ga1-3 provides a system in which cell elongation in the hypocotyl can be rescued conditionally by exogenous application of GA, while gai provides a control for the effects of exogenous GA application. Hypocotyl growth kinetics in wild-type (WT) (Ler), ga1-3, and gai seedlings were established in a continuous light environment with plates positioned horizontally. Hypocotyl growth was measured during a period of 10 d after the culture plates were transferred to the growth room, in the presence and absence of 1 μM exogenous GA₄ (Figure 1A), a concentration that restores hypocotyl length of ga1-3 to WT length 36. In the absence of exogenous GA, WT hypocotyls elongate between 2 and 7 d, and have a final length of around 2 mm. ga1-3 required an extra day to germinate, after which hypocotyl elongation was minimal, reaching only 0.6 mm. gai hypocotyls elongate for up to 6 d, but at a slower rate than the WT, with a maximum length of about 1.6 mm. In the presence of exogenous GA, WT hypocotyls elongate between 2 and 7 d, and have final lengths of approximately 3.5 mm, and such hypocotyls grow longer and at a faster rate than without GA. ga1-3 hypocotyls respond to exogenous GA, elongating for up to 7 d, with final lengths of around 3 mm. Finally, gai does not respond to exogenous GA, having the same hypocotyl growth kinetics and final length as in the absence of the growth regulator, thus confirming its insensitivity to GA. These results are consistent with those reported previously 36. However, in our analysis, final hypocotyl lengths are shorter, probably as a consequence of the inhibitory effects of the continuous light regime used.

Our analysis of WT, ga1-3, and gai hypocotyls and their cell walls used material taken at an equivalent developmental stage; in our case defined as approximately 50% of final hypocotyl length, estimated from the growth curves in Figure 1A and indicated by arrows. This was set at 3 d, both in the presence and absence of GA. However, for ga1-3, in the absence of GA, hypocotyls barely grow following germination. Therefore we analysed hypocotyls at 3 d, the earliest time point following germination. The general morphology of 3-d-old seedlings of average hypocotyl length is shown in Figure 1B. In the absence of exogenous GA, WT hypocotyls are approximately 1 mm long, but are almost twice as long (1.8 mm) when grown in the presence of exogenous GA. In contrast, ga1-3 seedlings are severely dwarfed with hypocotyls at approximately 0.5 mm in length. When grown in the presence of exogenous GA, ga1-3 hypocotyl length is restored to that of untreated WT. In the absence of GA, gai seedlings have slightly shorter hypocotyls than WT, at about 0.8 mm, and are unaffected by exogenous GA. GA-regulation of hypocotyl growth is mediated through elongation of the pre-existing cells with little or no contribution from cell division 36. To test whether continuous light affects this process, epidermal cells were imaged with a field-emission scanning electron microscope (FESEM) (Figure 1B). In the absence of exogenous GA, WT epidermal cells are almost twice as long as those of ga1-3, while gai epidermal cells are slightly shorter than WT. In the presence of exogenous GA, WT epidermal cells approximately double in length, ga1-3 epidermal cell length is increased 2 to 3 fold and gai epidermal cell length is unchanged. The relative differences in epidermal cell length closely match the relative differences in hypocotyl length. As the same relative differences in cell length have also been observed in the cortical and endodermal layers 37, the differences in hypocotyl length are likely to reflect differences in cell length and therefore in cell elongation.

FTIR microspectroscopy has been used to measure the composition of plant cell walls 383940. Small areas of tissue can be selected for analysis, and other advantages include the speed of both sample preparation and data collection. We used FTIR microspectroscopy to quickly ascertain if DE% was associated with dwarfism in primary cell walls of Arabidopsis hypocotyls. Spectra were collected from a 200 × 100 μm area in the central region along the length of WT and ga1-3 hypocotyls, grown in the presence and absence of exogenous GA and at the developmental stages indicated in Figure 1B. The central stele was avoided to prevent contamination from secondary cell wall components. For each population of hypocotyls, DE% was determined semi-quantitatively based on the method described by Filippov and Kohn 41. Table 1 shows cell walls of WT hypocotyls have a DE of about 60% when grown both in the presence and absence of GA. In contrast, DE is lowest in walls of ga1-3 hypocotyls grown without GA, at about 40%, but rises to around 55% when grown in the presence of GA. Thus, GA-promoted cell elongation in ga1-3 hypocotyls is associated with a corresponding rise in DE%.

