Abstract Plant cell walls exhibit architectural and compositional changes throughout their development and in response to external cues. While tubulins are involved in cell wall biogenesis, much remains unknown about the scope of their involvement during the orchestration of this resource-demanding process. A transgenic approach coupled with cell wall compositional analysis, RNA-seq and mining of publicly available diurnal gene expression data was used to assess the involvement of tubulins in poplar leaf cell wall biogenesis. Leaf cell walls of transgenic poplar lines with constitutive overexpression of α-tubulin (TUA) exhibited an increased abundance of homogalacturonan, along with a reduction in xylose. These changes were traced to altered expression of UDP-glucuronic acid decarboxylase (GADC) in the transgenic leaves. A model is postulated by which altered diurnal control of TUA through its constitutive overexpression led to a metabolic tradeoff affecting cellular utilization of GADC substrate UDP-glucuronic acid. While there were no effects on cellulose, hemicellulose or lignin abundance, subtle effects on hemicellulose composition and associated gene expression were noted. In addition, expression and enzymatic activity of pectin methylesterase (PME) decreased in the transgenic leaves. The change is discussed in a context of increased levels of PME substrate homogalacturonan, slow stomatal kinetics and the fate of PME product methanol. Since stomatal opening and closing depend on fundamentally contrasting microtubule dynamics, the slowing of both processes in the transgenic lines as previously reported appears to be directly related to underlying cell wall compositional changes that were caused by tubulin manipulation. Introduction Plant growth depends largely on carbon dioxide uptake coupled with evaporative water loss at the leaf surface. Specifically, the exchange of these gases is controlled by stomatal pores that open and close in accordance with turgor changes in the flanking guard cells (Dow and Bergmann 2014, Lawson and Blatt 2014). How guard cell geometry and stomatal pore diameter respond to turgor depends on rapidly changing interactions between cortical microtubules (MTs) and cellulose microfibrils (MFs) (Eisinger et al. 2012, Rui and Anderson 2016, Rui et al. 2016). Microtubule stabilization and bundling promote stomatal opening, while MT depolymerization facilitates stomatal closing (Eisinger et al. 2012). Although rapid MT and cellulose dynamics direct stomatal aperture changes, structural properties conferred by guard cell wall composition contribute to overall flexibility. Cellulose constrains longitudinal guard cell expansion under turgor, slowing stomatal opening, while xyloglucans promote stomatal opening (Rui et al. 2016). Pectin composition and methylation also influence guard cell rigidity and stomatal aperture response kinetics (Jones et al. 2003, Amsbury et al. 2016). Pectins are substantially more abundant than cellulose in primary cell walls of Arabidopsis (Zablackis et al. 1995), and this suggests a magnified importance in primary cell wall dynamics. Greater pectin enrichment in guard cells compared with adjacent pavement cells has been reported (Jones et al. 2003, 2005, Merced and Renzaglia 2014), and some pectins are specific to the walls of guard cells (Majewska-Sawka et al. 2002). The movement of carbohydrates toward sites of cell wall incorporation is mediated by the cytoskeleton. During cell division, massive actin and MT-dependent delivery of pectinaceous sugars and uronic acids gives rise to the phragmoplast (Boruc and Van Damme 2015, Voxeur and Hofte 2016). Extensive remodeling occurs as subsequent hemicellulose and cellulose deposition within the pectin-rich phragmoplast gives rise to the cell plate during cytokinesis, and ultimately to the formation of new cell wall (Lee and Liu 2013). The continued dynamic partnership of cortical MTs and cellulose MFs during subsequent cell expansion has long been recognized as a cornerstone of plant form (Lloyd and Chan 2002). Cellulose synthase (CesA) complexes move from the Golgi bodies to the plasma membrane, and are guided by cortical MTs (Paredez et al. 2006, Crowell et al. 2009). Microtubules also appear to be directly involved in pectin trafficking to maturing cell walls (Tian et al. 2004, McFarlane et al. 2008, Kong et al. 2015, Zhu et al. 2015), and to have a role during pectin incorporation in wood-forming tissues as we reported recently (Swamy et al. 2015). Specifically, transgenic poplars that expressed C-terminal truncated α-tubulins (TUA) exhibited altered solvent-extractability of xylem cell wall pectins. Other reports leave open the possibility of MT-independent trafficking of pectins, for example, where delivery of pectin polymers to the growing cell wall matrix continues despite impaired MT function (Domozych et al. 2014). It is also important to note that cell wall biogenesis exhibits diurnal programming, and takes place in diurnally overlapping stages under complex orchestration (Solomon et al. 2010, Mahboubi et al. 2015). Scattered evidence that MT functioning also exhibits diurnal oscillations (Fukuda et al. 2000, Jacobshagen et al. 2001) argue for the possibility of cooperativity between cytoskeletal (MT) and metabolic processes during cell wall biogenesis. The suite of tubulin-modified transgenic poplars we generated previously exhibited no pleiotropic effects on plant development or morphology. However, stomatal closure in response to drought, and stomatal opening in response to light were impaired (Swamy et al. 2015). The present study was undertaken to measure the impact of tubulin manipulation on leaf primary cell wall composition. A model is discussed in which constitutive overexpression of tubulins interfered with diurnal control of pectin and hemicellulose accrual, and thereby altered primary cell wall composition and functionality. Materials and methods Plant growth and drought stress treatments Wild-type, dY (one line) and dEY (two lines) plants were propagated from single-node cuttings and grown to a height of 1.5 m in 1-gallon pots of commercial Fafard 3B potting mix (Fafard, Agawam, MA, USA) supplemented with Osmocote Plus 15-9-12 4-month slow-release fertilizer (Scotts, Marysville, OH, USA). The drought treatments were performed as previously described (Swamy et al. 2015). Plants were watered twice daily until water was withheld for 24–30 h. When petioles of expanding leaves at leaf plastochron index (LPI) 5–8 began to lose turgor, stomatal conductance of fully expanded leaves (LPI 15) was measured using a Licor LI-6400XS (Frost et al. 2012) to confirm the stress state. Adjacent leaves (LPI 14 and 16) were immediately harvested into liquid nitrogen, and stored at −80 °C until use. ABA feeding ABA-feeding was carried out using LPI 12–16. Two days prior to the experiment, leaves were trimmed to equal leaf area using a template while still attached to the plant. For feeding, leaves were excised at the base of the petiole and then cut under degassed water to leave a petiole length of 4 cm. Leaves were inserted into 15-ml graduated Falcon tubes filled with degassed water and jacketed by white PVC tubing to prevent heating. Excised leaves were equilibrated in the tubes for 30 min to ensure that water uptake was similar for all leaves. Water uptake was then measured for 60 min before leaves were transferred to new tubes containing 25 μM ABA. Comparative uptake was estimated by carefully marking the Falcon tubes at 20–30 min time intervals with a fine-tipped felt pen. The rate of water uptake during ABA feeding was calculated and divided by the average rate of water uptake before feeding for each genotype. Feeding with a control solution of 25 μM butyric acid was found not to affect water uptake. Therefore, to minimize the possibility of unknown effects of a long pre-incubation with butyric acid on subsequent ABA feeding, butyric acid was not used during pre-ABA water uptake measurements. All steps of the feeding were conducted under full daylight (midday) in a greenhouse and care was taken to ensure that airflow and light conditions were identical for all leaves. Excised leaves can be maintained in this fashion for at least one full day before water uptake begins to decrease. Gene expression analysis Total RNA was extracted using the Direct-zol kit (Zymo Research, Irvine, CA, USA) with Plant RNA Reagent (Ambion, Thermo Fisher Scientific, Waltham, MA, USA). RNA quality was evaluated by gel electrophoresis and quantified using the Qubit RNA HS Assay Kit with a Qubit fluorometer (Thermo Fisher