated by the PI staining (Figure S3a and a′). These Al-damaged cells retained the PI dye in their nuclei (Figure S3a–a″ and Movie S3). These sequential events were similar in both Al-resistant and Al-sensitive plants, but the extent of swelling and the level of ROS production were significantly higher in the Al-sensitive plants (Figure S3b–b″ and Movies S2, S4).

After 3 days of Al treatment, the ROS-stained band as well as the swollen cells in the DTZ disappeared in the Al-resistant line and ROS labeling became more diffuse in the TZ (Figure S3c–c″, d and Movie S5). These findings are again consistent with activation of Al resistance mechanisms in roots of the resistant line such that the root cells now passing through the DTZ are protected from Al damage (Figure S3c–c″ and Movie S5) compared with the worsening conditions including the sloughing off epidermal cell layers from the roots for the Al-sensitive plants (Figure S3d–d″ and compare Movies S1–S6).

We also visualized callose (1,3-beta-d-glucan) production via staining with aniline blue, which is another established marker for Al stress in roots (Jones et al., 2006). Both callose and the other Al markers were found at high levels in the swollen epidermal cells of the DTZ 1 day after Al treatment in the Al-resistant plants (Figure S4e–h). After 3 days in Al, a reduction in callose, Al accumulation and cell swelling was observed in the DTZ for the Al-resistant line (compare Figure S4e–h and i–l). In the Al-sensitive NIL there were sustained high levels of Al accumulation and callose production in the cells of the DTZ accompanied by extensive cellular swelling (Figure S4m–p).

Al-induced SbMATE gene expression is specifically localized to the epidermal and outer cortical cells of the DTZ

To investigate cell- and tissue-specific expression of SbMATE, we used a laser capture microdissection (LCM) technique to first dissect the root meristem from the TZ and then more specifically dissect the root epidermis/outer cortical layers and the root stele into two different tissues in both the root meristem and the TZ from root apices of the Al-resistant line. Total RNA was extracted from these samples and first-strand cDNA was synthesized, followed by quantitative reverse-transcriptase (RT) real-time PCR analyses (Figure 2Ia–t).

Figure 2.

SbMATE gene expression is localized to the specific external cell layers of the distal root transition zone (DTZ) in the Al resistant sorghum near-isogenic line. The laser capture microdissection (LCM) technique was used with bright field light microscope images. Panel I explains how the individual root cell layers were isolated. Panels II and III depict SbMATE gene expression for these dissected regions and Panel IV shows a schematic representation of the different growth zones within the root apices and the regions where specific cell regions were laser captured for SbMATE mRNA isolation. Panel I: The entire root meristematic zone before capture (a), after capture (b) and the captured segment on the LCM cap (c). The entire DTZ before capture (d) and after capture (e) and the captured segment on the LCM cap (f). The epidermal and outer cortex cells from the meristematic zone before capture (g) and after capture (h), the captured segment on the LCM cap (i) and pooled similar segments from multiple root apices (j). The captured inner cortex and stele cells segments from the meristematic zone after capture (k), the root tip after capture (l) and the pooled similar segments from multiple root apices on the LCM cap (m). The epidermal and outer cortical cells from the DTZ before LCM capture (n) and after capture (o), the captured segment on the LCM cap (p) and pooled similar segments from multiple root apices (q). The captured inner cortex and stele cells from the DTZ (r), the root tip after capture (s) and the pooled similar segments from multiple root apices captured on the LCM cap (t). The scale bar in I(f) represents 500 μm for all microscopic images. Panel II: Quantitative real-time PCR analysis of SbMATE gene expression in the entire meristematic zone and DTZ of the root apex. Panel III: Quantitative real-time PCR analysis of SbMATE gene expression in the epidermis/outer cortex versus the inner cortex/stele from both the meristematic and transition zones in response to the indicated Al treatments. Panel IV: Schematic representation of the different growth zones within the root apex as well as the specific cellular regions that were laser captured for SbMATE mRNA expression analysis. The zones with highest Al impact as well as Al response are highlighted. Standard error bars for quantitative gene expression data in the graphs represent the standard deviation from two independent experiments.