To more accurately determine pectin DE%, we measured HG content as uronic acid, and methyl-ester content as the amount of methanol released, at the developmental stages described in Figure 1B. Average hypocotyl lengths used in all experiments are shown in Figure 2A. When grown without exogenous GA, WT (Ler) hypocotyls measured 1.06 ± 0.02 mm, and increased to 1.74 ± 0.02 mm in the presence of GA. Dwarf ga1-3 hypocotyls were 0.55 ± 0.02 mm but increased to 1.31 ± 0.03 mm with exogenous GA. Finally, gai hypocotyls measured 0.82 ± 0.01 and 0.86 ± 0.01 mm, when grown without or with GA respectively. Uronic acid and methanol content are expressed as amount per hypocotyl. Since hypocotyl growth is essentially division-free, a change in the amount of a particular wall component can be correlated primarily to cell elongation.

When grown in the absence or presence of GA, WT uronic acid content was 2.31 ± 0.09 and 2.43 ± 0.10 nmol per hypocotyl, respectively, and so was not significantly different between the two treatments (Figure 2B). In ga1-3 hypocotyls, grown without GA, the values were lower than WT, measuring 2.00 ± 0.16 nmol per hypocotyl, and was unchanged at 2.10 ± 0.11 nmol per hypocotyl when grown in the presence of GA. gai hypocotyls contained the lowest amount of uronic acid, at 1.81 ± 0.04 and 1.75 ± 0.09 nmol per hypocotyl when grown without or with GA respectively. As with WT and ga1-3, GA did not affect the uronic content of gai hypocotyls. GA also did not significantly affect methanol released in WT. In the absence of GA, methanol released was 1.38 ± 0.04 nmol per hypocotyl. When grown in the presence of exogenous GA, methanol released from WT was 1.44 ± 0.03 nmol per hypocotyl. Therefore, GA affected the amount of neither uronic acid nor methanol released in WT cell walls, even though hypocotyl length increased almost two-fold over the same period of growth. In contrast, GA increased the amount of methanol released from ga1-3 cell walls, rising from 0.97 ± 0.08 nmol per hypocotyl in the absence of exogenous GA, to 1.23 ± 0.06 nmol per hypocotyl when grown in the presence of GA. GA-stimulated growth therefore correlates with an increase in cell wall methyl-esterification. Finally, gai hypocotyls contained similarly reduced amounts of methanol to ga1-3, at 0.97 ± 0.05 nmol per hypocotyl when grown without GA, and was not significantly altered with GA, at 0.91 ± 0.04 nmol per hypocotyl.

The ratio of methanol to uronic acid content was used to calculate DE% (Figure 2C). In WT hypocotyls this was 60.04 ± 2.23% and 59.08 ± 2.31% in the absence and presence of exogenous GA, respectively. GA therefore promotes cell elongation and hypocotyl growth in WT but does not affect DE%. In contrast, GA did affect DE% in ga1-3 hypocotyls. In the absence of exogenous GA, DE was 48.23 ± 4.00%, rising to 58.89 ± 3.12% when grown in the presence of GA. GA-stimulated growth in the dwarf ga1-3 hypocotyls therefore correlated with the recovery of DE% to WT levels in this mutant. In the semi-dwarf hypocotyls of gai, DE was 53.91 ± 1.08 and 52.25 ± 2.52% when grown either without or with GA, respectively. A correlation therefore exists, between hypocotyl length and DE%. The shortest hypocotyls of ga1-3 have the lowest DE%, but stimulation of hypocotyl extension by GA also increases DE% to the WT level. gai hypocotyl length is intermediary between ga1-3 and WT regardless of GA, as is the measured DE% in this mutant.

In summary, an increase in hypocotyl length, and therefore cell elongation, is also accompanied by an increase in DE%. However, enhanced growth of WT induced by GA does not affect DE%. These data suggest that the degree of pectin esterification may affect cell elongation in a GA-deficient and GA-insensitive background.