Scientific). Illumina TruSeq RNA libraries were prepared according to the manufacturer’s instructions and sequenced on a NextSeq 500 at the Georgia Genomics Facility, University of Georgia. Four to nine million paired-end 75-bp reads were generated per sample, with biological replicates (n = 4 for wild-type (WT), n = 11 for dY and n = 8 for dEY). The RNA-seq data has been deposited to the NCBI Sequence Read Archive under accession SRP106027. After quality control filtering, sequences were mapped to the variant-substituted Populus tremula × alba genome v1.1 (Xue et al. 2015) using Tophat2 (Kim et al. 2013). Differential expression between genotypes (WT vs dY or WT vs dEY) or treatments (well-watered (WW) vs drought (DR)) was assessed using DEseq2 (Love et al. 2014) with P ≤ 0.01 and fold-change ≥1.5. Cell wall biogenesis genes were extracted based on the following terms/keywords in Gene Ontology annotation: cell wall biogenesis, cell wall organization, cell wall modification, glucan and nucleotide-sugar metabolic process. Diurnal gene expression data of Filichkin et al. (2011) were downloaded from ArrayExpress (accession E-MEXP-2509). Raw data were processed by MAS5 implemented in the affy package (Gautier et al. 2004). To examine the diurnal pattern of gene expression, the GeneCycle R package (Ahdesmäki et al. 2007) was applied to test the periodicity of 24 h. The sums of expression values from multiple probe-sets of the same gene or of closely related gene family members were used. Cell wall analysis Freeze-dried leaf tissues were Wiley-milled (60-mesh), Soxhlet-extracted with 100% ethanol, and air-dried to generate alcohol-insoluble cell wall residue (AIR) samples. To remove lipids, AIR were sequentially incubated for 30 min at room temperature in chloroform/methanol (1:1 v/v), and 100% acetone. Starch was removed using porcine α-amylase (Sigma-Aldrich, St. Louis, MO, USA, A6255, 35 U per 0.1 g AIRs) in 100 mM Tris buffer (pH 7.0) for 48 h at room temperature, washed with H2O and 100% acetone, and then air-dried. Serial extraction to isolate ionically bound pectins, hemicellulose-esterified pectins, xylan-rich hemicelluloses and xyloglucoan-rich hemicelluloses from AIR samples was carried out as described (Pattathil et al. 2012). Briefly, 50 mg AIR samples were suspended in 50 mM ammonium oxalate (AO, pH 5.0), and incubated overnight at room temperature with constant shaking. The suspension was centrifuged at 4000g for 15 min, the supernatant was collected, and the pellet was washed three times with ddH2O. The washed pellet was then incubated as above using 50 mM sodium carbonate (SC, pH 10), followed by 1 M KOH and 4 M KOH. The 1 and 4 M KOH extracts were carefully neutralized with glacial acetic acid. The supernatants from each step were dialyzed against ddH2O (1:60 v/v, 3500 MWCO tubing) for 48 h with stirring at room temperature, with a complete change every 12 h. Dialysates were then lyophilized, weighed and stored under vacuum over silica gel desiccant. Glycosyl residue composition was determined according to Merkle and Poppe (1994). Briefly, lyophilized samples were weighed and resuspended in ddH2O to 1 mg ml–1. Inositol (20 μg) was added as internal standard to a 500 μg aliquot of resuspended cell wall extracts and freeze-dried. Methyl glycosides were prepared by methanolysis in 600–800 μl of 1 M methanolic HCl (Supelco, Sigma-Aldrich) at 80 °C for 16 h. Samples were dried under nitrogen, resuspended in isopropanol and dried under nitrogen. Residue was per-O-trimethylsilylated with Tri-Sil (Pierce, Thermo Fisher Scientific) at 80 °C for 0.5 h. Gas chromatography–mass spectrometry (GC–MS) analysis was performed on an Agilent Technologies (Santa Clara, CA, USA) 7890A GC interfaced to a 5975C MSD, using an Agilent DB-5MS fused silica capillary column (30 m × 0.25 mm ID). Samples were injected at 80 °C followed by a 2 min hold at 80 °C; a 20° min–1 ramp to 140 °C; 2 min hold; ramp 2° min–1 to 200 °C; ramp 30° min–1 to 250 °C; and a 5 min hold. A mix of derivatized authentic standard monosaccharides was run with each batch of samples for retention time and sensitivity monitoring. AIR cellulose was estimated according to Updegraff (1969). Washed and vacuum-dried cell wall residue from the AO, SC, 1 M KOH and 4 M KOH fractions (~35 mg) was digested in 3 ml acetic–nitric reagent (acetic acid:nitric acid:water = 8:1:1) at 95 °C for 30 min, pelleted, resuspended and washed sequentially with acetic-nitric reagent, water and methanol, vacuum-dried and weighed as cellulose. Lignin was estimated as the differential between starting mass and cellulose. ELISA analysis of the AO fraction was performed as described (Pattathil et al. 2012, Swamy et al. 2015). Significance of genotypic differences (WT versus transgenics) in cell wall mass yield and carbohydrate content was evaluated using two-sample t-tests. Where indicated, data from both transgenic groups were pooled for comparison with WT. PME activity assay An aliquot of liquid nitrogen-ground leaf powder (100 mg) was extracted with 400 μl of 0.1 M citric acid, 0.2 M Na2HPO4, 1 M NaCl (pH 7.5) at 4 °C for 3 h according to Hewezi et al. (2008). Extract was centrifuged for 20 min at 16,000 rpm at 4 °C, and soluble protein quantified using the BCA protein assay kit (EMD Millipore, Temecula, CA, USA). Pectin methylesterase (PME) activity was estimated colorimetrically according to Richard et al. (1994). Briefly, a 10 μl aliquot of the protein extract was added to 3 ml of substrate solution containing 0.75% esterified pectin from citrus fruit (Sigma-Aldrich, P9561), 0.75 M NaCl and 0.03% methyl red (pH 7.5), and incubated for 1 h at 37 °C. The OD change at 525 nm was measured using a SpectraMax M2 microplate reader (Molecular Devices, Sunnyvale, CA, USA). Pectin methylesterase activity was expressed as nmol H+ min–1 mg–1 total protein, based on a calibration curve of 0–100 mM HCl in substrate solution. Results Transgenic Populus over-expressed TUA in leaves The transgenic P. tremula × alba (INRA 717-1B4) lines used in this study were reported previously (Swamy et al. 2015). The lines constitutively expressed C-terminal truncated variants of TUA1, lacking either the terminal Y (dY) or EY (dEY) residues. TUA sequences are highly conserved except for the C-termini (Oakley et al. 2007). The C-terminal residues of TUA are important for post-translation modification via the detyrosination cycle, a mammalian pathway which was determined not to operate in plants (Swamy et al. 2015, Hu et al. 2016). The original study documented a complex regulation of tubulin gene expression, as 35S:TUA1dY/dEY transcripts and proteins were barely detected in xylem but abundant in leaves (Swamy et al. 2015). As a follow-up to the previous study with its focus on wood-forming tissues, the present investigation aimed to examine cell wall changes in the leaves where TUA1 overexpression (OE) was strong. RNA-Seq analysis confirmed 10- and 30-fold increases of TUA1 transcripts in dY and dEY lines, respectively (Figure 1). Because the transgenic dY and dEY phenotypes were morphologically similar (see below and Swamy et al. 2015), the dY and dEY transgenics are collectively referred to as TUA1-OE lines. Expression of TUA5, the most abundant TUA transcript in Populus leaves, was unchanged in the TUA1-OE lines (Figure 1). Figure 1. View largeDownload slide Transcript abundance of TUA1 (Potri.002G111900) and TUA5 (Potri.009G085100) in mature leaves based on RNA-Seq. Well-watered (WW) and drought-stressed (DR) wild-type (WT) and TUA1-OE transgenic (dY and dEY) plants are shown. Transcript abundances of TUA1 represent both endogenous and transgenes. Data represent the means ± SE of n = 4 WT or n = 7–11 transgenic plants. Figure 1. View largeDownload slide Transcript abundance of TUA1 (Potri.002G111900) and TUA5 (Potri.009G085100) in mature leaves based on RNA-Seq. Well-watered (WW) and drought-stressed (DR) wild-type (WT) and TUA1-OE transgenic (dY and dEY) plants are shown. Transcript abundances of TUA1 represent both endogenous and transgenes. Data represent the means ± SE of n = 4 WT or n = 7–11 transgenic plants. Pectin abundance and composition were altered in leaf cell walls of TUA1-OE plants Mature leaves of WT and TUA1-OE plants were subjected to cell wall composition analysis to probe the effects of tubulin manipulation under WW as well as DR conditions (Swamy et al. 2015). The AIR of leaves, as well as the cellulose and lignin content of AIR, were similar between WT and transgenics (Table 1). The mass totals of non-lignocellulosic AIR fractions recovered following sequential extractions in AO, SC, 