The diagram of the root apex (Figure 2IV) shows the microdissection of the root cellular regions in relation to the specific root growth zones. Our results indicated that under control conditions, although the SbMATE gene was constitutively expressed at a basal level in both the root meristem zone (MZ) and the TZ, the level of SbMATE transcript abundance was about 2.4-fold higher in the TZ than that in the MZ (Figure 2II). Treatment with Al gradually induced SbMATE expression in both the TZ and the MZ and the degree of induction was much greater and occurred more rapidly in the TZ. Compared with the control (–Al treated) roots, the transcript abundance of SbMATE in the TZ increased by about 25% after 1 day of Al treatment, while there was only a 10% increase in the MZ. After 3 days of Al treatment, the level of SbMATE expression increased by about 90% in the TZ compared with a 60% increase in the MZ. By 6 days of Al treatment, the level of SbMATE expression was about 2.5-fold higher in the TZ compared with the MZ. These results indicate that SbMATE is significantly more strongly expressed in the TZ of the root apex. We then looked at the response of SbMATE expression to Al in the epidermis/outer cortex versus inner cortex/stele in both the MZ and TZ. Little change in the level of SbMATE expression was found in the inner cortex and stele in both the TZ and MZ with or without Al treatment. In the epidermis/outer cortex, significantly higher constitutive SbMATE expression was seen, and this expression was strongly induced by Al treatment over the 6-day period (Figure 2III). Taken together, our results indicate that expression of the SbMATE gene in root tips is induced by Al mainly in epidermis and outer cortical cells in the TZ region (Figure 2III).

SbMATE protein is induced by Al in epidermal and outer cortical cells in the DTZ

Consistent with the SbMATE gene expression patterns depicted in Figure 2, immunolocalization analysis indicated that in root tips under control (–Al) growth conditions, SbMATE protein was constitutively expressed at a basal level in the root TZ of the Al-resistant plants and the level of SbMATE protein accumulated in the root epidermal cells was much higher than in other tissues in this region (Figure 3b–b″). After 1 day of Al treatment, substantial swelling of the epidermal and outer cortical cells was apparent in the root TZ in the Al-resistant plant (Figure 3c–c″), an indication of cell damage, and at this point the level of accumulation of SbMATE protein remained largely unchanged in these cells (Figure 3c–c″). However, after 3 days of Al treatment, the level of SbMATE protein was substantially increased in the epidermal and the outer cortical cells in the TZ (compare Figure 3c′ and d′), relative to that in other cells in the MZ and TZ of the root apex (Figure 3c,d). Induction of the expression of SbMATE protein by Al was coincident with the increase in SbMATE gene expression in the TZ, the activation of root citrate exudation by Al in root tips (Furukawa et al., 2007) and the recovery of root cell damage in the DTZ (Figures 1, S1–S4). Moreover, expression of SbMATE protein in the meristem is in general restricted to the epidermal cells, while in the DTZ expression also is extended to the outer layer of cortical cells (Figure 3d′, d″ arrow). In the case of the Al-sensitive line, Al treatment resulted in highly distorted cells in the growth zone and specifically within the DTZ, and considerable sloughing of cells from the root surface resulted in a damaged root tip. This coincided with little or no apparent increase in SbMATE protein expression in these regions after 1 and 3 days of exposure to Al (Figure 3e–h).

Figure 3.

