Evidence is presented for the existence of a noncellulosic 8-1,3-glucan in cotton fibers. The glucan can be isolated as distinct fractions of varying solubility. When fibers are homogenized rigorously in aqueous buffer, part of the total fi-1,3-glucan is found as a soluble polymer in homogenates freed of cell walls. The proportion of total /P-1,3-glucan which is found as the soluble polymer varies somewhat as a function of fiber age. The insoluble fraction of the ,8-1,3-glucan remains associated with the cel waUl fraction. Of this cell wall ,8-1,3-glucan, a variable portion can be solubilized by treatment of waUls with hot water, a further portion can be solubilized by alkaline extraction of the walls, and 17 to 29% of the glucan remains associated with cellulose even after alkaline extraction. A portion of this glucan can also be removed from the cell walls of intact cotton fibers by digestion with an endo-,8-1,3-glucanase. The glucan fraction which can be isolated as a soluble polymer in homogenates freed of cell waUs is not associated with membranous material, and we propose that it represents glucan which is also extracellular but not tightly associated with the cel wall. Enzyme digestion studies indicate that al of the cotton fiber glucan is fl-linked, and methylation analyses and enzyme studies both show that the predominant linkage in the glucan is I -* 3. The possibility of some minor branching at C-6 can also be deduced from the methylation analyses.The timing of deposition of the f8-1,3-glucan during fiber development coincides closely with the onset of secondary wal celulose synthesis.Kinetic studies performed with ovules and fibers cultured in vitro show that incorporation of radioactivity from I'4Clglucose into ,8-1,3-glucan is linear with respect to time almost from the start of the labeling period; however, a lag is observed before incorporation into cellulose becomes linear with time, suggesting that these two different glucans are not polymerized directly from the same substrate pool. Pulse-chase experiments indicate that neither the f8-1,3-glucan nor cellulose exhibits significant turnover after synthesis.In our search for a cellulose synthetase in developing cotton fibers, we discovered, instead, a highly active UDP-glucose:,8-1,3-glucan synthetase, the properties of which were described recently in detail (4, 6). Previous work from our laboratory (13) Preparation of Cell Walls and Soluble ,-1,3-Glucan. Cell walls were prepared from fibers of various ages as described previously (13) but with the following modifications. Fibers were harvested into liquid N2, then placed directly into 0.05 M Tes buffer (pH 7.0) at 4 C and homogenized in a glass tissue homogenizer. This homogenate was centrifuged at 4 C for 10 min at 5,500g. The resulting supernatant was brought to 80%o (v/v) with absolute ethanol and allowed to stand for at least 2 h at -20 C or overnight at 4 C. The resulting precipitate was washed twice with cold 80o ethanol. This precipitate was the fraction containing the glucan re...

MATERIALS AND METHODSIn Vitro Synthesis of Mannolipid. Preparation of extracts and production of mannolipid were scaled up from the general procedure of Ericson and Delmer (5). Developing seeds of P. vulgaris (14-17 mm in length) were harvested and the seed coats and embryos removed. The cotyledons (220 g fresh wt) were chopped with a razor blade at 4 C, and ground in a chilled mortar with 110 ml of 0.05 M tris-HCl (pH 8.5) containing 1 mM EDTA. The homogenate was centrifuged at 800g for 10 min and the supematant used for synthesis of mannolipid. Onequarter of the extract was used for preparation of high specific radioactivity mannolipid and the remainder for synthesis of larger quantities of low specific radioactivity mannolipid. Conditions for synthesis of the high specific radioactivity preparation were: crude extract, 28.5 ml; 0.2 M MnC12, 1.5 ml; and [GDP-'4C]mannose (New England Nuclear, specific radioactivity 210 mCi/mmol, 20 ,.Ci/ml), 0.075 ml. For the low specific radioactivity preparation, conditions were: crude extract, 120 ml; 0.2 M MnCl2, 6 ml; [GDP-14C]mannose (2.03 mCi/mmol, 9.09 tCi/ ml), 0.22 ml. The reactions were incubated for 15 min at 23 C and terminated, and the mannolipid extracted into chloroformmethanol as described previously (5)

In our studies on cellulose biosynthesis in the developing cotton fiber, we have been investigating the utilization of UDPglucose by enzyme systems present in the cotton fiber. We have done this because, for a variety of reasons (3), UDP-glucose can be considered as a likely precursor to cellulose. However, UDPglucose can also serve as a precursor to such compounds as sucrose (4) and steryl glucosides (8, 15) in plants, and it is well known that in a variety of plant tissues, UDP-glucose can also serve as a substrate for the synthesis of /-,(1 -* 3)-glucan (2, 5, 6, 20-22), or for the synthesis of a glucan with mixed f3-(1 -+ 3) and 18-(1 --4) linkages (14, 19).We have studied the utilization of UDP-glucose by cotton fibers harvested at the onset of secondary wall cellulose synthesis, determining and comparing the activities both in fibers detached from the ovules ("semi-intact fibers"), and in a particulate fraction isolated from homogenized fibers. Since UDPglucose can serve as substrate for several different enzyme systems, we have analyzed for a variety of possible reaction prod- (224 ,uCi/,umol) was purchased from New England Nuclear. f8-(1--+3)-Glucanase purified from Rhizopus was the generous gift of E. T. Reese, U.S. Army Laboratories, Natick, Mass. Laminaribiose was prepared by digestion of laminarin (Sigma) with this 63(1-.3)-glucanase.The disaccharide was purified from the digestion mixture by chromatography on a Bio-Gel P-2 column.Determination of Radioactivity. Samples were dried on paper or filters and counted in toluene scintillation fluid (2.2 g PPO and 176 mg POPOP/l toluene). Aqueous samples, samples in chloroform-methanol, or undried fibers were counted in the same toluene scintillation fluid to which '/lo volume of Bio-Solve (Beckman) has been added. All samples were counted in a Packard Tri-Carb model 3390 liquid scintillation spectrometer. Quench curves for various counting conditions were constructed, and where appropriate, corrections to dpm were made.Detached Fiber Assay. Fibers of known age were removed from the boll, divided into 50-mg aliquots, and placed in tubes containing 2 ml 0.01 M TES buffer (pH 7.5) containing 1 mm EDTA. The fibers were agitated on a Vortex mixer with a small spatula in the tube until thoroughly wetted (about 30 sec). The individual aliquots were then removed from the tubes, drained on paper towels, and a single aliquot was placed into each reaction mixture at room temperature to start the enzyme assay.