1 M KOH and 4 M KOH trended slightly higher for the transgenics (Table 1). AO releases primarily pectins, while SC, 1 M KOH and 4 M KOH treatments de-esterify and remove hemicellulose-cross-linked pectins, xylan-enriched hemicellulose and xyloglucan-enriched hemicelluloses, respectively (Selvendran and Oneill 1987). Total hemicelluloses did not differ between genotypes, but AO dry mass yield was highest in dEY plants with the strongest TUA1 overexpression (Table 1 and Figure 1). There was no measurable drought treatment effect on fraction yields in terms of their mass or sugar composition (Table 2 and Table S1 available as Supplementary Data at Tree Physiology Online). Therefore, samples from WW and DR plants were pooled, and subsequent comparisons focused on genotype differences. The most abundant sugar overall was xylose (Xyl) (Table 2), 90% of which comprised 1 and 4 M KOH extractable hemicelluloses (see Figure S1 available as Supplementary Data at Tree Physiology Online). Chelating agents like AO remove readily extractable pectins with minimal degradation, and based on yield-adjusted recoveries, AO was substantially more enriched than the other extract fractions for pectin constituent galacturonic acid (GalA) (see Figure S1A available as Supplementary Data at Tree Physiology Online). Over 60% of cell wall GalA was AO-extractable, with GalA comprising about 44% of AO sugars, compared with just 7% of sugars in the remaining fractions combined (see Figure S1B available as Supplementary Data at Tree Physiology Online). Table 1. Yield and composition of cell wall material from WT and transgenic poplar leaves. AIR Cellulose Lignin Oxalate Carbonate 1 M KOH 4 M KOH Total % DW % AIR DW % AIR DW WT 66.7 ± 2.0 21.6 ± 0.9 10.3 ± 1.2 10.5 ± 1.2 6.9 ± 1.9 21.2 ± 6.1 12.3 ± 3.6 50.9 ± 6.0 dY 65.5 ± 3.2 20.7 ± 1.9 9.6 ± 1.7 10.8 ± 0.9 6.6 ± 1.4 23.5 ± 6.1 11.3 ± 1.5 52.2 ± 7.0 dEY 65.7 ± 2.1 22.3 ± 2.4 10.2 ± 1.2 13.1 ± 2.9 6.4 ± 2.4 22.1 ± 4.4 12.4 ± 2.4 54.0 ± 7.8 AIR Cellulose Lignin Oxalate Carbonate 1 M KOH 4 M KOH Total % DW % AIR DW % AIR DW WT 66.7 ± 2.0 21.6 ± 0.9 10.3 ± 1.2 10.5 ± 1.2 6.9 ± 1.9 21.2 ± 6.1 12.3 ± 3.6 50.9 ± 6.0 dY 65.5 ± 3.2 20.7 ± 1.9 9.6 ± 1.7 10.8 ± 0.9 6.6 ± 1.4 23.5 ± 6.1 11.3 ± 1.5 52.2 ± 7.0 dEY 65.7 ± 2.1 22.3 ± 2.4 10.2 ± 1.2 13.1 ± 2.9 6.4 ± 2.4 22.1 ± 4.4 12.4 ± 2.4 54.0 ± 7.8 Data are means ± SD of n = 5–8. Significance between WT and transgenic is shown in boldface (P ≤ 0.05). Table 2. Cell wall sugar yields of WT and transgenic leaves. Ara Rha GalA Fuc Xyl GlcA Man Gal Glc Xyl/GalA μg mg–1 AIR DW WW WT 79.8 ± 10.5 15.4 ± 1.1 77.9 ± 8.9 5.8 ± 0.6 174.7 ± 25.1 16.8 ± 2.5 6.5 ± 0.8 60.4 ± 8.2 139.0 ± 45.0 2.3 ± 0.5 dY 78.7 ± 4.7 14.3 ± 1.1 87.0 ± 10.1 5.5 ± 0.5 163.0 ± 5.7 16.5 ± 0.4 6.2 ± 0.8 56.4 ± 1.9 148.7 ± 22.1 1.83 ± 0.2 dEY 76.0 ± 10.2 15.5 ± 2.5 94.8 ± 13.7 5.5 ± 1.0 135.3 ± 16.9 14.0 ± 5.1 6.6 ± 1.5 61.8 ± 8.7 166.8 ± 53.9 1.44 ± 0.2 DR WT 78.7 ± 3.2 15.8 ± 0.6 76.1 ± 6.0 5.9 ± 0.4 178.1 ± 16.2 18.2 ± 1.7 6.4 ± 0.5 63.1 ± 4.5 134.0 ± 21.0 2.36 ± 0.4 dY 80.3 ± 7.7 15.1 ± 0.6 83.7 ± 2.2 5.8 ± 0.7 175.3 ± 29.0 16.8 ± 1.4 6.8 ± 0.5 59.4 ± 1.1 133.1 ± 41.3 2.09 ± 0.3 dEY 89.3 ± 11.6 17.4 ± 2.1 90.2 ± 9.6 5.6 ± 1.0 167.4 ± 20.8 14.4 ± 4.3 6.9 ± 1.2 65.4 ± 8.6 119.7 ± 51.2 1.86 ± 0.1 P (WT vs dY) 0.97 0.06 0.04 0.29 0.30 0.32 0.99 0.13 0.74 0.05 P (WT vs dEY) 0.52 0.36 0.01 0.41 0.04 0.08 0.55 0.62 0.77 0.00 Ara Rha GalA Fuc Xyl GlcA Man Gal Glc Xyl/GalA μg mg–1 AIR DW WW WT 79.8 ± 10.5 15.4 ± 1.1 77.9 ± 8.9 5.8 ± 0.6 174.7 ± 25.1 16.8 ± 2.5 6.5 ± 0.8 60.4 ± 8.2 139.0 ± 45.0 2.3 ± 0.5 dY 78.7 ± 4.7 14.3 ± 1.1 87.0 ± 10.1 5.5 ± 0.5 163.0 ± 5.7 16.5 ± 0.4 6.2 ± 0.8 56.4 ± 1.9 148.7 ± 22.1 1.83 ± 0.2 dEY 76.0 ± 10.2 15.5 ± 2.5 94.8 ± 13.7 5.5 ± 1.0 135.3 ± 16.9 14.0 ± 5.1 6.6 ± 1.5 61.8 ± 8.7 166.8 ± 53.9 1.44 ± 0.2 DR WT 78.7 ± 3.2 15.8 ± 0.6 76.1 ± 6.0 5.9 ± 0.4 178.1 ± 16.2 18.2 ± 1.7 6.4 ± 0.5 63.1 ± 4.5 134.0 ± 21.0 2.36 ± 0.4 dY 80.3 ± 7.7 15.1 ± 0.6 83.7 ± 2.2 5.8 ± 0.7 175.3 ± 29.0 16.8 ± 1.4 6.8 ± 0.5 59.4 ± 1.1 133.1 ± 41.3 2.09 ± 0.3 dEY 89.3 ± 11.6 17.4 ± 2.1 90.2 ± 9.6 5.6 ± 1.0 167.4 ± 20.8 14.4 ± 4.3 6.9 ± 1.2 65.4 ± 8.6 119.7 ± 51.2 1.86 ± 0.1 P (WT vs dY) 0.97 0.06 0.04 0.29 0.30 0.32 0.99 0.13 0.74 0.05 P (WT vs dEY) 0.52 0.36 0.01 0.41 0.04 0.08 0.55 0.62 0.77 0.00 Values were normalized to total sugar recovery in each sample, and represent means ± SD (n = 3–4). Statistical significance between WT and transgenics was determined by pooling well-watered (WW) and drought-treated (DR) samples from each genotype (n = 7–8), and indicated by boldface (P ≤ 0.05). No statistical difference was found for drought effect. The GalA content of AO was 19% and 28% greater in dY and dEY leaves, respectively, than WT (Figure 2). Therefore, even though the AO yield of pectin dry mass was significantly larger in just one of the transgenic groups (dEY) than in WT (Table 1), significant GalA enrichment was observed in both transgenic groups. Linear polymers of GalA form homogalacturonans (HGs), generally the most abundant pectin in primary cell walls, whereas rhamnogalacturonans (RGs) are branched, and greatly contribute to the structure and functional diversity of cell wall pectins. The rhamnose (Rha) residues of RGs form linkages with carbohydrate side chains such as arabinoses (Ara) that influence pectin properties and stomatal behavior (Oomen et al. 2002, Jones et al. 2003). It was noteworthy that the ratio of GalA to both Ara and Rha was higher in pectins of TUA1-OE than WT plants (see Table S2 available as Supplementary Data at Tree Physiology Online). This was driven more by the increase of GalA than of Ara or Rha in the transgenic leaves (Figure 2). The AO compositional data were therefore consistent with HG-enrichment relative to RGs in pectins of transgenic leaves. Interestingly, the GalA–Rha–Ara composition of hemicellulose-linked or entrapped pectins in the SC and 1 M KOH fractions did not differ between WT and transgenic lines (see Table S2 available as Supplementary Data at Tree Physiology Online). As was the case for AO, however, the ratio of GalA/Rha and GalA/Ara in 4 M KOH extracts was higher in the TUA1-OE lines. Figure 2. View largeDownload slide Carbohydrate composition of the ammonium oxalate-extractable cell wall fraction. The six most abundant sugars together comprising over 90% of the fraction mass are shown. Data represent the means ± SE of n = 8 plants for each genotype. Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). Figure 2. View largeDownload slide Carbohydrate composition of the ammonium oxalate-extractable cell wall fraction. The six most abundant sugars together comprising over 90% of the fraction mass are shown. Data represent the means ± SE of n = 8 plants for each genotype. Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). GalA increased and Xyl decreased in leaf primary cell walls of TUA1-OE plants While the AO fraction was GalA-enriched in transgenic compared to WT leaf cell walls, it was relatively depleted with respect to galactose (Gal), and more strikingly, to Xyl (Figure 2). Pooled across all fractions, Xyl was 10–15% more abundant in WT than transgenic leaf cell walls, while GalA was 10–20% more abundant in transgenic leaf cell walls (Table 2). Linear regressions revealed negative correlations between GalA and Xyl in the AO, SC and 4 M KOH fractions individually and when they were pooled (Figure 3A), consistent with the idea of a GalA vs Xyl tradeoff. The Xyl–GalA correlation was positive in the 1 M KOH fraction, consistent with a robust functional stoichiometry between GalA and Xyl in this hemicellulose fraction (Figure 3B). A striking feature about the 1 M KOH fraction was the abundance of strong negative correlations between glucose (Glc) and all other sugars, including Xyl (Figure 3B). In addition, Xyl levels trended lower and Glc higher in hemicelluloses of TUA1-OE lines (see Figure S2 available as Supplementary Data at Tree Physiology Online). Glucose and Xyl form the backbone of xyloglucans that predominate in primary cell wall hemicelluloses (Mellerowicz and Gorshkova 2012). Furthermore, the other saccharides are linked to the xyloglucan backbone via its Xyl residues (Hoffman et al. 2005). Therefore, given that the recovery of hemicellulose mass after 1 and 4 M KOH extraction was similar in WT and TUA1-OE, shortfalls in Xyl appear to have been compensated by Glc in a manner that sustained the xyloglucan backbone. Overall, the abundance and correlation data pointed, first, toward a biosynthetic increase of GalA at the expense of Xyl, and second, toward flexibility in substituting Glc for Xyl in certain hemicellulose