SbMATE protein is highly localized to the external cell layers of the distal root transition zone (DTZ) in the Al-resistant near-isogenic lines (NIL). Optical sections of root apices show the localization of SbMATE protein in the meristematic (a–d) and the DTZ (a′–d″ and e–h) regions after the indicated Al treatments in Al-resistant and Al-sensitive sorghum lines. Western blots of recombinant proteins show the specificity of the SbMATE antibody (Panel I). Note the substantial swelling of epidermal (asterisks in c and c″) and outer cortical cells in the root apex of the Al-resistant line 1 day after Al treatment, which disappears after 3 days of Al treatment and is accompanied by a dramatic increase in SbMATE protein expression in the same epidermal (d, d′) and outer cortical cells (arrow in d″). But the swelling and sloughing off cells in the same regions persisted with no apparent increase in SbMATE protein expression in the Al-sensitive NIL (e–h). The scale bar in D″ represents 50 μm for all images.


Significant progress has been made over the past decade in our understanding of the genetic and molecular mechanisms underlying Al resistance of crops, which includes the identification of several Al-tolerance genes and associated physiological mechanisms (Ryan et al., 1993; Matsumoto, 2000; Horst et al., 2010). The best known and most widely documented Al-resistance mechanism involves exclusion of Al by the root tip mediated by exudation from the roots of OA anions, primarily malate and citrate, that are strong chelators of toxic Al3+ in the rhizosphere (Ryan et al., 1997; Matsumoto, 2000; Kochian et al., 2004; Rengel, 2006). Aluminum resistance in wheat and sorghum is mainly achieved by Al-activated root malate (wheat) and citrate (sorghum) exudation. The efflux of these OA anions from the root tip has been shown to be mediated by OA transporters encoded by the Al-resistance genes TaALMT1 and SbMATE, in wheat and sorghum respectively (Sasaki et al., 2004; Magalhaes et al., 2007). In our previous publication we demonstrated that the SbMATE gene is induced by Al specifically in the root tip of Al-resistant sorghum lines, and the induction of SbMATE expression by Al is associated with increased Al-activated citrate exudation from root tips and induction of sorghum Al resistance. At that time, the assumption was that SbMATE gene and protein expression were expressed throughout the entire root tip.

In the current study, we used microscopic imaging of biochemical markers of Al toxicity to more specifically localize the site of Al toxicity within the root tip. Then we employed LCM techniques and immunolocalization to precisely determine the cellular location of SbMATE gene and protein expression within the root tip. We found that the tissue-specific pattern of Al-induced root damage, ROS formation and root tip Al accumulation was most highly localized to an approximately 1–3 mm wide band that coincides with the root DTZ, indicating that this is the site of maximal Al toxicity. In this region we observed the greatest Al-induced cell swelling and breakage, and increased PI staining, consistent with breaching of the plasma membrane. We also found that this was the region of greatest Al accumulation and Al-induced ROS production. Earlier work on Al toxicity in maize by Sivaguru and colleagues had suggested that this zone within maize roots could be a key site of Al toxicity (Sivaguru and Horst, 1998; Sivaguru et al., 1999). It should be noted that this root region is unique, as it is the junction between the root elongation zone and root meristem, and is the region where cells are transitioning from cell division to cell elongation and are undergoing a preparatory phase for rapid cell elongation (Baluska et al., 2010).

Within the DTZ our 3D optical sectioning and orthogonal views of this region (Figure S3 and Movie S5) clearly indicate that the increased ROS production is mainly localized to the epidermal and outer cortical cells. The production of ROS can be associated not only with cell damage but also plant signaling responses (Apel and Hirt, 2004). Hence the high ROS levels in response to Al might serve as a monitoring signal triggering events that lead to protection of root cells in this region from Al toxicity. We had previously reported on Al-induced ROS production in the maize root apex (Jones et al., 2006), but did not have the resources in that study to determine the spatial profile of ROS production. There are several well-documented signaling events for which Al could be associated with increased ROS production. These include electrical depolarization of the plasma membrane (Sivaguru et al., 2003), changes in calcium influx (Sivaguru et al., 2005), cell wall lipid peroxidation (Cakmak and Horst, 1991), changes in rhizosphere, apoplastic and symplastic pH (Degenhardt et al., 1998), interactions with hormone transport (Hasenstein and Evans, 1988; Kollmeier et al., 2000), proton efflux and influx (Ahn et al., 2002) and secondary effects inside root cells including a disturbed cytoskeleton (Sivaguru et al., 1999, 2002). Furthermore, ROS production has been shown to be associated with similar sorts of processes in response to other stimuli, including activation of anion channels and inhibition of proton pumping in connection with hormones (Trouverie et al., 2008). All of these findings lead us to suggest that Al-induced production of ROS could be involved in signaling in sorghum roots that helps to trigger induction of Al resistance via increased SbMATE expression.

