Water binding and water release by plant-based meat analogues

The sustainability of the human diet could be improved by lowering meat consumption and offering consumers plant-based meat substitutes as an alternative. Consumers are thought to prefer meat substitutes with a high similarity to real meat. Therefore, meat analogues that accurately mimic the different textural properties of real meat are being developed. Recent advancements in high-moisture extrusion cooking and shear cell processing have enabled the production of fibrous, meat-like products from plant-based ingredients. This has resulted in a wider research scope, which includes other textural properties such as the juiciness. For meat, the juiciness is associated with the water holding capacity (WHC). We have aimed to better understand the uptake, distribution, and release of water of meat analogues. Since these properties relate to the water holding capacity (WHC), this work could help the development of juicier meat analogues.Meat analogues were simplified to gels consisting of two protein phases, such as soy protein isolate (SPI) and gluten. In Chapter 2 we studied the WHC of two-phase gels with different ratio's between SPI and gluten. The WHC of single-phase SPI gels as a function of applied pressure could be described adequately with Flory-Rehner theory. The WHC of gluten gels was very low and could not be described with Flory-Rehner theory as the WHC did not depend on the applied pressure. Therefore, the WHC of gluten was approximated as a constant when describing the WHC of the two-phase gels. The WHC of SPI gels was found to depend on the polymer content at gelation. This was accounted for by determining the water partitioning before gelation using Flory-Rehner theory. The WHC of SPI in the two-phase gels decreased as gluten content increased. This was explained as the result of a mechanical interaction by which the continuous gluten network lowers the WHC of the entrapped SPI phase. In Chapter 3 the effect of gluten on the WHC was further studied using protein isolates from soy, pea and fababean (SPI, PPI, FPI resp.). Gels made from these protein isolates differed in terms of WHC, with SPI having the highest WHC and FPI the lowest. When combined with gluten, the relative reduction in WHC as a function of gluten content was the same for all three proteins. This suggested that the interaction between gluten and leguminous proteins is universal. A reduction in WHC with increasing gluten content was also observed after thermo-mechanical processing in a shear cell, albeit to a lesser degree. Visual inspection of the structures obtained after thermo-mechanical processing indicated that gluten should contribute ≥ 50% to the total protein content to obtain fibrous structures. Since fibres were also obtained after processing hydrated gluten, we hypothesized that gluten is primarily responsible for fibre formation. The second protein might merely act as a filler and could be easily replaced. The WHC experiments suggested that a continuous gluten network could form already at low gluten contents. Since no fibres were observed at low gluten contents it was suggested that the fraction of gluten should be high enough for them to be observed as fibres.In Chapter 4 we studied whether the WHC of meat analogues could be controlled during post-processing by varying the pH and ionic strength of the marinade, or by altering the cross-link density. Our experiments showed that the WHC can be increased by lowering the ionic strength or increasing the pH of the marinade. Lowering the cross-link density also led to an increase of the WHC. Similar results were obtained from qualitative simulations using a model based on Flory-Rehner and an extension to account for charge effects. The simulations showed that the difference between the iso-electric point (pI) and the internal pH of the meat analogue strongly affects the WHC. At low ionic strengths, the meat analogue's internal pH can deviate from the pH of the marinade due to the buffering effect of proteins and the requirement of electro-neutrality inside and outside the protein network. The results showed that marinade pH and ionic strength offer control over the WHC and that low-salt marinades might improve juiciness. %, which do not have internal cavities, , which do have internal cavities.The release of water during mastication is considered essential to the perception of juiciness. In Chapter 5 we developed a confined compression cell to measure the dynamics of juice release for SPI gels and meat analogues. The release rates for SPI gels were in good agreement with model simulations based on Flory-Rehner theory and Darcy's law. The release rates for meat analogues were greatly underestimated by the model, especially for low applied pressures. Time domain nuclear magnetic resonance (TD-NMR) results indicated the presence of water-filled cavities within the meat analogue. This water was expelled at a lower pressure than the water inside the protein matrix. The water-filled cavities can provide a path of low resistance, which explains the higher measured water release rates. The porous structure of meat analogues, therefore, appears to be important for their water release properties. Control over the porous structure of meat analogues could, therefore, offer an additional tool to control the juiciness.Since the WHC of meat analogues is related to their structure, control over the WHC, and presumably the juiciness, will require a good understanding of their production process. Meat analogues are most commonly produced through thermo-mechanical processing with high moisture extrusion cooking (HMEC) or with a shear cell. In Chapter 6 we reviewed these two processes by describing the physicochemical changes induced during the different processing steps. The processes consist of the same three steps: mixing and hydration, thermo-mechanical treatment, and cooling. Knowledge gaps were identified concerning the effect of thermo-mechanical treatment on protein-protein interactions and the formation of the fibrous structure. Filling these gaps will require more experimentation at process conditions and could improve control of the process and product. The use of food and non-food model systems could support this by allowing for more controlled experimentation.In Chapter 7 we reflect on our findings in a broader context. The different underlying assumptions of Flory-Rehner theory are discussed. Most assumptions are considered plausible for protein-based meat analogues, although the physical meaning of the model parameters might be limited. The use of physical models is expected to contribute to the development of better meat analogues in the future. Some suggestions are made on how to produce a juicy meat analogue under the assumption that the juiciness largely depends on the WHC. Although the relation between the WHC and the sensory juiciness is yet to be confirmed, we propose that juiciness could be considered a structure-function relationship. The juiciness is a complex sensory attribute with several contributing factors. Since we have only focused on the WHC, future research should explore the relative importance of the WHC as well as the other contributing factors, such as flavour, texture and salivating action. Given the interconnectedness of the different factors and the possibility of cross-modal perception, a high level of control over the product properties will be required. The level of control will largely depend on the level to which we understand the relations between ingredient properties, process parameters, and product properties.