fractions of TUA1-OE leaves. Figure 3. View largeDownload slide Correlation coefficients between individual sugars recovered in cell wall fractions extracted by ammonium oxalate, sodium carbonate and 4 M KOH (A), and between individual sugars recovered in the 1 M KOH-extractable fraction (B). Boldfaced values were significant at P < 0.01. Figure 3. View largeDownload slide Correlation coefficients between individual sugars recovered in cell wall fractions extracted by ammonium oxalate, sodium carbonate and 4 M KOH (A), and between individual sugars recovered in the 1 M KOH-extractable fraction (B). Boldfaced values were significant at P < 0.01. Was UDP-glucuronic acid utilization perturbed in the TUA1-OE plants? We next investigated whether the observed primary cell wall compositional differences might be linked to transcriptional changes in TUA1-OE leaves. We identified 109 cell wall biogenesis-associated genes that exhibited differential expression (DE) in response to drought and/or transgenic manipulation (see Table S3 available as Supplementary Data at Tree Physiology Online). Strong agreement was observed between the two TUA1-OE groups with respect to their expression differentials versus WT under either WW or DR condition (see Figure S3 available as Supplementary Data at Tree Physiology Online). Many more DE genes were affected by drought (89 genes) than by TUA1 overexpression (50). For instance, two cellulose synthase genes, CesA6C (Potri.005G194200) and CesA6E (Potri.013G019800), were significantly down-regulated by drought in WT, but their expression was similar across all lines at each respective treatment condition. Similarly, expression of a xyloglucan endotransglucosylase (Potri.007G008500) was significantly down-regulated by drought in WT, but the response was attenuated in the TUA1-OE plants. Several cellulose synthase-like (SCL) genes involved in hemicellulose biosynthesis were affected in apparently compensatory fashion. For example, Potri.002G114200 belonging to the CSLC subfamily was down-regulated in both TUA1-OE groups relative to WT under DR, whereas genes of the CSLA (Potri.009G149700) and CSLG (Potri.003G142300) subfamilies were up-regulated. Two Walls-Are-Thin (WAT) genes encoding tonoplast-localized auxin transporters in fiber cells (Ranocha et al. 2010, 2013) were more highly expressed in the transgenic lines under both WW and DR conditions (see Table S3 available as Supplementary Data at Tree Physiology Online). With reciprocity in the mean abundances of GalA and Xyl (Table 2), and with negative GalA–Xyl correlations between individual samples (Figure 3A), we reasoned that the primary metabolic tradeoff was between GalA and Xyl, and we focused on identifying a basis for that tradeoff. UDP-Xyl and UDP-GalA originate from UDP-glucuronic acid (UDP-GlcA) via UDP-GlcA decarboxylase (GADC) and UDP-GlcA 4-epimerase (GAE), respectively (Harper and Bar-Peled 2002, Gu et al. 2009). We therefore mined the RNA-seq data for evidence of a transgenic effect on these enzymatic steps. Transcript levels of a GADC ortholog Potri.014G129200 were lower in TUA1-OE than WT under WW, and significantly so under DR (see Table S3 available as Supplementary Data at Tree Physiology Online). Transcript levels for an UDP-apiose/UDP-xylose synthase (AXS, Potri.004G189900), which like GADC mediates the conversion of UDP-GlcA to UDP-Xyl (Ahn et al. 2006), were also significantly down-regulated in TUA1-OE lines under DR (see Table S3 available as Supplementary Data at Tree Physiology Online). Two GAE orthologs exhibited opposing patterns of differential expression; one (Potri.018G100400) was down-regulated in TUA1-OE lines under DR, though significantly only in dY, while the more abundant Potri.001G320000 trended higher in the transgenic lines at DR. Overall, the expression data were consistent with a possible transcriptional basis for the preferential accumulation of GalA over Xyl in TUA1-OE leaves. Cell wall biogenesis is known to be under diurnal regulation (Mahboubi et al. 2015). To address whether diurnal control played a role in the observed transgenic responses, we explored a published Populus trichocarpa leaf microarray dataset with diurnal time courses (Filichkin et al. 2011). We determined that only one of the leaf-expressed GAEs exhibited a slight transcript level oscillation under light–dark conditions (Figure 4), but reduced expression with no oscillation when the light–dark cycle followed a 48-h cycle of continuous light (see Figure S4 available as Supplementary Data at Tree Physiology Online). In contrast, all three leaf-expressed GADCs exhibited multifold oscillations under both light regimes (Figure 4 and Figure S4 available as Supplementary Data at Tree Physiology Online). Broadly speaking, expression ratios of GAE:GADC decreased during the day and increased at night (Figure 4D). As was the case for GADC, transcript levels of the most strongly expressed TUA and TUB (β-tubulin) genes in leaves exhibited diurnal oscillation under both light regimes (Figure 4 and Figure S4 available as Supplementary Data at Tree Physiology Online). The oscillating transcript levels of tubulin, GADC and GAE genes in P. trichocarpa are therefore consistent with the idea that metabolic fluxes through GAE and GADC might be differentially impacted by tubulin manipulation Figure 4. View large Download slide Diurnal leaf expression of Populus trichocarpa tubulin (TUA1, TUA5 and TUB15), glucuronic acid-4-epimerase (GAE) and glucuronic acid decarboxylase (GADC) genes. The most abundantly expressed members of each gene family are represented: (A) TUA1: Potri.002G111900; TUA5: Potri.009G085100 and TUB15: Potri.001G272800; (B) GADC: Potri.010G207200, Potri.001G237200 and Potri.014G129200; (C) GAE: Potri.001G320000, Potri.003G114600 and Potri.018G100400. Diurnal expression data were from Filichkin et al. (2011), using plants that had been acclimated to 12 h light/12 h dark cycles. (D) Diurnal changes in GAE/GADC expression ratio were plotted as the average of the ratios collected for light–dark acclimated plants and 48 h continuous light pretreated plants (diurnal patterns are shown in Figure S4 available as Supplementary Data at Tree Physiology Online). Histogram values represent the ratio and range of n = 2. Figure 4. View large Download slide Diurnal leaf expression of Populus trichocarpa tubulin (TUA1, TUA5 and TUB15), glucuronic acid-4-epimerase (GAE) and glucuronic acid decarboxylase (GADC) genes. The most abundantly expressed members of each gene family are represented: (A) TUA1: Potri.002G111900; TUA5: Potri.009G085100 and TUB15: Potri.001G272800; (B) GADC: Potri.010G207200, Potri.001G237200 and Potri.014G129200; (C) GAE: Potri.001G320000, Potri.003G114600 and Potri.018G100400. Diurnal expression data were from Filichkin et al. (2011), using plants that had been acclimated to 12 h light/12 h dark cycles. (D) Diurnal changes in GAE/GADC expression ratio were plotted as the average of the ratios collected for light–dark acclimated plants and 48 h continuous light pretreated plants (diurnal patterns are shown in Figure S4 available as Supplementary Data at Tree Physiology Online). Histogram values represent the ratio and range of n = 2. Pectin methylation was altered in the TUA1-OE lines One of the signature developmental events that takes place in cell wall pectin is progressive demethylation of HG-backbone GalA residues by PME (Zhang and Staehelin 1992, Goubet et al. 1998). PME activity regulates the calcium-binding potential of HGs with consequences for pectin properties and cell wall functions (Micheli 2001, Senechal et al. 2014). Our RNA-Seq analysis revealed a striking constitutive decrease in transcript levels of a leaf-abundant PME (Potri.014G149700) in both transgenic groups (Figure 5A). As was the case for TUA and GADC, PME gene expression oscillates diurnally in poplar leaves (Figure 5C). Measurement of PME enzyme activity in leaf extracts revealed lower activity in transgenic than WT lines (Figure 5B). The calcium-binding capacity of HG is determined by the number of carboxyl groups made available by methyl de-esterification (Micheli 2001). Calcium content was not measured directly, but the degree of pectin methylation was inferred based on ELISA analysis of the pectin-rich AO cell wall fraction. ELISA signals were similar between genotypes with a monoclonal antibody (mAb) directed against unmethylated HG (CCRC M38), but were significantly higher in the TUA1-OE lines with mAbs recognizing partially (JIM5) or highly (JIM7) methylesterified HG (Figure 6). This suggested that cell wall pectin in the TUA1-OE lines was more heavily methylesterified and thus less able to interact with calcium. Figure 5. View largeDownload slide Pectin methylesterase gene expression (A), enzyme activity (B) and diurnal expression pattern (C). Mature leaves of WT and transgenic (dY and dEY) plants from well-watered (WW) or drought (DR) treatments were used for (A) and (B). Values represent means ± SE of n = 4 WT or n = 7–11 transgenic samples in (A), or means ± SD of n = 3–4 in (B). Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). Asterisks above the bars denote treatment effect, and asterisks inside the bars indicate transgenic effect. Diurnal expression data in (C) was the same as described in Figure 4. The most abundantly expressed pectin methylesterase gene (Potri.014G149700) is shown in (A) and (C). Figure 5. View largeDownload slide Pectin methylesterase gene expression (A), enzyme activity (B) and diurnal expression pattern (C). Mature leaves of WT and transgenic (dY and dEY) plants from well-watered (WW) or drought (DR) treatments were used for (A) and (B). Values represent means ± SE of n = 4 WT or n = 7–11 transgenic samples in (A), or means ± SD of n = 3–4 in (B). Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). Asterisks above the bars denote treatment effect, and asterisks inside the bars indicate transgenic effect. Diurnal expression data in (C) was the same as described in Figure 4. The most abundantly expressed pectin methylesterase gene (Potri.014G149700) is shown in (A) and (C). Figure 6. View largeDownload slide Methylesterification of ammonium oxalate-extractable homogalacturonan as determined by ELISA. (A) CCRC-M38 mAb recognizes unmethylated HG backbone; (B) JIM5 mAb recognizes partially or weakly methylated HG backbone, and JIM7 recognizes methylated HG backbone (data were normalized with respect to the CCRC-M38 signal). Histogram values represent means ± SD of n = 3 biological replicates. Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). Figure 6. View largeDownload slide Methylesterification of ammonium oxalate-extractable homogalacturonan as determined by ELISA. (A) CCRC-M38 mAb recognizes unmethylated HG backbone; (B) JIM5 mAb recognizes partially or weakly methylated HG backbone, and JIM7 recognizes methylated HG backbone (data were normalized with respect to the CCRC-M38 signal). Histogram values represent means ± SD of n = 3 biological replicates. Statistical significance was determined by two-sample t-test (*P < 0.05; **P < 0.01). ABA-induced stomatal closure was delayed in the TUA1-OE lines We reported previously that stomatal opening at daybreak, and closing during a drought treatment were slower in TUA1-OE than WT plants (Swamy et al. 2015). Delayed stomatal closure in DR-stressed transgenic leaves was based on data collected at one time point during the drought treatment. To obtain higher resolution data of the stomatal behavior in TUA1-OE plants, an ABA petiole feeding experiment was carried out (see Figure S5 available as Supplementary Data at Tree Physiology Online). Water uptake was essentially identical for excised leaves between WT and transgenic plants prior to ABA feeding (see Figure S5A available as Supplementary Data at Tree Physiology Online). There was a clear reduction in all three genotypes within the first 40 min of feeding (see Figure S5B available as Supplementary Data at Tree Physiology Online). Uptake was reduced more substantially in WT than TUA1-OE leaves, leveling off at less than 10% of pre-treatment uptake in WT compared with 20–30% in the TUA1-OE lines by the time ABA feeding was terminated at 90 min (see Figure S5B available as Supplementary Data at Tree Physiology Online). Discussion TUA1 overexpression stimulated HG accrual The data address outstanding questions with regard to the basis of altered cell wall properties in poplar leaves that overexpress TUA1, and support a new insight about the role of MTs in cell wall biogenesis and remodeling. Our previous work with these transgenic lines focused on secondary xylem and concluded that integration of pectins with other cell wall polymers was altered (Swamy et al. 2015). The present work focused on leaf cell walls, and our findings argued that there were quantitative sugar composition differences in the cell walls due to TUA1 overexpression. The two findings are not mutually exclusive since leaf cell walls do not possess the thick secondary layer characteristic of wood-forming xylem cell walls, and since transgene expression was much greater in leaves than in xylem (Swamy et al. 2015). AO-extractable pectins comprise the middle lamella and are ionically or otherwise loosely associated with the cell wall, while most pectin in the hemicellulose-rich fractions is ester-linked (Brummell 2006). GalA in the 4 M KOH extract may represent cellulose cross-linked pectins (Ryden and Selvendran 1990). While HG increased, there was no indication that RG-pectin levels increased. Higher GalA/Rha ratios suggest less pectin branching in TUA1-OE lines, but reduced branching did not appear to affect pectin extraction from the leaf cell wall matrix. This assessment is supported by the fact that the ratio of GalA/Xyl in SC and 1 M KOH extracts did not differ between WT and TUA1-OE leaves. On the basis of extractability, therefore, integration of pectin with hemicelluloses did not change. The increased GalA content of AO-pectin in transgenics thus reflected increased GalA biosynthesis, an increase which in the case of dEY plants resulted in an AO-pectin mass increase. There was little indication on the basis of either gene expression or cell wall compositional data that cellulose accrual was affected in the TUA1-OE lines. Likewise, total hemicellulose content was not altered. However, the ratio of Glc:Xyl appeared to trend higher in the hemicellulose fractions of transgenic leaf cell walls compared with WT. Interestingly, one gene of the CSLC subfamily known to be involved in xyloglucan biosynthesis (Liepman and Cavalier 2012) exhibited lower expression in TUA1-OE than WT leaves under DR (see Table S3 available as Supplementary Data at Tree Physiology Online). An opposing expression pattern was observed for orthologs of the CSLA and CSLG subfamilies, the former known to be involved in glucomannan biosynthesis (Goubet et al. 2009, Liepman and Cavalier 2012). The contrasting expression patterns are consistent with the idea that compensatory metabolism related to hemicellulose composition may occur and that the process was altered in the transgenic lines. Do tubulins mediate the partitioning of GlcA? Current understanding of MT-regulated trafficking of sugars and uronic acids to cell walls does not offer a scenario on how increasing tubulin expression would lead to the GalA increase we observed. It has long been recognized that MTs guide MF deposition, and that MTs and actins collaborate in the vesicular trafficking of a variety of complex sugars from the Golgi network to cell wall sites (Baluska et al. 2002, Zhu et al. 2015, Voxeur and Hofte 2016). However, our cell wall composition data are consistent with a putative tradeoff that pitted pectin vs hemicellulose accrual in primary cell walls. HGs increased, and although hemicellulose levels did not decrease in transgenic plants, their Xyl content decreased. We speculate from the mature state of the leaves in which GAE and GADC transcript levels were affected that the compositional trade-off was not developmentally limited, for example, to cell wall biogenesis in rapidly expanding leaves. We therefore considered whether ongoing utilization of the GAE and/or GDAC products was perturbed in the transgenic leaves. Tubulins have long been regarded as so-called housekeeping genes for the normalization of transcript levels of target genes in various studies. As a result, little has been reported on whether tubulin transcript levels exhibit diurnal oscillation. Our results support the idea of diurnal oscillations of tubulin expression in whole poplar leaves (Figure 4). Diurnal oscillation of tubulin transcript levels in phase with MT stability and function has been reported in guard cells of bean (Fukuda et al. 1998, 2000). TUA protein levels increased by at least 150% due to the constitutive TUA1 overexpression in transgenic poplar leaves (Swamy et al. 2015). This points toward the possibility that a damping of diurnal regulation by TUA1 overexpression mitigated MT capacity to orchestrate cell wall incorporation of UDP-GalA and UDP-Xyl. The data led us to posit that constitutive elevation of tubulin protein levels promoted utilization of GAE product UDP-GalA in a way that was costly to