The most interesting finding in this study is the fine regulation of the cell-specific SbMATE gene and protein expression in the root tips of the Al-resistant near-isogenic line. In order to most effectively confer Al resistance via Al-activated citrate exudation, it is logical to assume that the highest density of root citrate transporters should be specifically in the root regions most affected by Al toxicity. This is exactly what was found in this study. The localization of SbMATE gene expression is in the same region of the root, the epidermis and outer cortex of the DTZ (Figure 2), where the greatest Al-induced injury, Al accumulation and Al-induced ROS production occurs (Figure 1). Furthermore, localization of SbMATE protein is also highest in this same region. The release of citrate from the sorghum root to mediate Al3+ chelation and detoxification in the rhizosphere comes with a potential cost to the plant. Organic acids are a valuable carbon resource essential for plant growth and development. This is especially true of citric acid, which is an important intermediate in the Krebs or tricarboxylic acid cycle and plays an essential role in the metabolism of virtually all living things (Berg et al., 2010). Therefore, the use of root citrate efflux itself for resistance against Al stress can carry a significant energy cost for the plant, which may negatively affect plant yields. In a recent study we swapped the promoters of the two primary Arabidopsis Al resistance genes, AtALMT1, which encodes the root malate efflux transporter, and AtMATE, which encodes the root citrate efflux transporter, in transgenic Arabidopsis and determined the effect on Al resistance (Liu et al., 2012). We found that the most effective combination was the AtALMT1 promoter driving expression of AtMATE. This promoter is much more specifically localized to the root tip, while AtMATE1 transports citrate, which is a much stronger Al3+ chelator than is malate. In that study evidence was presented that one component of the better root growth in response to Al in the transgenic AtALMT1P::AtMATE Arabidopsis line may be the reduced carbon cost to the plant with regards to loss of OAs from the root. This idea is supported by the findings presented here that a finely regulated Al-activated root citrate exudation may be important not only for Al resistance but also for maintaining sustainable growth and yields under Al stress. In line with this, yet another recent study by Maron et al. (2013) revealed that copy number variation in the gene MATE1 (there are three copies of the gene MATE1 in Al-tolerant plants compared with one copy in normal plants) is responsible for Al tolerance in maize; it will be very interesting to see whether this copy number variation is cell/tissue/zone specific upon Al stress.