To directly test our hypothesis that a low average DE% may constrain growth, we artificially manipulated DE% using reverse genetics. Our prediction would be that reducing the DE% should inhibit hypocotyl elongation. T-DNA insertions into putative PMEs might in principle reduce the potential for de-esterification and ionic cross-linking, leading to an increase in wall extensibility. In Arabidopsis, 67 putative PMEs, in Carbohydrate Esterase Family 8, have been identified based on protein sequences 42. Therefore, the scope for functional redundancy in this family is high, and gene knock-outs might not reveal clear phenotypes. In addition, no PMEs have been biochemically characterised in this species, and some may actually be pectin trans-esterases 4344. For the same reasons, homologous over-expression of endogenous or other plant putative PMEs, without biochemical characterisation, may give results that are difficult to interpret 1920. In contrast, several bona fide PMEs have been reported in bacteria and fungi 4546. In Aspergillus aculeatus, the PME1 gene has been rigorously tested and biochemically characterised 47. We therefore transformed the PME1 cDNA clone into Arabidopsis under the control of a constitutive promoter. Interestingly, constitutive expression of PME1 yielded no transformants and therefore is probably lethal.

Analysis of the predicted signal peptide region using pSORT showed a low probability of the PME1 protein localising to the cell wall in plants. Therefore, we removed the signal peptide sequence and replaced it with one from a putative PME from Arabidopsis (At4g12390) that had a high probability of targeting the protein to the cell wall. The ethanol-inducible expression system was used 48, in which the chimeric construct was cloned downstream of the AlcA promoter, and then transformed into line P5-3 carrying the AlcR promoter. Several independent lines carrying the transgene were identified by PCR using gene-specific primers. To induce expression of the transgene, seedlings were grown for 3 d in continuous light with plates in a near vertical position, and then transferred to induction medium containing 0.1% ethanol in the solidified medium. Transfer at this time point, allowed germination to take place and hypocotyls to enter the rapid phase of elongation. Two lines, PME01 and PME08, in which hypocotyl growth was affected only in the presence of ethanol, were selected for further analysis.

Hypocotyl growth kinetics are shown in Figure 3. In the absence of ethanol, P5-3 hypocotyls grew over a period of 6 d, from day 2 to day 8, with a final length of 5.56 ± 0.17 mm (Figure 3A). The concentration of ethanol used to induce PME1 expression did not affect either the growth profile or final length of P5-3 hypocotyls, which measured 5.77 ± 0.29 mm at day 10. However, compared to previous experiments (Figure 1A), the duration and extent of hypocotyl elongation was increased when plates were positioned vertically, and may be the result of additional nutrient uptake and/or touch responses from being in contact with the surface of the growing medium. In the absence of ethanol, both PME01 and PME08 hypocotyls followed a similar growth profile as P5-3. Final lengths were 5.67 ± 0.22 and 5.25 ± 0.21 mm in lines PME01 and PME08, respectively (Figure 3B, C). However, transfer of the seedlings to induction medium resulted in a deflection of the growth curve for both expressing lines. Hypocotyls stopped growing about 1 d earlier, and final lengths were 4.63 ± 0.24 and 4.27 ± 0.23 mm, respectively, representing a length reduction of about 20%.

Transcriptional and cell wall analysis was performed on excised hypocotyls after 2 d growing on control/induction medium (arrows in Figure 3). At this time point (day 5), the A. aculeatus PME was strongly expressed in both lines when grown in the presence of ethanol, whereas no expression was detected in seedlings grown on ethanol-free medium or in P5-3 (Figure 4). Expression was stronger in line PME08 compared to PME01. Both parental lines had reduced seed yield, which may be a consequence of auto-induced PME1 expression during seed set, and/or during pollen tetrad separation, the latter involving PME 49. Thus, it was difficult to collect enough transgenic hypocotyls for direct chemical analysis. Therefore, to confirm that the growth effects were due to pectin de-esterification, we again used FTIR microspectroscopy of individual hypocotyls to measure DE% indirectly (Table 2). At this time point, hypocotyl lengths in P5-3 were 4.52 ± 0.19 and 4.46 ± 0.30 mm when grown in the absence and presence of ethanol, respectively. In the absence of ethanol, PME01 hypocotyls were 4.25 ± 0.19 mm long, compared to 3.60 ± 0.24 mm when grown on induction medium. Similarly, PME08 hypocotyls were 4.08 ± 0.33 and 3.15 ± 0.29 mm after 2 d growth on control and induction medium, respectively. Induced expression of PME1 therefore corresponded to a 15% reduction in average hypocotyl length in line PME01, and a 22% reduction in line PME08, compared to non-induced seedlings. DE in P5-3 hypocotyls was about 48% in the absence of ethanol, and about 45% in the presence (Table 2). In line PME01, DE was about 48% in the absence of ethanol, but only about 40% following induction. In line PME08, DE was about 42% in the absence of ethanol, and reduced to about 38% when induced. The overall reduction in DE in P5-3, from about 60% (Table 1) to about 48% (Table 2), may be due to the slowing down of hypocotyl elongation at day 5, as opposed to day 3 when they are growing fastest. Nevertheless, the lowest DE% we measured, in both lines, followed PME1 induction. In summary, PME1 expression corresponded to a reduction in cell wall DE% and hypocotyl length in both lines. Expression was strongest in line PME08 in which we measured both the lowest DE% and the shortest hypocotyls.