utilization of GADC product UDP-Xyl. GAE expression is generally stronger and more constant than that of GADC, which oscillates several-fold across the diel (Figure 4). Integrated across the diel, constitutively elevated tubulin protein levels would therefore potentially facilitate a greater MT-mediated flux of GAE than of GADC product toward the cell wall (Figure 7). Based on the diurnal expression data (Figure 4 and Figure S4 available as Supplementary Data at Tree Physiology Online), endogenous TUA transcript levels oscillate roughly synchronously with those of GADC. Therefore, in normally functioning WT leaves, GADC expression decreases at a time when MT-mediated trafficking of all substrates for cell wall biogenesis may also be on a downward trajectory. Although GAE expression appears to remain sustained, MT-trafficking of GAE enzyme product GalA would also be limited (Figure 7). This coordination is presumably compromised in transgenic leaves because TUA1 transgene expression and MT function remain artificially high throughout the diel. The temporal organization of cell wall biogenic processes has been studied in tissues undergoing extensive secondary cell wall growth (Hosoo et al. 2002, Gou et al. 2007, Solomon et al. 2010, Mahboubi et al. 2015). In developing xylem of Eucalyptus, for example, expression of pectinesterase genes is higher during the day than at night (Solomon et al. 2010). UDP-Xyl and hemicellulose biosynthesis appear to peak at night, along with the expression of pectin acetylesterases (Solomon et al. 2010), presumably to stabilize pectin when it is not undergoing remodeling (Gou et al. 2012). Figure 7. View largeDownload slide Proposed model of TUA1 overexpression effects on cell wall composition. (A) UDP-GlcA is utilized by GAE and GADC for biosynthesis of pectins and hemicelluloses, respectively. (B) Tubulin and GADC gene transcript levels normally exhibit multifold oscillations across the diel. (C) When TUA1 is constitutively expressed at a high level, periodic reductions in GADC expression against a background of sustained GAE expression result in cell walls with elevated UDP-GalA:UDP-Xyl ratios, consistent with a metabolic tradeoff. Figure 7. View largeDownload slide Proposed model of TUA1 overexpression effects on cell wall composition. (A) UDP-GlcA is utilized by GAE and GADC for biosynthesis of pectins and hemicelluloses, respectively. (B) Tubulin and GADC gene transcript levels normally exhibit multifold oscillations across the diel. (C) When TUA1 is constitutively expressed at a high level, periodic reductions in GADC expression against a background of sustained GAE expression result in cell walls with elevated UDP-GalA:UDP-Xyl ratios, consistent with a metabolic tradeoff. Exactly how increased expression of TUA1 contributed to the observed decreases in GADC transcript level was not determined. However, metabolite repression of plant gene expression has long been known to occur (Graham et al. 1994). Although gene repression by metabolite feedback has not been studied for GAE or GADC, in vitro inhibition of both enzymes by GADC product UDP-Xyl has been reported (Hewezi et al. 2008, Gu et al. 2009). In addition, transcriptional control of nucleotide-sugar biosynthesis related to cell wall biogenesis has been described (Seifert 2004). The potential for UDP-Xyl feedback inhibition of GAE would presumably oscillate in accordance with GADC expression and activity. Since the expression of GADC exhibited a greater tendency to be reduced in the TUA1-OE lines of this study, we conclude that feedback due to reduced utilization of UDP-Xyl had a less significant impact on GAE than GADC activity. This is consistent with our depiction of diel expression and trafficking trends (Figure 7). The leaf cell wall composition and gene expression data offer a basis to argue for a contribution by tubulins to previously described developmental and diurnal influence over cell wall biogenesis (Hosoo et al. 2002, Gou et al. 2007, Solomon et al. 2010, Mahboubi et al. 2015). In light of the downstream consequences reported here, further study of the ways in which MT function interfaces with utilization and metabolic control of cell wall carbohydrates is warranted. Implications for guard cell function Stomatal opening and closing were slower in TUA1-OE than WT plants (Swamy et al. 2015). In addition, data from an ABA feeding experiment were consistent with a capacity for normal initiation of stomatal closure, followed by a much slower response for the duration of the feeding period in the TUA1-OE plants (see Figure S5 available as Supplementary Data at Tree Physiology Online). Cellulose and hemicelluloses constrain and facilitate, respectively, stomatal opening (Rui et al. 2016), while pectin composition and methylation also contribute (Jones et al. 2003, Amsbury et al. 2016). Pectin modification by PME modulates calcium cross-linking and porosity in the cell wall, as well as the ionic makeup of cell wall spaces (Bosch and Hepler 2005). In turn, those changes affect pectin fluidity and the accessibility as well as the activity of cell wall loosening enzymes (Micheli 2001, Samaj et al. 2004, Senechal et al. 2014). Cellulose and hemicellulose contents were not altered in primary cell walls of the transgenic plants. We therefore argue that the observed changes in pectin abundance, composition and methylation strongly contributed to the altered dynamics of transgenic guard cells. Reductions in the proportion of RG-linked Ara side chains and in the degree of de-esterification as observed in transgenic leaves would be expected to decrease porosity, rigidity and flexibility of primary cell wall pectins (Jones et al. 2003, Amsbury et al. 2016). Loss of both rigidity and flexibility may seem contradictory, but the assessment is based on two considerations. First, highly methylated HGs do not coordinate with calcium to form ‘eggbox’ like structures important for rigidity (Micheli 2001). Second, Ara side chains in RGs facilitate molecular shifts within pectin structures that allow for flexibility during stomatal opening (Jones et al. 2003, Moore et al. 2008). Removal of Ara side chains is thought to result in a loss of spacing between HG backbones, and thus a loss of cell wall flexibility (Jones et al. 2003). Because the frequency of HG domains with rigid eggbox structures or with Ara side chains is predicted to be lower in TUA1-OE lines, their cell walls may lose both rigidity and flexibility. We also found evidence that the decrease in Xyl was compensated to some degree in the hemicellulose fraction by Glc (see Figure S2 available as Supplementary Data at Tree Physiology Online). The ability of hemicelluloses to facilitate stomatal movements may partly depend on properties such as hemicellulose branching. A Xyl decrease in xyloglucan would reduce branching frequency with as yet uncharacterized impacts on hemicellulose function. A final question regards the negative effect of TUA1 overexpression on PME transcription and enzyme activity. We found that PME transcript levels tend to increase toward the end of the light cycle and then to decrease sharply at daybreak or slightly before (Figure 5). It has been reported that leaf methanol levels are at their lowest after stomata have been open for some time, and highest at daybreak after stomata have been closed for some time (Huve et al. 2007). According to these authors, methanol production is largely determined by PME activity and methanol emission is determined by stomatal aperture (Huve et al. 2007). Therefore, in light of the anti-parallel diurnal relationship between PME expression and PME product methanol, our results are consistent with the idea of a negative feedback of methanol on PME expression that may have become enhanced due to higher PME substrate levels and slower stomatal opening at daybreak in the transgenic plants. Overall, the data support a model in which diurnal control of MT function may condition cell wall biogenesis. Common abiotic stresses perturb circadian regulation (Sanchez et al. 2011), MT dynamics (Olinevich and Khokhlova 2003, Ban et al. 2013, Nick 2013) and also cell wall composition (Solecka et al. 2008, de Lima et al. 2014, Fernandes et al. 2016). Therefore, our findings add to the foundation for future investigation of MT participation in the regulation of cell wall composition, especially in stressful environments. Supplementary Data Supplementary Data for this article are available at Tree Physiology Online. Acknowledgments We thank Vanessa Michelizzi and Zach Webster for assistance with RNA extraction and tissue processing, and