It might be useful to speculate why the cells of the root DTZ are more sensitive to Al than cells of other root tip regions; this then requires the plant to afford this region a greater degree of protection against Al. One can envision that some critical Al-sensitive processes or molecules are more active or abundant in cells of this root apical zone. The cells of the DTZ are unique from all other root cells from the perspective of endocytosis, vesicle recycling, polar auxin transport and as the first region of the root apex which is not covered with mucilage (Baluska et al., 2010). Al has been shown to inhibit shootward basipetal auxin transport driven by the PIN2 auxin efflux transporter and exogenous auxin reduces the Al sensitivity of this root apical region (Kollmeier et al., 2000). Interestingly in this respect, exogenous auxin inhibits endocytosis (Paciorek and Friml, 2006; Robert et al., 2008) and oxygen influx (McLamore et al., 2010), particularly at the Al-sensitive DTZ. Moreover, Al resistance in one Arabidopsis enhancer-tagged mutant line was shown to be based on over-expression of an auxin-like protein, reducing endocytosis and internalization of Al into cells, thereby reducing oxidative damage (Ezaki et al., 2005). On the other hand, exogenous Al has been shown to inhibit endocytosis in roots, and this effect is greatest in cells of the DTZ (Illes et al., 2006; Shen et al., 2008; Amenos et al., 2009). Because of this, the auxin efflux transporter, PIN2, is trapped at the plasma membrane and PIN2-driven shootward auxin transport is inhibited (Shen et al., 2008). Possibly, the high level of endocytosis in this region may be related to the unique Al sensitivity of the DTZ cells, which resembles the high Al sensitivity of tip-growing plant cells and neurons (Baluska et al., 2010). In addition, a recent report confirmed the presence of l-Glu glutamate receptor genes in the plant roots of Arabidopsis, shown to affect the architecture of roots via the MEKK1 kinase pathway (Forde et al., 2013); antagonists of this receptor have been shown to block Al-induced microtubule depolymerization and membrane depolarization through calcium signaling (Sivaguru et al., 2003), and in line with this the l-Glu glutamate receptors are shown to increase production of ROS in neuronal cells in mammals.

We employed the LCM technique to quantify the abundance of SbMATE mRNA in a spatially relevant manner instead of conventional in situ hybridization techniques because the LCM technique better quantifies SbMATE expression in discrete cellular regions of the root tip. Although there are several mRNA in situ hybridization protocols in the literature that have been used extensively in plants to quantify RNA transcripts (Birnbaum et al., 2003), several technical hurdles prevent in situ hybridization from accurately quantifying mRNA abundance. These include non-specific binding of mRNA probes, autofluorescence, light scattering and interference in tissues when using fluorescently tagged probes, differences in quantum efficiencies for different fluorescent probes and whether the probe's ability to fluoresce is affected by the local chemical environment (Kupper et al., 2007). We previously published a quantitative in situ hybridization technique (Kupper et al., 2007) which attempted to resolve these problems by using spectral confocal microscopic techniques and ratiometric analysis of fluorescent oligomeric probes to quantify the expression of the gene of interest relative to an internal control gene. Even with the improvements afforded by this technique, it is still a method that indirectly quantifies mRNA abundance of the gene of interest.

In summary, here we have shown how highly regulated the spatial localization of Al-tolerance genes and proteins is within the external cell layers of the TZ within the growing root tip. The time course for the onset and then recovery from the highly localized biochemical responses of the root to Al stress, including Al-induced cell damage, ROS and callose production and Al accumulation, correlates well with the relatively slow (4–6 days) induction of Al resistance in sorghum. This leads us to hypothesize that some of these localized Al stress responses, particularly ROS production, may play a signaling role in the induction of SbMATE gene expression, which also takes 4–6 days to be fully expressed. As we begin to gain a deeper understanding of the molecular and biochemical regulation of the response of plants to Al toxicity, we are also gaining a deeper appreciation for the sophisticated and elegant regulatory processes that the sorghum plant employs to maximize Al resistance while at the same time minimizing the carbon cost to the plant.

Experimental Procedures

Plant materials and growth conditions

Sorghum seeds were surface-sterilized with 0.5% (w/v) NaOCl for 15 min, rinsed with distilled H2O and germinated for 4 days at 26°C. Six to 10 seedlings were then transplanted to 8-L tubs containing complete Magnavaca nutrient solution at pH 4.0. After 24 h, the solution was changed to either –Al or +Al (27 μm activity) nutrient solution for subsequent treatment (Magalhaes et al., 2007).