Arabidopsis thaliana (L. Heynh) ecotype Landsberg erecta (Ler) was used as the reference wild-type (WT). In the over-expression experiment, line P5-3 (also in the Ler background) was used as WT. Seeds were surface-sterilised by immersion for 5 min in 5% (v/v) Vortex bleach (Procter & Gamble Ltd, containing 5 to 15% chlorine-based bleach), and washed three times in sterile distilled water (sdH₂O). Following sterilisation, to allow seeds of ga1-3 to germinate, they were incubated at 4°C for 5 d in a solution of 1 μM GA₄ (Sigma-Aldrich, UK) 36. Ler and gai do not require this treatment but were included for consistency. Next, seeds were rinsed five times in sdH₂O and sown onto medium containing 1× Murashige and Skoog (MS) basal salts (micro and macro elements) (Duchefa) supplemented with 3% (w/v) sucrose (pH adjusted to 5.7) and solidified with 0.5% (w/v) Phytagel™ (Sigma-Aldrich, UK). Approximately 20 seeds were evenly sown per 9 cm Petri plate (Bibby Sterilin Ltd) containing 20 mL of growing medium, and plates sealed with Parafilm^® laboratory film (Pechiney Plastic Packaging, Menasha, USA). Plates were placed in darkness at 4°C for 48 h to stimulate and synchronise germination. Following cold treatment, plates were transferred to a growth room maintained at 25°C and incubated horizontally under fluorescent lamps (70 μmol m^-2 s^-1) in a continuous white light regime.

Hypocotyl length was determined as the distance between the top of the collet root hairs, to the 'V' made by the cotyledon shoulder 73. Hypocotyl lengths were measured using a Leica WILD M10 binocular microscope fitted with an eye-piece graticule, and the mean ± SE calculated for each data set.

Seedlings were mounted in a horizontal position on adhesive carbon tabs (Agar Scientific Ltd) and plunge-frozen at -210°C in liquid nitrogen slush. After freezing, samples were immediately loaded into the cryo chamber of the scanning electron microscope, equilibrated with the stage and sublimed at -100°C for 2 min. The temperature was returned to -110°C, the samples were sputter-coated with platinum for 2 min at 10 mA, and then transferred to the imaging stage at -130 to -150°C for analysis. FESEM images of hypocotyl epidermal cells were obtained using a Philips XL30 FEG scanning electron microscope (FEI Co., Eindhoven, The Netherlands) fitted with a cryostage (CT1500 HF; Oxford Instruments, Abingdon, Oxford, UK), operating at 3 kV and a working distance of between 5 and 15 mm.

Whole hypocotyls were excised from seedlings and suspended on the surface of water-soaked tissue paper to prevent tissue dehydration during sample collection. This also effectively rinsed the samples. The samples were compressed onto barium fluoride (BaF₂) windows (13 × 2 mm) (Crystran Ltd, Poole, UK), dried at 60°C for 1 h and used immediately for spectral acquisition, or stored overnight at 4°C and used the next day. Windows were supported on the stage of a UMA500 microscope accessory of a Bio-Rad FTS175c spectrometer equipped with a liquid nitrogen-cooled mercury cadmium telluride detector and absorbance spectra obtained. Sixty-four interferograms were collected in transmission mode with 8 cm^-1 resolution and co-added to improve the signal-to-noise-ratio for each sample. An area of approximately 200 × 100 μm in the middle region (along the longitudinal axis) of each hypocotyl was selected, avoiding the central stele. One spectrum was collected from each hypocotyl and between 10 and 28 samples for each genotype/treatment used. For each population the spectra were averaged between 790 and 1810 cm^-1 and each average spectrum baseline-corrected and area-normalised to account for differences in sample thickness. Processing of spectral data was done using OMNIC E.S.P. 5.0 software. For each spectrum, a two-point baseline was constructed between 870 and 1810 cm^-1. The absorbance maxima of bands υ_as(COO^-) 1605 cm^-1 and υ(C = O)_ester 1745 cm^-1 from the baseline were measured, and the log ratio of these values used to semi-quantitatively calculate DE% from the calibration curve of Filippov and Kohn 41. For each genotype/treatment, values were averaged ± SE.

