The chemical composition of standard cotton dust.

Cotton dust samples from Cotton Incorporated were investigated by X-ray fluorescence and proximate analysis methods. These dust samples are known as "standard cotton dust" and have been used by many researchers investigating the causative agent(s) and physiological mechanisms of byssinosis. Silicon, calcium, potassium, and aluminum were present in relatively high concentrations (1-4%) in the dust fractions. The ash content of the dust fractions increased as the fraction particle size decreased. Proximate analyses demonstrate that "noncellulosic organics" are the major class of constituents (35-45%) in cotton dust. Cellulose comprises only 10-15% of the dust, while water-extractable materials comprise approximately 20% of the dust. Capillary gas chromatography performed on silylated, freeze-dried, aqueous extracts of the less than or equal to 38 micron dust fraction revealed the presence of phosphate, malate, arabitol, citrate, fructose, glucose, and mannitol.


Introduction
Searches for the causative agent(s) of byssinosis, proposed models for acute and chronic physiological response to cotton dust inhalation, and even the definition of the term, "byssinosis," have been objects of considerable research, debate, and controversy over the past decade (1)(2)(3)(4)(5)(6)(7). Direct comparison of experimental findings has been hampered by use of cotton dusts or leaf and bract materials from various sources. To alleviate this problem, Cotton Incorporated proposed production and use of a single source cotton dust that would be made available to cotton dust researchers worldwide (8). Fractions of -38 ,um, 38-75 ,um, and , 75 ,um cotton dust were sieved from crude dust collected in a West Texas cotton mill by Cotton Incorporated in 1981-1982 and 1982-1983. Samples of "standard cotton dust" furnished by Cotton Incorporated have been used by researchers in the United States, Sweden, and Wales to study the chemical, physical, biological, and physiological properties of cotton dust. Use of this dust allowed scientists to compare their data and hypotheses with those produced by other scientists. In order to provide fundamental data on the chemical composition of "standard cotton dust," we have chemically characterized "standard dust" fractions by X-ray fluorescence and proximate analysis methods. In addition, silylated, freeze-dried, aqueous extracts of the -38 ,um fraction were analyzed by capillary gas chromatography. *Southern Regional Research Center, P. 0. Box 19687, New Orleans, LA 70179.

Experimental
The proximate analysis method consists of a serial solvent extraction procedure followed by gravimetric analysis to determine moisture content as the loss on oven-drying overnight at 110°C, water-extractables as the loss on extraction with boiling 0.1% Triton X-100 in deionized water, ethanol-extractables as the loss on extraction with boiling 100% ethanol, ethanolamine-extractables as the loss on extraction with boiling ethanolamine, cellulose as the loss on extraction with 1.00 M cupriethylenediamine (cuene), and insoluble residue. The method of Brown and co-workers was modified through use of Triton X-100 in the initial extraction to ensure thorough wetting of the dust sample (9).
Ash content is determined by slowly heating sample replicates to 725°C. The replicates are then ashed at 725°C for 2 hr.
Elemental analyses were performed by energy-dispersive X-ray fluorescence on an EG&G Ortec Spectrometer.
Samples were prepared for gas chromatographic analysis by extracting 40 mg of dust with 1.0 mL of pyrogen-free water for 1 hr at room temperature, passing the solution through a syringe filter (0.45 ,um, Gelman Acrodisc), and pipetting 200 ,uL portions of the filtrate into derivatization vials-10 ,iL of glutaric acid standard (equivalent to 2.0 x 10-g) was added, and the solutions were freeze-dried at 40 mTorr; 200 ,uL of Tri-Sil was added to each vial and was allowed to react at 50°C in a heating module. Gas chromatograms were recorded after 0.5 hr and 1.5 hr reaction times. Gas chromatography was performed on an HP 5880 Chromatograph equipped with a crosslinked dimethyl silicone high-performance capillary column (12.5 m, 0.20 mm I.D.) and a flame ionization detector. The following temperature and flow parameters were used: injector, 200°C; detector, 250°C; oven temperature program, 100°C for 1.00