Reactive oxygen species, plasma membrane integrity, callose and Al localization

After designated treatments (1 day –Al, 1 day +Al, 3 days +Al), the first 10 mm of the root apex of intact plants was suspended in a fluorescent dye cocktail consisting of 25 μm carboxy-H2DCFDA [5- (and-6)-carboxy-2′,7′-dichlorodihydrofluorescein diacetate] and 100 μg ml−1 PI in 200 mm of CaSO4 and KCl for 30 min, while the remaining parts of the roots were wrapped in a moist filter paper and positioned vertically. After labeling, the images were obtained in a fluorescence stereomicroscope (Setereolumar v12, Zeiss, with a 1.5 × objective under alternating appropriate standard filters for DCFDA [fluorescein isothiocyanate (FITC) filter] and PI (rhodamine filter) at the same position. The same roots or the replicate roots of the same treatments were then taken immediately to a confocal laser scanning microscope (CLSM) for more detailed high-magnification imaging. In any case (stereomicroscopy with or without confocal microscopy), the roots were not kept for more than 60 min at horizontal positions to avoid any gravity-induced changes in the Al-induced ROS pattern and membrane integrity. Aluminum accumulation was tracked using the dye lumogallion in either intact live roots or in paraffin-embedded sections together with visualization of Al-induced callose via staining with aniline blue (0.1%) in 2-amino-2-(hydroxymethyl)1,3-propanediol (TRIS) buffer (pH more than 9.5) as described earlier (Sivaguru et al., 1999). The lumogallion fluorescence was visualized using the same Zeiss fluorescence stereomicroscope with a 1.5 × objective with an FITC filter that was used to visualize DCFDA fluorescence.

Confocal laser scanning microscopy, 3D rendering of ROS and PI

Confocal laser scanning microscope imaging was performed by again employing a 0.8 numerical aperture (NA) 20 × Planapochromat air objective, a 25 × LD-Planapochromat 0.8 NA water immersion or a 40 × C-apochromat 1.2 NA water immersion objective on a Zeiss LSM 710 CLSM (Carl Zeiss). The images were obtained using a 405-nm diode laser for detecting the aniline blue fluorescence associated with Al-induced callose production, a 488-nm Ar laser for DCFDA or lumogallion fluorescence (500–550 nm emission), or a 561 nm DPSS laser for PI imaging (650–700 nm). The ZEN 2011 software or the Imaris suite software (Bitplane, Zurich, were used to create 3D animations or projections showing all optical planes and saved in a five frames per second avi movie format with 5 or 10% compression. All final images were organized and assembled in Photoshop software (Adobe Systems, and the levels were adjusted in all images at the same time (entire panel) to display the best signal to noise ratio of grey values.

Western blots

The SbMATE coding sequence was amplified by PCR and the PCR fragment was cloned into the pGEX4T1 vector (GE Healthcare, to generate a construct for expression of the GST-SbMATE fusion protein. The pGEX4T1-SbMATE plasmid was transformed into Rosetta2 competent cells (Novagen, EMD Millipore, The expression of the GST-SbMATE recombinant protein in the Rosetta2 cells was induced at 22°C by 0.1 mm of isopropyl β-d-1-thiogalactopyranoside. The expressed GST-SbMATE recombinant proteins were purified by the Glutathione Sepharose High-Performance kit (GE Healthcare, The purified GST-SbMATE recombinant proteins were then separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and transferred from the SDS-PAGE gel to a polyvinylidene fluoride (PVDF) membrane, where they were probed by the affinity-purified SbMATE antibodies. The Western blot produced a single band for the SbMATE antibody (Figure 3I).

Immunofluorescence localization of SbMATE protein and imaging of SbMATE protein expression

Root tip samples of the Al-resistant sorghum NIL line were labeled with custom developed affinity-purified SbMATE polyclonal antibody (see Methods S1 for details) and the localization of SbMATE protein expression was tracked with a goat anti-rabbit secondary antibody conjugated with Alexa Fluor 488. The fluorescently stained sections were imaged in an apotome optical sectioning system (Sivaguru et al., 2012) coupled with X-Cite 120 illumination (Carl Zeiss). Further details are provided in Methods S1.