Steve Pettis and Suzzanne Tate for greenhouse plant care. Conflict of interest None declared. Funding This work was supported by the Office of Biological and Environmental Research within the Department of Energy (Grant no. DE-SC0008470). References Ahdesmäki M, Lähdesmäki H, Gracey A, Shmulevich L, Yli-Harja O ( 2007) Robust regression for periodicity detection in non-uniformly sampled time-course gene expression data. BMC Bioinformatics 8: 233. Google Scholar CrossRef Search ADS PubMed Ahn JW, Verma R, Kim M, Lee JY, Kim YK, Bang JW, Reiter WD, Pai HS ( 2006) Depletion of UDP-D-apiose/UDP-D-xylose synthases results in rhamnogalacturonan-II deficiency, cell wall thickening, and cell death in higher plants. J Biol Chem 281: 13708– 13716. Google Scholar CrossRef Search ADS PubMed Amsbury S, Hunt L, Elhaddad N et al. ( 2016) Stomatal function requires pectin de-methyl-esterification of the guard cell wall. Curr Biol 26: 2899– 2906. Google Scholar CrossRef Search ADS PubMed Baluska F, Hlavacka A, Samaj J, Palme K, Robinson DG, Matoh T, McCurdy DW, Menzel D, Volkmann D ( 2002) F-actin-dependent endocytosis of cell wall pectins in meristematic root cells. Insights from brefeldin A-induced compartments. Plant Physiol 130: 422– 431. Google Scholar CrossRef Search ADS PubMed Ban Y, Kobayashi Y, Hara T, Hamada T, Hashimoto T, Takeda S, Hattori T ( 2013) Alpha-tubulin is rapidly phosphorylated in response to hyperosmotic stress in rice and Arabidopsis. Plant Cell Physiol 54: 848– 858. Google Scholar CrossRef Search ADS PubMed Boruc J, Van Damme D ( 2015) Endomembrane trafficking overarching cell plate formation. Curr Opin Plant Biol 28: 92– 98. Google Scholar CrossRef Search ADS PubMed Bosch M, Hepler PK ( 2005) Pectin methylesterases and pectin dynamics in pollen tubes. Plant Cell 17: 3219– 3226. Google Scholar CrossRef Search ADS PubMed Brummell DA ( 2006) Cell wall disassembly in ripening fruit. Funct Plant Biol 33: 103– 119. Google Scholar CrossRef Search ADS Crowell EF, Bischoff V, Desprez T, Rolland A, Stierhof YD, Schumacher K, Gonneau M, Hofte H, Vernhettes S ( 2009) Pausing of Golgi bodies on microtubules regulates secretion of cellulose synthase complexes in Arabidopsis. Plant Cell 21: 1141– 1154. Google Scholar CrossRef Search ADS PubMed de Lima RB, dos Santos TB, Vieira LGE, Ferrarese MDL, Ferrarese O, Donatti L, Boeger MRT, Petkowicz CLD ( 2014) Salt stress alters the cell wall polysaccharides and anatomy of coffee (Coffea arabica L.) leaf cells. Carbohydr Polym 112: 686– 694. Google Scholar CrossRef Search ADS PubMed Domozych DS, Sorensen I, Sacks C, Brechka H, Andreas A, Fangel JU, Rose JKC, Willats WGT, Popper ZA ( 2014) Disruption of the microtubule network alters cellulose deposition and causes major changes in pectin distribution in the cell wall of the green alga, Penium margaritaceum. J Exp Bot 65: 465– 479. Google Scholar CrossRef Search ADS PubMed Dow GJ, Bergmann DC ( 2014) Patterning and processes: how stomatal development defines physiological potential. Curr Opin Plant Biol 21: 67– 74. Google Scholar CrossRef Search ADS PubMed Eisinger WR, Kirik V, Lewis C, Ehrhardt DW, Briggs WR ( 2012) Quantitative changes in microtubule distribution correlate with guard cell function in Arabidopsis. Mol Plant 5: 716– 725. Google Scholar CrossRef Search ADS PubMed Fernandes JC, Goulao LF, Amancio S ( 2016) Regulation of cell wall remodeling in grapevine (Vitis vinifera L.) callus under individual mineral stress deficiency. J Plant Physiol 190: 95– 105. Google Scholar CrossRef Search ADS PubMed Filichkin SA, Breton G, Priest HD et al. ( 2011) Global profiling of rice and poplar transcriptomes highlights key conserved circadian-controlled pathways and cis-regulatory modules. PLoS One 6: 16. Google Scholar CrossRef Search ADS Frost CJ, Nyamdari B, Tsai CJ, Harding SA ( 2012) The tonoplast-localized sucrose transporter in Populus (PtaSUT4) regulates whole-plant water relations, responses to water stress, and photosynthesis. PLoS One 7: 10. Fukuda M, Hasezawa S, Asai N, Nakajima N, Kondo N ( 1998) Dynamic organization of microtubules in guard cells of Vicia faba L. with diurnal cycle. Plant Cell Physiol 39: 80– 86. Google Scholar CrossRef Search ADS PubMed Fukuda M, Hasezawa S, Nakajima N, Kondo N ( 2000) Changes in tubulin protein expression in guard cells of Vicia faba L. accompanied with dynamic organization of microtubules during the diurnal cycle. Plant Cell Physiol 41: 600– 607. Google Scholar CrossRef Search ADS PubMed Gautier L, Cope L, Bolstad BM, Irizarry RA ( 2004) A ffy-analysis of Affymetrix GeneChip data at the probe level. Bioinformatics 20: 307– 315. Google Scholar CrossRef Search ADS PubMed Gou JY, Wang LJ, Chen SP, Hu WL, Chen XY ( 2007) Gene expression and metabolite profiles of cotton fiber during cell elongation and secondary cell wall synthesis. Cell Res 17: 422– 434. Google Scholar CrossRef Search ADS PubMed Gou JY, Miller LM, Hou GC, Yu XH, Chen XY, Liu CJ ( 2012) Acetylesterase-mediated deacetylation of pectin impairs cell elongation, pollen germination, and plant reproduction. Plant Cell 24: 50– 65. Google Scholar CrossRef Search ADS PubMed Goubet F, Council LN, Mohnen D ( 1998) Identification and partial characterization of the pectin methyltransferase ‘homogalacturonan-methyltransferase’ from membranes of tobacco cell suspensions. Plant Physiol 116: 337– 347. Google Scholar CrossRef Search ADS Goubet F, Barton CJ, Mortimer JC et al. ( 2009) Cell wall glucomannan in Arabidopsis is synthesised by CSLA glycosyltransferases, and influences the progression of embryogenesis. Plant J 60: 527– 538. Google Scholar CrossRef Search ADS PubMed Graham IA, Denby KJ, Leaver CJ ( 1994) Carbon catabolite repression regulates glyoxylate CYCLE gene-expression in cucumber. Plant Cell 6: 761– 772. Google Scholar CrossRef Search ADS PubMed Gu XG, Wages CJ, Davis KE, Guyett PJ, Bar-Peled M ( 2009) Enzymatic characterization and comparison of various Poaceae UDP-GlcA 4-epimerase isoforms. J Biochem 146: 527– 534. Google Scholar CrossRef Search ADS PubMed Harper AD, Bar-Peled M ( 2002) Biosynthesis of UDP-xylose. Cloning and characterization of a novel Arabidopsis gene family, UXS, encoding soluble and putative membrane-bound UDP-glucuronic acid decarboxylase isoforms. Plant Physiol 130: 2188– 2198. Google Scholar CrossRef Search ADS PubMed Hewezi T, Howe P, Maier TR, Hussey RS, Mitchum MG, Davis EL, Baum TJ ( 2008) Cellulose binding protein from the parasitic nematode Heterodera schachtii Interacts with Arabidopsis pectin methylesterase: cooperative cell wall modification during parasitism. Plant Cell 20: 3080– 3093. Google Scholar CrossRef Search ADS PubMed Hoffman M, Jia ZH, Pena MJ, Cash M, Harper A, Blackburn AR, Darvill A, York WS ( 2005) Structural analysis of xyloglucans in the primary cell walls of plants in the subclass Asteridae. Carbohydr Res 340: 1826– 1840. Google Scholar CrossRef Search ADS PubMed Hosoo Y, Yoshida M, Imai T, Okuyama T ( 2002) Diurnal difference in the amount of immunogold-labeled glucomannans detected with field emission scanning electron microscopy at the innermost surface of developing secondary walls of differentiating conifer tracheids. Planta 215: 1006– 1012. Google Scholar CrossRef Search ADS PubMed Hu H, Gu X, Xue LJ, Swamy PS, Harding SA, Tsai CJ ( 2016) Tubulin C-terminal post-translational modifications do not occur in wood forming tissue of Populus. Front Plant Sci 7: 10. Google Scholar PubMed Huve K, Christ MM, Kleist E, Uerlings R, Niinemets U, Walter A, Wildt J ( 2007) Simultaneous growth and emission measurements demonstrate an interactive control of methanol release by leaf expansion and stomata. J Exp Bot 58: 1783– 1793. Google Scholar CrossRef Search ADS PubMed Jacobshagen S, Whetstine JR, Boling JM ( 2001) Many but not all genes in Chlamydomonas reinhardtii are regulated by the circadian clock. Plant Biol 3: 592– 597. Google Scholar CrossRef Search ADS Jones L, Milne JL, Ashford D, McQueen-Mason SJ ( 2003) Cell wall arabinan is essential for guard cell function. Proc Natl Acad Sci USA 100: 11783– 11788. Google Scholar CrossRef Search ADS PubMed Jones L, Milne JL, Ashford D, McCann MC, McQueen-Mason SJ ( 2005) A conserved functional role of pectic polymers in stomatal guard cells from a range of plant species. Planta 221: 255– 264. Google Scholar CrossRef Search ADS PubMed Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL ( 2013) TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol 14: R36. Google Scholar CrossRef Search ADS PubMed Kong ZS, Ioki M, Braybrook S et al. ( 2015) Kinesin-4 functions in vesicular transport on cortical microtubules and