Results and Discussion
Proximate analyses data shown in Table 1 demonstrate that cotton dust is composed primarily of ethanolamine-extractable material (35-45%) and water-extractable material (ca. 20%). The boiling ethanolamine extraction procedure is used to remove a class of compounds known as "noncellulosic organics." Water extraction removes polar organic molecules and some of the inorganic salts, particularly the potassium, calcium, aluminum, and magnesium salts. Cellulose, the major constituent of raw cotton lint, accounts for only 10 to 15% of the dust. This is comparable to the amount of cellulose (ca. 10%) found in dried cotton leaf tissue (unpublished observation). Cellulose contents ranging from 38 to 59% have been reported for "standard dust" samples (8). However, this report did not indicate the method of cellulose determination. Our method removes water-soluble, ethanol-soluble, and ethanolamine-soluble substances prior to dissolution of the cellulose. Our method takes advantage of the structure of plant materials in that the outer layers of extracellular and cellular material are removed before the cellulose framework is dissolved. The ash contents of the standard dust fractions increase with decreasing particle size ( Table 2). This trend of increasing ash content with decreasing particle size is also observed for cotton dust samples generated by aMean ± SEM, n = 2. b Samples received as separate shipments from Cotton Inc. a variety of techniques (10). The -38 ,um fraction has a high ash content: 34%. This is not unexpected for Texas cotton dust samples. For example, cardroom dust samples from three Texas-grown cottons were shown to have ash contents of 31 to 36% (11). The high ash contents were attributed to soil contamination.
Consistent with the trends in ash contents, elemental analyses show higher levels of silicon, iron, zinc, and manganese in the -38 ,um fraction than in the 3 75 and 38-75 ,unm fractions (Table 3). Silicon, iron, and manganese are commonly found in soils. Variations of silicon, iron, manganese, and ash content with particle size indicate that soil particles in the dust are smaller on the average than many ofthe particles ofplant origin. Thus, for these fractions, the inorganic content of the dust increases with decreasing particle size.
In order for a dust to serve as a standard, it is necessary that samples taken from the standard be chemically invariant. Elemental analyses of two separate 1981-1982 samples and one 1982-1983 sample were conducted to test for variations. The results are shown in Table 4. Variations in the levels of several important elements are minor. Calcium, chlorine, sulfur, and zinc levels were slightly higher in the 1982-1983 sample than b .W." ; d 6' 0 U, S in the 1981-1982 samples. The -38 pum standard cotton dust samples have silicon contents of 3-4%, calcium contents of 2%, aluminum contents of 2%, potassium contents of 1%, and magnesium contents of 1%. By comparison, Wirtz and Olenchock (12) investigated several grain dusts, including flax dust, and found silicon contents of 0.35 to > 1%, calcium contents of 0.1 to > 1%, aluminum contents of 0.05 to > 1%, potassium contents of 0.17 to > 1%, and magnesium contents of 0.039 to > 1%.
Gas chromatograms of derivatized, freeze-dried, aqueous extracts of s-38 ,um standard cotton dust show approximately 10 to 15 major volatile components other than those present in controls (Fig. 1). Retention time matching and peak enhancement studies are consistent with assignment of ten of these peaks to phosphate, malate, arabitol, citrate, fructose (three major peaks), glucose (two major peaks), and mannitol. Retention times and peak identities are shown in Table 5. These compounds are characteristic of plant extracts and have been found in aqueous extracts of cotton lint (13,14), cardroom dust (15), and cotton bract (unpublished observation). Glycerol, an expected plant product, is also present in the controls. It is leached from the plastic syringe filters. These chro- matographic studies demonstrate that vegetative matter is an important component of the standard cotton dust. The high levels of malate and citrate and the relatively low levels of arabitol and mannitol indicate that the vegetative matter in the dust was recently photosynthetically active and was only in the early stages of degradation at the time of harvest (15). The level of senescence or degradation of the vegetative matter may be an important determinant of pulmonary response to cotton dust. Studies in this area are continuing in our laboratories.

Conclusions
Standard cotton dust contains approximately 20% water-extractable material, approximately 40% ethanolamine extractable material, and 10 to 15% cellulose. Ash contents and levels of silicon, iron, and manganese of standard cotton dust fractions increase with decreasing particle size. Silicon, calcium, potassium, aluminum, and magnesium compounds are present in the 1 to 4% range in the dust fractions. The chemical composition of standard cotton dust fractions from 1981-1982 and 1982-1983 are similar. Malate, citrate, fructose, glucose, arabitol, and mannitol are major components of derivatized, freeze-dried, aqueous extracts of standard cotton dust. These elemental, proximate, and chromatographic analyses support the contention that standard cotton dust can be an important resource for all scientists trying to develop meaningful bioassays for byssinosis, in order to eventually identify and eliminate the causes of the disease.
We thank R. R. Jacobs of Cotton Incorporated, Raleigh, NC, for samples of "standard cotton dust." We thank Chauncy Williams for proximate and ash analyses, Biagio Piccolo and Charles McCombs for X-ray fluorescence analyses, and Sharon McDonald and Virginia Fernandez for gas chromatographic analyses.
Use of a company and/or product name does not imply approval or recommendation of the product or company to the exclusion of others which may also be suitable.