Laser capture and microdissection

After designated treatments, plant root apices (0–10 mm) were harvested in 4% paraformaldehyde (Electron Microscopy Sciences, in PBS and fixed for 1 h with the first 10 min under vacuum. Fixed roots were dehydrated in a graded ethanol series, processed in an automatic tissue processor (Leica ASP300,, infiltrated and embedded in paraffin wax. Sections 6–10 μm thick were made in a motorized microtome (Leica RM2255). Sections were expanded in an ultrapure (18 MΩ) water bath and secured on PEN membrane glass slides (LCM 0522, Microdissect Gmbh, for laser capture. After thorough drying in a hood, they were dewaxed in ultrapure xylene, which was part of a Picopure RNA isolation kit (Invitrogen, The dewaxed slides were loaded into an Arcturus Microdissection system (Molecular Devices, and target cell layers were divided into the whole MZ and TZ or the epidermis/outer cortex or inner cortex/stele. Further details are provided in Methods S1.

Synthesis of cDNA and quantitative real-time RT-PCR for SbMATE gene expression

The extracted total RNA was heated at 64°C for 15 min and then immediately used for first-strand cDNA synthesis. First-strand cDNAs were synthesized in a reaction cocktail containing 12 μl of total RNAs, 4 μl of 5 × reaction buffer, 50 ng of random primers, 1 mm of each dNTP and 1 μl of SuperScript III reverse transcriptase (Invitrogen, in a total volume of 20 μl. The reaction was carried out at 37°C for 90 min, followed by heating at 72°C for 10 min. Quantitative real-time RT-PCR was conducted using an ABI 7500 Real Time PCR System and the SYBR Green Kit (Applied Biosystems, To quantify SbMATE gene expression, each real-time RT-PCR reaction contained 3 μl of the first-strand cDNAs, 0.15 μl of primers and 10 μl of 2 × SYBR Green Master Mix in a final volume of 20 μl. The conditions used for quantifying the 18S internal control were the same as that for SbMATE expression except that 2 μl of the 100 × diluted first-strand 18S cDNA was used in each reaction. Three technical replicates were included for each sample. The optimal gene-specific quantification real-time PCR primer sequences for SbMATE are: 5′-CTG-GTG-AGG-AAG-GTC-GATGTC-3′, 5′-AGG-AAG-CGC-CGG-AAT-TTG-3′; and for 18S 5′-AATCCC-TTA-ACG-AGG-ATC-CAT-TG-3′, 5′-CGC-TAT-TGG-AGC-TGGAAT-TAC-C-3′. The real-time PCR program contained an initial denaturation at 95°C for 10 min, followed by 40 cycles of 94°C for 15 sec, 60°C for 1 min and a final dissociation stage of 95°C for 15 sec, 60°C for 1 min and 95°C for 15 sec. The data were analyzed with the 7500 System SDS Software (Applied Biosystems) and plotted in SigmaPlot version 10 (Systat Software,


We thank Donna Epps for her excellent technical help in histology, Sarah Boomgarden Williams and Nick Vasi, IGB-UIUC communications department for editing the language for accuracy and Frantisek Baluska, University of Bonn, Germany for critical reading and comments on the manuscript. We also thank Drs. Jurandir Magalhaes and Robert Schaffert of Embrapa Maize and Sorghum, Sete Lagoas, Brazil, for generously proving us with seed of the sorghum near isogenic lines, ATF8B and ATF10B.

Conflict of Interest

Authors declare no conflict of interests.

Author Contribution

M.S., J.L. and L.V.K. designed research; M.S. and J.L. performed research; M.S., J.L. and L.V.K. provided reagents and analytic tools; M.S., J.L. and L.V.K. analyzed data; and M.S., J.L. and L.V.K. wrote the paper.