regulates cell wall mechanics during cell elongation in plants. Mol Plant 8: 1011– 1023. Google Scholar CrossRef Search ADS PubMed Lawson T, Blatt MR ( 2014) Stomatal size, speed, and responsiveness impact on photosynthesis and water use efficiency. Plant Physiol 164: 1556– 1570. Google Scholar CrossRef Search ADS PubMed Lee YRJ, Liu B ( 2013) The rise and fall of the phragmoplast microtubule array. Curr Opin Plant Biol 16: 757– 763. Google Scholar CrossRef Search ADS PubMed Liepman AH, Cavalier DM ( 2012) The CELLULOSE SYNTHASE-LIKE A and CELLULOSE SYNTHASE-LIKE C families: recent advances and future perspectives. Front Plant Sci 3: 109. Google Scholar CrossRef Search ADS PubMed Lloyd C, Chan J ( 2002) Helical microtubule arrays and spiral growth. Plant Cell 14: 2319– 2324. Google Scholar CrossRef Search ADS PubMed Love MI, Huber W, Anders S ( 2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15: 550. Google Scholar CrossRef Search ADS PubMed Mahboubi A, Linden P, Hedenstrom M, Moritz T, Niittyla T ( 2015) 13C tracking after 13CO2 supply revealed diurnal patterns of wood formation in aspen. Plant Physiol 168: 478– 489. Google Scholar CrossRef Search ADS PubMed Majewska-Sawka A, Munster A, Rodriguez-Garcia MI ( 2002) Guard cell wall: immunocytochemical detection of polysaccharide components. J Exp Bot 53: 1067– 1079. Google Scholar CrossRef Search ADS PubMed McFarlane HE, Young RE, Wasteneys GO, Samuels AL ( 2008) Cortical microtubules mark the mucilage secretion domain of the plasma membrane in Arabidopsis seed coat cells. Planta 227: 1363– 1375. Google Scholar CrossRef Search ADS PubMed Mellerowicz EJ, Gorshkova TA ( 2012) Tensional stress generation in gelatinous fibres: a review and possible mechanism based on cell-wall structure and composition. J Exp Bot 63: 551– 565. Google Scholar CrossRef Search ADS PubMed Merced A, Renzaglia K ( 2014) Developmental changes in guard cell wall structure and pectin composition in the moss Funaria: implications for function and evolution of stomata. Ann Bot 114: 1001– 1010. Google Scholar CrossRef Search ADS PubMed Merkle RK, Poppe I ( 1994) Carbohydrate-composition analysis of glycoconjugates by gas-liquid chromatography/mass spectrometry. In: Lennarz WJ, Hart GW (eds) Guide to techniques in glycobiology . Elsevier Academic Press Inc., San Diego, CA, pp 1– 15. Micheli F ( 2001) Pectin methylesterases: cell wall enzymes with important roles in plant physiology. Trends Plant Sci 6: 414– 419. Google Scholar CrossRef Search ADS PubMed Moore JP, Vicre-Gibouin M, Farrant JM, Driouich A ( 2008) Adaptations of higher plant cell walls to water loss: drought vs desiccation. Physiol Plant 134: 237– 245. Google Scholar CrossRef Search ADS PubMed Nick P ( 2013) Microtubules, signalling and abiotic stress. Plant J 75: 309– 323. Google Scholar CrossRef Search ADS PubMed Oakley RV, Wang Y-S, Ramakrishna W, Harding SA, Tsai C-J ( 2007) Differential expansion and expression of α- and β-tubulin gene families in Populus. Plant Physiol 145: 961– 973. Google Scholar CrossRef Search ADS PubMed Olinevich OV, Khokhlova LP ( 2003) Effects of abscisic acid, low temperature, and plant age on cytoskeleton and phosphorylated proteins. Biochem-Moscow 68: 678– 687. Google Scholar CrossRef Search ADS Oomen R, Doeswijk-Voragen CHL, Bush MS et al. ( 2002) In muro fragmentation of the rhamnogalacturonan I backbone in potato (Solanum tuberosum L.) results in a reduction and altered location of the galactan and arabinan side-chains and abnormal periderm development. Plant J 30: 403– 413. Google Scholar CrossRef Search ADS PubMed Paredez AR, Somerville CR, Ehrhardt DW ( 2006) Visualization of cellulose synthase demonstrates functional association with microtubules. Science 312: 1491– 1495. Google Scholar CrossRef Search ADS PubMed Pattathil S, Avci U, Miller J, Hahn M ( 2012) Immunological approaches to plant cell wall and biomass characterization: glycome profiling. In: Himmel ME (ed) Biomass conversion . Humana Press, New York, NY, pp 61– 72. Google Scholar CrossRef Search ADS Ranocha P, Denance N, Vanholme R et al. ( 2010) Walls are thin 1 (WAT1), an Arabidopsis homolog of Medicago truncatula NODULIN21, is a tonoplast-localized protein required for secondary wall formation in fibers. Plant J 63: 469– 483. Google Scholar CrossRef Search ADS PubMed Ranocha P, Dima O, Nagy R et al. ( 2013) Arabidopsis WAT1 is a vacuolar auxin transport facilitator required for auxin homoeostasis. Nat Commun 4: 9. Google Scholar CrossRef Search ADS Richard L, Qin LX, Gadal P, Goldberg R ( 1994) Molecular cloning and characterization of a putative pectin methylesterase cDNA in Arabidopsis thaliana (L). FEBS Lett 355: 135– 139. Google Scholar CrossRef Search ADS PubMed Rui Y, Anderson CT ( 2016) Functional analysis of cellulose and xyloglucan in the walls of stomatal guard cells of Arabidopsis. Plant Physiol 170: 1398– 1419. Google Scholar PubMed Rui Y, Yi HJ, Kandemir B, Wang JZ, Puri VM, Anderson CT ( 2016) Integrating cell biology, image analysis, and computational mechanical modeling to analyze the contributions of cellulose and xyloglucan to stomatal function. Plant Signal Behav 11: 4. Google Scholar CrossRef Search ADS Ryden P, Selvendran RR ( 1990) Cell wall polysaccharides and glycoproteins of parenchymatous tissues of runner bean (Phaseolus coccineus). Biochem J 269: 393– 402. Google Scholar CrossRef Search ADS PubMed Samaj J, Baluska F, Voigt B, Schlicht M, Volkmann D, Menzel D ( 2004) Endocytosis, actin cytoskeleton, and signaling. Plant Physiol 135: 1150– 1161. Google Scholar CrossRef Search ADS PubMed Sanchez A, Shin J, Davis SJ ( 2011) Abiotic stress and the plant circadian clock. Plant Signal Behav 6: 223– 231. Google Scholar CrossRef Search ADS PubMed Seifert GJ ( 2004) Nucleotide sugar interconversions and cell wall biosynthesis: how to bring the inside to the outside. Curr Opin Plant Biol 7: 277– 284. Google Scholar CrossRef Search ADS PubMed Selvendran RR, Oneill MA ( 1987) Isolation and analysis of cell walls from plant material. Methods Biochem Anal 32: 25– 153. Google Scholar PubMed Senechal F, Wattier C, Rusterucci C, Pelloux J ( 2014) Homogalacturonan-modifying enzymes: structure, expression, and roles in plants. J Exp Bot 65: 5125– 5160. Google Scholar CrossRef Search ADS PubMed Solecka D, Zebrowski J, Kacperska A ( 2008) Are pectins involved in cold acclimation and de-acclimation of winter oil-seed rape plants? Ann Bot 101: 521– 530. Google Scholar CrossRef Search ADS PubMed Solomon OL, Berger DK, Myburg AA ( 2010) Diurnal and circadian patterns of gene expression in the developing xylem of Eucalyptus trees. S Afr J Bot 76: 425– 439. Google Scholar CrossRef Search ADS Swamy PS, Hu H, Pattathil S et al. ( 2015) Tubulin perturbation leads to unexpected cell wall modifications and affects stomatal behaviour in Populus. J Exp Bot 66: 6507– 6518. Google Scholar CrossRef Search ADS PubMed Tian GW, Smith D, Gluck S, Baskin TI ( 2004) Higher plant cortical microtubule array analyzed in vitro in the presence of the cell wall. Cell Motil Cytoskeleton 57: 26– 36. Google Scholar CrossRef Search ADS PubMed Updegraff DM ( 1969) Semimicro determination of cellulose inbiological materials. Anal Biochem 32: 420– 424. Google Scholar CrossRef Search ADS PubMed Voxeur A, Hofte H ( 2016) Cell wall integrity signaling in plants: ‘To grow or not to grow that’s the question’. Glycobiology 26: 950– 960. Google Scholar CrossRef Search ADS PubMed Xue L-J, Alabady MS, Mohebbi M, Tsai C-J ( 2015) Exploiting genome variation to improve next-generation sequencing data analysis and genome editing efficiency in Populus tremula × alba 717-1B4. Tree Genet Genomes 11: 82. Google Scholar CrossRef Search ADS Zablackis E, Huang J, Muller B, Darvill AG, Albersheim P ( 1995) Characterization of the cell-wall polysaccharides of Arabidopsis thaliana leaves. Plant Physiol 107: 1129– 1138. Google Scholar CrossRef Search ADS PubMed Zhang GF, Staehelin LA ( 1992) Functional compartmentation of the Golgi apparatus of plant cells – immunocytochemical analysis of high-pressure frozen-substituted and freeze-substituted sycamore maple suspension culture cells. Plant Physiol 99: 1070– 1083. Google Scholar CrossRef Search ADS PubMed Zhu CM, Ganguly A, Baskin TI et al. ( 2015) The Fragile Fiber1 Kinesin contributes to cortical microtubule-mediated trafficking of cell wall components. Plant Physiol 167: 780– 792. Google Scholar CrossRef Search ADS PubMed Author notes handling Editor Christophe Plomion © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: email@example.com
Tree Physiology – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera