Application and sensory evaluation of enzymatically texturised vegetable proteins in food models

Abstract

Using simplified model systems, the effects of salts and oil on enzymatic texturisation of protein isolates from soy (Glycine max (L.) Merr.; SPI), pea (Pisum sativum L.; PPI) and sweet lupine (Lupinus albus L.; LPI) were evaluated. In aqueous systems, protein cross-linking by microbial transglutaminase (MTG) was significantly improved when NaCl (1–2 g hg−1) was added, but respective doses of CaCl2 reduced gel strengths. As shown by emulsion model systems of PPI and SPI with oil/protein ratios of 1 and 2 g g−1, emulsification of corn oil into aqueous protein suspensions prior to enzymatic cross-linking enhanced gel formation depending on the emulsification technique. The impact of NaCl and oil varied among the protein isolates as to obtainable maximum gel strengths and optimum doses of these ingredients. The applicability of MTG to leguminous proteins in complex plant foodstuffs was finally deduced from their performance in complex food models of the liquid (thickened soup), foamed (mousse) and solid (sausage-like substitute) type, respectively. The sensory characteristics of the latter were evaluated by trained panellists relative to their milk-, gelatin- and meat-based counterparts. Texture was appealing in the foamed and solid food models, but the liquid soup model suffered from unfavourable grittiness. Without masking their beany off-flavour, the food models containing leguminous proteins deviated from the reference products. On the whole, MTG-catalysed cross-linking rendered the leguminous proteins suitable for the food applications in terms of visual appearance, texture and colour. Especially, the gels representing mousse-type foams and cuttable sausage-like vegetarian substitutes were very promising.