Legume protein concentrates

Abstract

A legume protein concentrate having a protein content of about 40-70%, preferably 50-60%. The concentrate exhibits trypsin inhibitor activity (TIA) at up to about 2mg / g and has a water binding capacit

Description

Field of the Invention This invention relates to new legume protein concentrates having reduced off-flavours and low amounts of antinutritional factors. The protein concentrates are useful in food and drink products and may provide a sustainable replacement to animal protein and also can be used as feed material for feed or pet food. The invention also relates to a process for producing these legume protein concentrates. Background Legume protein concentrates have gained significant attention due to their nutritional value and sustainability. Derived from legume sources like peas, chickpeas, lentils, and beans, these concentrates offer several advantages: • Nutritional Richness: Legume protein concentrates provide essential amino acids, making them valuable for human nutrition. • Sustainability: As a plant-based alternative to animal protein, they contribute to reducing the environmental impact. • Functional Properties: These concentrates exhibit emulsifying, gelling, and foaming properties, enhancing their versatility in food formulations. These legumes are typically contained in pods at harvest, and are used as food either as dry or immature seeds. They are in the class Leguminosae. Most grain legumes, as harvested, typically contain protein in at around 18-50% by weight. Other significant components are carbohydrates (e.g. starch), oil and fibre. Many legumes are converted into protein concentrates, which typically have higher protein levels (e.g. around 40 to 70% by weight) and so offer additional utility in food and drink products. Key classes of legume protein include storage protein (e.g. albumins, globulins, prolamins and glutelins) which is the most abundant class in legumes, biologically active protein (e.g. lectins, enzymes and enzyme inhibitors) and structural proteins (e.g. ribosomal, chromosomal and membrane proteins). However, leguminous materials are often undesirable for use in food products because of their undesirable flavours. For instance, pea flour obtained from dry milling of peas contains pea flavours and a bitter taste. Consequently, when pea flour is used in food products, the flavour and bitter taste of raw peas is transferred to these products. Taste is a critical factor in food production, and as a result, protein isolates are often preferred for such uses. Protein isolates have a higher protein content and typically have a cleaner taste profile than protein concentrates, but their manufacture is more resource intensive. The listing or discussion of an apparently prior-published document in this specification should not necessarily be taken as an acknowledgement that the document is part of the state of the art or is common general knowledge. Disclosure of the Invention According to a first aspect of the invention, there is provided a legume protein concentrate having a protein content of from about 40% to about 70% by weight, wherein the concentrate exhibits trypsin inhibitor activity at up to about 2 mg / g, and the concentrate has a water binding capacity of at least about 2.0 g / g. These legume protein concentrates are hereinafter referred to as "protein concentrates of the invention" or "concentrates of the invention". We have found that protein concentrates of the invention have a number of advantageous properties, such as sensory and / or functional attributes, that make them particularly suited for use in food and drink products for human consumption. In particular, the concentrates of the invention may have one or more of an improved flavour profile, lower antinutritional factor content, and an increased water binding capacity, each in comparison to commercially available legume protein concentrates. In addition, these properties can be maintained whilst ensuring that the oil binding capacity of the product can be tuned. The concentrates of the invention are protein concentrates, and these differ from protein isolates in a number of respects. For example, the protein concentrates of the invention contain a lower weight percentage of protein than is commonly found in protein isolates. The protein concentrates of the invention typically contain protein at from about 40% to about 70% by weight of the concentrate, whereas protein isolates which typically contain protein at about 80% or above. Protein isolates are often preferred for use in foods as taste is a critical factor and protein isolates typically have a cleaner taste profile than protein concentrates. However, the protein concentrates of the present invention have been found to have an improved taste profile, as evidenced by tests in human volunteers. This allows protein concentrates to be used in foods instead of protein isolates, and this is advantageous as the protein concentrates of the invention provide other benefits, including their production requires significantly lower amounts of natural resources (particularly water and energy, with a lower need for drying) compared to protein isolates, and protein concentrates have greater functionality. Different methodologies used for manufacture of protein isolates include alkaline extraction with isoelectric precipitation, alkaline extraction with ultrafiltration, salt extraction, or micellization. In one embodiment of the invention, the protein concentrate contains protein at from about 45% to about 65% (e.g. from about 50% to about 65%, or from about 50% to about 60%) by weight. The protein content in a sample may be determined according to any method known to the person skilled in the art. For example, the nitrogen content may be first determined, and then the protein content may be calculated by multiplying the determined nitrogen content by a standard nitrogen-to-protein conversion factor (e.g. by a factor of 6.25, which is based on the assumption that al proteins contain 16% nitrogen by weight - other conversion factors may be selected depending on the material in question). A suitable method for determining nitrogen content is detailed in ISO 1871:2009, which uses the Kjeldahl method. The protein concentrates of the invention are obtained from one or more legumes. A legume is a plant in the family Fabaceae (or Leguminosae), or the fruit or seed of such a plant. When used as a dry grain, the seed is called a pulse. For the avoidance of doubt, such seeds may also be used as a source material for the concentrates of the invention. Legumes that may be mentioned in this respect include alfalfa, clover, beans, peas, chickpeas, lentils, lupins, mesquite, carob, soybeans, peanuts, and tamarind. Particular legumes include fava beans, peas, lentils, and chickpeas. References to "pea" or "peas" include references to the yellow pea and the brown(grey) pea. The concentrate of the present invention is advantageously obtainable from a single source. The use of a single source facilitates efficient and economic processing of the raw legume material. By the phrase "obtainable from a single source" (or "obtained from a single source"), we mean that the protein concentrate is obtainable from one or more organisms of a single taxonomic class. In a particular embodiment, the protein concentrate is not derived from multiple organisms across different taxonomic classes. Instead, the protein concentrate of the invention (or the starting material from which the protein concentrate can be obtained, e.g. by extrusion processing) is obtainable from a single population of organisms, for example, a single legume species. For the avoidance of doubt, the phrase "obtainable from a single source" does not exclude the use of multiple organisms of the same species as a source of the protein concentrate, i.e. it does not exclude the use of multiple plants or plant seeds that are of the same species. Said multiple organisms are preferably all of the same species, or from the same breeding line, or of the same plant variety, or of the same production stock or batch. The protein concentrates of the invention advantageously contain low quantities of antinutritional factors, in particular low quantities of trypsin inhibitor. Antinutritional factors are compounds that interfere with the absorption of nutrients, and they typically act by binding to vitamins and minerals, preventing their uptake, or inhibiting enzymes. A trypsin inhibitor is a naturally occurring protein that acts as a serine protease inhibitor and reduces the biological activity of trypsin by controlling the activation and catalytic reactions of proteins. Trypsin is an enzyme involved in the breakdown of many different proteins, primarily as part of digestion. The presence of trypsin inhibitor has been found to result in delayed growth as well as metabolic and digestive diseases. Additionally, pancreatic hypertrophy is a common occurrence with trypsin inhibitor consumption. The presence of trypsin inhibitor in a product reduces the protein efficiency and therefore results in the consumer's body not being able to efficiently and fully utilise the protein. In particular embodiments of the invention, the protein concentrate has a trypsin inhibitor activity level of up to about 1 mg / g, such as up to about 0.5 mg / g. In the examples, the amount of trypsin inhibitor activity was significantly reduced compared to the raw plant material used to form the protein concentrate as that raw plant material contained trypsin inhibitor activity level of about 2.92 mg / g. In another embodiment, the total amount of antinutritional factors in the protein concentrate is up to about 20 mg / g. Other antinutritional factors that may be mentioned include vicine and convicine, both of which are pyrimidine glycosides. These substances are notably present in fava beans. In spite of the vast nutritional and environmental benefits provided by fava bean cultivation, its use as a food crop has been restricted, primarily due to the presence of vicine and convicine. Ingestion of these substances can cause favism in individuals with a genetically inherited deficiency in glucose-6-phosphate dehydrogenase (G6PD). In monogastric animals, vicine and convicine can cause decreased feeding efficiency. Vicine and convicine are stored in cotyledons of most faba beans at about 1% of dry matter, and are thermostable, but their concentration can be greatly reduced by soaking the seeds in water or in a weak acid solution prior to cooking. Thermal processing such as boiling, roasting, microwave irradiation, and frying can reduce the vicine and convicine content in faba bean seeds. In addition, the combination of enzyme treatment with fermentation or of alkaline extraction with acid precipitation can reduce vicine and convicine content by more than 99% (H. Khazaei, et al. Trends in Food Science &Technology, Vol. 91, 2019, 549-556). However, removal or destruction of vicine and convicine by dry milling for protein concentration on an industrial scale is problematic because air classification of faba bean protein concentrates these substances up to nearly four-fold in the protein fraction. Wet processing methods for protein purification, e.g., isoelectric precipitation, can remove anti-nutritional factors such as these from protein fractions, but these methods are costly and energy-intensive. The best solution for the reduction of vicine and convicine is breeding for low vicine and convicine faba beans, and the discovery of a low-vicine and low-convicine accession with up to 95% reduction in their content compared to wild type has enabled the transfer of the low-vicine and low-convicine trait to faba bean cultivars by sexual crosses. In particular embodiments of the invention, the protein concentrate contains vicine in an amount up to about 10 g / kg, such as up to about 8 g / kg or up to about 7 g / kg. In the examples, the amount of vicine present was surprisingly slightly reduced compared to the raw plant material used to form the protein concentrate as that raw plant material contained vicine in an amount of about 6.5 g / kg. In another embodiment, the protein concentrate contains convicine in an amount up to about 5 g / kg, such as up to about 4.5 g / kg or up to about 4 g / kg. In the examples, the amount of convicine present was slightly reduced compared to the