Characterization of coffee silverskin: opportunities and problems for a real use

Coffee represents one of the most widely consumed beverages obtained by roasting green coffee. During this technological phase, the characteristic properties of coffee beverage (such as flavor and aroma) are developed, and the remnant thin tegument that covers and protects the outer layer of the green coffee bean (coffee silverskin, CS) is removed. The re-utilization of CS could be an alternative to its environmental disposal; due to its interesting chemical composition, a demand for CS revalorization has lately increased. Many authors have proposed CS as natural source of several compounds having positive effects on human health. However, CS could also contain undesirable compounds, such as ochratoxin A (OTA), which has been classified by the International Agency for Research on Cancer as a possible human carcinogen (group 2B); contradictory results about the effect of roasting process on reduction of OTA in coffee beans have been reported, though. The lipid fraction of CS could be a source of bioactive molecules, such as phytosterols (PS), but it is necessary to control them in terms of their conversion into risky molecular species for human health, such as phytosterols oxidation products (POPs). A study on lipophilic and hydrophilic components of CS was carried out, by weekly collecting it (5 samplings) from a local medium enterprise. The determination of OTA (controlled for 10 weeks) and POPs was performed, together with the determination of fiber content, phenolic compounds, carbohydrates composition and caffeine amount. In addition, different methods for extracting the lipid fraction were tested; the fatty acid composition and total phytosterol content were also investigated. Moisture (2.68-10.34% CS) significantly changed (P≤0.05) during the study, probably due to the uncontrolled hydration system used to press and store CS. Total dietary fibre (68.9-79.7% CS) was composed by about 85% insoluble dietary fibre and 15% soluble dietary fibre, as reported in literature. In addition, the high content of caffeine (0.80-1.04% CS) detected suggested CS as possible source of such xanthine. Total carbohydrates (0.94-1.28% CS) were mainly composed by fructose (0.40-0.56% CS), sucrose (0.10-0.27% CS), mannitol (0.13-0.20% CS), glucose (0.10-0.19% CS) and inositol (0.09-0.19% CS). The composition of total polyphenols (3918.7-7292.2 mg/kg eq gallic acid) was investigated by HPLC-MS; neochlorogenic, chlorogenic, caffeic, feruloylquinic and dicaffeoylquinic acids were the main phenols found in CS. Different methods were tested in order to obtain the lipid fraction; in particular, the method suggested by Folch et al. (1957) (A) was compared with the Soxhlet one (1879) (B). The lipid matter (5.2%) obtained by method A was mainly composed by free fatty acids (58%), free sterols (17%), diacylglycerols (12%), triacylglycerols (9%) and esterified sterols (3.7%), whereas the lipid fraction (3.4%) extracted with method B was essentially constituted by triacylglycerols (48%), followed by free fatty acids (21%), free sterols (13%), esterified sterols (15%) and diacylglycerols (4%). The most abundant fatty acid was linoleic acid (0.36-0.70% CS), followed by palmitic acid (0.35-0.63% CS) and oleic acid (0.07-0.15% CS). Total phytosterols ranged from 7.4% and 10.5% of lipid fraction, being β-sitosterol the main sterol (5.25-8.02% lipids, 0.16-0.25% CS), followed by campesterol (0.95-1.56% lipids, 0.04-0.05% CS), stigmasterol (0.79-1.11% lipids, 0.03-0.04% CS) and 5-avenasterol (0.38-0.47% lipids, 0.01-0.02% CS). It should be noted that PS levels in CS were about 10 times higher than those found in green coffee; however, they can undergo oxidation during roasting and give rise to POPs. Total POPs was about 3.6% of lipid matter, being the most abundant those arising from β-sitosterol; in fact, 7α-hydroxysitosterol (1.70% lipids) was the main oxyphytosterol, followed by 7-ketositosterol (0.60% lipid), 7β-hydroxysitosterol (0.42% lipids), β-epoxysitosterol (0.09% lipids), α-epoxysitosterol (0.07% lipids) and triolsitosterol (0.07% lipids). Lower amounts of POPs deriving from oxidation of campesterol and stigmasterol were found, being 7α-hydroxyl and 7-keto derivatives were the main ones. Epoxy isomers of these minor phytosterols and 5-avenasterol oxidation products, were not found, probably due to the low amount of such sterols present in CS. Finally, the content of OTA (18.7-34.4 μg/kg CS) was higher than those defined by the Commission Regulation (EC) (2006), which states that the maximum level of OTA in roasted coffee and soluble coffee should be 5 μg/kg and 10 μg/kg, respectively. However, it must be pointed out that CS covers the coffee bean, so it is reasonable that the isolated CS showed a higher level of such mycotoxin. Therefore, a solution for the removal or dilution of OTA from CS should be studied, if CS is intended for human or livestock consumption. Data obtained in the present work suggests that CS could be used as source of cellulose for paper production, as well as a source of bioactive compounds to be employed as ingredients in pharmaceutical/cosmetic industries or for development of functional food. In the case of human or livestock use, a rigorous quality control should be performed, eventually providing a procedure set-up to reduce the amount of OTA and POPs.