Development of Soy Protein Meat Analogues Using High Moisture Extrusion Cooking

Global food shortage, aggravated by a steadily increasing world population and diminishing cultivable land, as well as evolving consumer preferences favouring vegetarian/vegan diets have prompted an escalating surge in meat-reduced or meat-less products products. To satisfy this demand, the focus of food manufacturers has been redirected from traditional meat production towards the exploration and utilisation of plant-based proteins. Due to their exceptional nutritional profile and versatile functional properties, soybean proteins in form of isolates (SPI) and concentrates (SPC) have emerged as frontrunning ingredient for meat analogues, surpassing other plant-based protein sources such as pea or lupin. Particularly the soy proteins’ high protein solubility, water holding capacity (WHC), and gelling ability account for superior texturization capabilities, aiding in approximating the fibrous, multilayered characteristics of meat. However, the proteins’ physio-chemical properties were found to vary depending on cultivation and/or extraction conditions. Whilst this topic is exhaustively investigated, the impact of the samples’ distinct characteristics on their texturization capabilities upon high moisture extrusion cooking (HMEC) has not yet been fully explored. The first part of this Ph.D. project explored the functional diversity among six commercially available soy protein powders (one SPC and five SPIs) from different manufacturers, countries and/or batches. This served as foundation for the primary aim of the thesis, i.e., the investigation into how the samples’ unique properties translated into their viscosity pattern upon thermo-mechanical treatment, and further, their ability to be texturized. Moreover, from an application point of view, it was analysed whether (and how) distinct post-processing conditions changed the extrudates’ properties (e.g., water retention capacity). Similarly, it was investigated whether these changes were governed by the inherent physio-chemical properties of the protein powders. Building upon that, the project explored the potential of crosslinker genipin (GNP) in improving the samples’ ability to develop a cohesive, yet fibrous product reminiscent of meat. The first experimental chapter found significant variations in solubility, protein dispersibility index (PDI), WHC, zeta potential (various pH-levels), and particle size distribution. Protein solubility and PDI values exhibited strong correlations, whereas WHC was notably influenced by particle size distribution. Those soy proteins with high solubility (> 57 %, pH 7.0) demonstrated the greatest increase in diameter upon hydration (factor 2.9 – 4.6) and exhibited the highest WHC at pH 7.0 (3.45 – 4.40 g protein/mL water), holding promise in their conversion into meat-like fibres. Dispersibility test is frequently used for screening plant proteins, however it does not provide the functional properties of protein powders when subjected to thermomechanical processing. Texturization is commonly achieved by means of high moisture extrusion cooking (HMEC), which exposes the material to high pressure, shear, and temperatures exceeding 100°C. In the second experimental chapter, these conditions were approximated by rheometer and Rapid Visco Analyser (RVA), which examined the samples’ visco-elastic behaviour upon heating to 95 °C, holding, and cooling down to 25 °C. Both instruments revealed strong correlation between solubility, PDI, and final viscosity, underscoring the importance of high solubility for effective protein-protein interaction during thermal treatment. Variations in viscosity patterns were observed between RVA and rheometer due to different recordings of protein aggregation. The addition of SPC/fibre increased the final viscosity of soy protein blends, with the least soluble SPI being impacted the most (factor 5.2), and vice versae. Hence, adjusting the percentage of added fibre should consider the inherent quality of protein powders. Besides pH value and cooking temperature, the post-processing resilience of extrudates also largely hinges on the material’s inherent properties. The third experimental chapter investigated the retainment of texture of two distinct extrudates, with SPI A outperforming SPI B in revealing a meat-like texture due to good functional properties. Whilst the pH value did not impact the samples’ spatial/dimensional (?) extension, it caused varying effects on water absorption and firmness. When heated at pH 7.0, both samples absorbed water (by 3.32 ± 1.84 %, SPI A and 11.56 ± 1.88 %, SPI B), resulting in decreased hardness, which would be perceived less favourable. Similarly, exposure to pH 4.0 and high temperature (95 °C) resulted in reduced hardness for SPI B, however, SPI A experienced water exclusion and network strengthening, making it suitable for low-acidic dishes like Bolognese sauce. Further elaborating on the divergences in the samples’ quality, the fourth experimental chapter aimed to elucidate how the cross-linking of GNP influenced the proteins’ viscoelastic behaviour, thereby building upon chapter 2. Samples with favourable functional properties either exhibited further increases in viscosity and gel hardness or were not/only slightly impacted. This was possibly owed to disparities in protein folding affecting their accessibility to GNP. Conversely, SPIs with inferior physicochemical properties benefited from the natural crosslinker; so that 0.10 % GNP brought their viscosity and gel texture to the same level as the other SPIs without the chemical. SPC remained unaffected by GNP addition, likely due to the presence of polysaccharides forming a protective barrier. In terms of extrusion processing, the fifth experimental chapter revealed the impact of 0.10 % GNP to be most pronounced for samples with inherently good protein quality, with their hardness significantly increasing by +164 to +162 %, while solubility declined by -10.3 % to -11.6 %. However, their overly cohesive, resilient structure raised doubts about GNP's efficacy for these extrudates. Conversely, GNP aided the other two SPIs in steadily enhancing their initially low texturization, transforming them from a homogeneous, loosely connected protein mass to a fibrous, multilayered extrudate with minimal impact on hardness, gumminess, and colour. Hence, GNP enhances texturization in low-quality samples, while its effect on powders with inherently good functionalities is neglectable. This research demonstrated the strong interrelation of the physio-chemical properties solubility, PDI value and WHC, which were found to differ significantly among the five protein samples. Further, superiority in these inherent quality parameters directly translated into promising texturization capabilities as well as high structural integrity upon post-processing steps. These findings provide valuable insights into the choice of ingredients and formulation of extrusion pre-mixes as well as the processing conditions (95 °C, pH 4.0) required to include extrudates into ready-to-eat dishes. The fact that 0.10 % GNP led to significant texture improvement in low-quality samples set the fundament for future investigations into its application as crosslinking agent in food formulations involving high temperature treatment.