*4.4. Extensiometry*

Segments of 20 mm were cut from the top growing part of the sunflower hypocotyls, longitudinally bisected using a double-edged razor blade, and then frozen using freezing spray (R.A. Lamb, Eastbourne, UK). After 60 s, the segments were thawed in MES buffer. The segments were then pressed between microscope slides wrapped with absorbent paper, using about 2 kg of weight for 60 s, and returned to MES buffer. Segments of cellulose or cellulose composite 2 mm × 1 mm in cross-section and about 10 mm in length were

cut from pellicles using a pair of razor blades held in a spaced block and transferred to MES buffer.

Constant load extensiometry was carried out as described in [21]. Segments of plant material, bacterial cellulose, or cellulose composite were fixed into a reservoir tube with about 10 mm exposed between a clamp at the bottom of the tune and another clamp attached to one arm of a counterweighted cantilever. The reservoir tube was then filled with buffer. The exact initial length of the exposed portion of the segment was measured using a magnifying eyepiece with a graticule (extensiometer custom built by the Biological Sciences workshop at the Lancaster University, Lancaster, UK).

The counterweight positions were adjusted so that a negligible force was applied to the samples for 20 min to allow them to settle, after which the force was increased in increments of 0.098 N (10 g) by sliding the counterweights along the cantilever arms. The change in length was measured using a linear variable displacement transducer (Schlumberger DFG 2.5 from RS Components Ltd., Corby, Northants, UK), which gave a voltage output based on the position of a core balanced on the opposite end of the cantilever arm in relation to the one to which the clamped segment was attached. The mean of 1000 individual readings was recorded every 30 s by a Bede PCADH24 analogue input card (Bede Technology, Jarrow, Tyne and Wear, UK). The extensiometer can be used to test up to six samples at a time. All extensiometry experiments were conducted at least twice, using at least two replicates of each treatment in each run. The data presented are examples of single tests, because although the replicate data exhibited the same patterns of response, they are not normally closely enough matched for pooling of multiple runs to be feasible.

The buffer used throughout (referred to as MES buffer) was 10 mM MES buffer containing 5 mM KCl and 1 mM CaCl2, titrated to pH 5.0 with 1 M NaOH. The pH of buffers containing PEG or other osmotica was adjusted to pH 5.0 after their addition. The effect of changing the water potential on segment extension was observed by draining MES buffer from the extensiometer reservoirs and replacing it with the same buffer containing PEG 6000 or by carrying out the opposite exchange. PEG osmotic pressure (and therefore water potential) was determined by vapor pressure osmometry using a Vapor® 5520 osmometer (Wescor Inc., South Logan, UT, USA). Melting point depression measurements of the osmotic pressures of PEG solutions are higher than those made using vapor pressure [60], but it has been plausibly argued that this is because of the effect of temperature on higher order virial coefficients of the Van't Hoff equation in PEG solutions and therefore that measurements from melting point, which are necessarily at low temperatures, do not accurately reflect the osmotic pressure under the measurement conditions [61], and vapor pressure measurements were preferred. From these data, 0.27 g PEG 6000 g water−<sup>1</sup> gave an osmotic potential of −0.62 MPa.

#### *4.5. Viscosity of Pectin Solutions*

The viscosity of 2% (*w*/*v*) solutions of apple pectin (Sigma-Aldrich, cat. No. 76282, 70–75% esterified) containing 5 mM CaCl2 adjusted to the required pH by dropwise addition of 2 M KOH were tested by timing how long the solutions took to flow between two graduations of a 1 mL graduated glass pipette. The same pipette was used throughout for consistency and was wetted with the pectin solutions before measurements were carried out. The viscosity of 2% (*w*/*v*) solutions of apple pectin (Sigma-Aldrich, cat. No. 76282, 70–75% esterified), containing 10 mM CaCl2 adjusted to the required pH by dropwise addition of 2 M KOH, were also visualized by inverting capped disposable centrifuge tubes for 5 min. Measurements were carried out at room temperature.