Hypocotyls were excised precisely using fine-tipped forceps and a razor blade. Upon excision, samples were transferred to a 1.7 mL microfuge tube containing 1 mL absolute ethanol and heated to 85°C for 20 min to extract chlorophyll, sugars and other small molecules. An additional extraction was made in 1 mL 80% (v/v) ethanol at 85°C for a further 20 min, and then rinsed three times in 1 mL sdH₂O. Samples were suspended in a small volume of sdH₂O and freeze-dried. Each tube contained between 50 and 100 hypocotyls. Uronic acid assays were performed on these as described previously 74. Methyl-esters were determined as the amount of methanol released following saponification using the method described by Kim and Carpita 17. Values are expressed as nmol per hypocotyl. For each genotype and treatment, duplicate or triplicate samples were used in each experiment, and each experiment performed three times. In total, 600–900 hypocotyls were used to independently calculate average uronic acid and average methanol values. The ratio of methanol to uronic acid was used to calculate DE%. Thus in total, between 1200 and 1800 hypocotyls were used to derive the average DE% for each genotype/treatment. Standard error values were ratioed as described previously 75.

The open reading frame of PME1 (Accession no: U49378) from Aspergillus aculeatus 47, minus the predicted signal peptide sequence, was PCR amplified out of pYES 2.0 using the forward primer OVEXP3 (5'-CTGCCAATCCACCATAGCCGCCAGCCGTACCACGGCTCC-3') and the reverse primer OVEXP4 (5'-GGC

GAATTC

GGATCC

Seeds were prepared as described above and sown onto sterile filter paper in contact with growing medium containing 1% (w/v) sucrose. Sealed plates were incubated in a near vertical position. This allowed hypocotyls to be measured each day without opening plates, which would have resulted in some loss of ethanol vapour (see below). After 3 d seedlings were carefully transferred to the same medium containing no ethanol (control medium) or to induction medium containing 0.1% (v/v) ethanol. Induction medium was prepared by adding the appropriate volume of 50% (v/v) of ethanol to the molten medium cooled just to the point at which it started to solidify in order to prevent ethanol evaporation. Following transfer, plates were resealed with Parafilm. Hypocotyl lengths were imaged digitally and measured using Photoshop 5.0 software.

RNA was extracted from whole seedlings at 2 d after transfer to induction/control medium, using a QIAGEN RNeasy Plant minikit according to the manufacturer's instructions. RNA yield was quantified by spectrophotometry and concentrations equalised with RNase-free water. After DNase treatment (40 units DNaseI; Amersham Pharmacia) for 20 min at 37°C, 2.5 μg was reverse transcribed for 60 min at 42°C in a final volume of 20 μL in the presence of 20 units RNA guard, 1 mM dNTPs, 5 mM MgCl₂, 0.3 μM oligo(dT) primers and 4 units M-MLV reverse transcriptase (Life Technologies) in the reaction buffer provided. Reactions were stopped by heat inactivation and 80 μL H₂O added. 2 μL of the reverse transcription reaction were used for PCR amplification. The forward primer PMEfor (5'-GTACCACGGCTCCCTCCG-3') and the reverse primer PMErev (5'-GTAGGAGGTATCGACCCAGC-3') gave a 932 bp product for the transgene cDNA. The forward primer Actin2-5' (5'-CTAAGCTCTCAAGATCAAAGGCTTA-3') and the reverse primer Actin2-3' (5'-ACTAAAACGCAAAACGAAAGCGGTT-3') amplified a 220 bp fragment of ACT2 cDNA and used as a semi-quantitative control 77. For controls, 25 cycles of PCR were conducted (30 s at 94°C, 30 s at 55°C, 1 min at 72°C) in a final volume of 20 μL containing 2 μL cDNA, 1 mM dNTPs, 5 mM MgCl₂, 0.3 μM Actin forward/Actin reverse primers and 0.5 units of Taq DNA polymerase (Life Technologies) in the reaction buffer provided. For quantification of the PME1 transgene 30 cycles of PCR were conducted as described above using PMEfor/PMErev primers. The latter reaction was also used to confirm presence of the transgene following Basta selection.