raw plant material used to form the protein concentrate as that raw plant material contained convicine in an amount of about 5.13 g / kg. In yet another embodiment, the protein concentrate contains vicine in an amount less than about 7 mg / kg and / or convicine in an amount less than about 5 mg / kg. The amount of antinutritional factor present in a sample may be determined according to any method known to the skilled person. One suitable method is that disclosed in ISO 14902:2001. In such a method, trypsin inhibitors are extracted from a sample at pH 9.5. The remaining trypsin activity is measured by adding benzoyl-L-arginine-p-nitroanilide (L-BAPA) as substrate. The quantity of released p-nitroaniline is measured spectrometrically. Vicine and convicine levels may be determined using the method disclosed in Example 6, or any other method known to the skilled person. Water binding capacity (WBC) is a further important characteristic of the protein concentrates of the invention. The WBC of a protein is defined as the amount of water absorbed by 1 g of protein. It is a crucial functional property as it affects a product's texture and flavour binding. Water binding occurs through a combination of ion-dipole, dipole-dipole, dipole-induced dipole, and hydrophobic interactions. Several theories have been presented to explain how water interacts with proteins. The WBC can be determined by amino acid composition of a protein. Water is more electrostatically attracted to highly charged proteins which is a consequence of the polar nature of water and the charged state of proteins, leading to the formation of a hydration layer that plays a crucial role in the solubility and structural stability of the proteins in water (A.K. Stone, et al., Food Sci. Biotechnol. 24 (3) (2015) 827-833). The WBC is also correlated with other protein functional properties, such as solubility, gelation, and emulsifying properties. The WBC is also lowest at a protein's isoelectric pH since protein-protein interactions are at their peak. Furthermore, it is reported that the WBC of pea protein concentrates increased with the severity of heat treatment. Without wishing to be bound by theory, it is believed that the exposure of hydrophilic groups may be the reason for the increase in WBC, thus more denatured isolate products have higher WBC. The WBC of proteins is critical for end-product quality in terms of mouthfeel, texture, and flavour retention. A higher WBC is essential to prevent losses in formulations such as meat alternatives during processing, helping to maintain an acceptable texture of end products. For example, a high WBC can help improve the juiciness and texture of plant-based meat products. The ability to retain more water can lead to a more meat-like mouthfeel and better moisture retention during cooking, making the final product more appealing to consumers. In baked goods, i.e. bakery products, a higher WBC can enhance the texture and extend the shelf life by keeping products moist for longer periods. This can be particularly useful in products like bread, muffins, and cakes. In dairy alternatives, such as non-dairy yogurt or cheese alternatives and other products that are not fermented, increased WBC can help achieve a creamy and smooth texture, which is often desired in these products. Maintenance of the pH at higher levels (e.g. at a pH of from 7 to 8) can help to achieve increased WBC. A lower WBC can be advantageous in certain circumstances. For example, in the production of fermented foods like vegan cheeses or yogurt alternatives, lower WBC (which has been found to occur at acidic pH) can help manage the texture, making it firmer and less watery. This is beneficial for achieving the desired consistency in such fermented food products. Lower WBC levels in protein shakes or acidic protein-rich beverages can be helpful for preventing the beverage from becoming too thick or viscous, maintaining a more drinkable consistency. In certain acidic sauces and dressings, reduced WBC can help maintain a smoother and more pourable consistency, which is often preferred by consumers. The ability to manipulate the WBC by adjusting the pH offers other benefits of great significant for food production. In this respect, it allows for greater control over the texture of food products. This can be particularly useful in developing new products or improving existing ones to meet specific textural requirements. In products where the moisture content needs to be controlled to prevent spoilage or ensure stability, the pH-dependent WBC can be a useful tool. For instance, in dry mixes and powdered formulations, maintaining low WBC can help in achieving the desired dryness. Legume protein, such as pea protein, is a high-quality, plant-based protein source. Utilizing it effectively by controlling its WBC can enhance the nutritional profile of various food products without compromising on texture and quality. The concentrates of the invention have a water binding capacity of at least about 2.0 g / g. In a particular embodiment, the water binding capacity is at least about 2.5 g / g, or at least about 3.0 g / g. These values are appropriate when the pH is about 7 (e.g. from 6 to 7.5). Methods of measuring WBC are known to the skilled person, and include for example the methods described elsewhere herein. The protein concentrates of the invention, particularly those obtained using the extrusion methods disclosed herein, have been found to have higher WBC values than protein concentrates obtained from the same legumes by other methods. In comparative studies using fava beans, WBC values increased for all samples in comparison to a reference fava bean protein concentrate that had been obtained using a different method. These increases in WBC ranged from 38 % to 94 %. Thus, in another embodiment of the invention, the protein concentrate has a WBC that is at least 20%, such as at least 30% higher when compared to a protein concentrate obtained using standard dehulling, milling and classification (e.g. air classification) methods. In such protein concentrates, the WBC may be at least 35% higher, at least 40% higher, at least 45% higher or at least 50% higher. The WBC may also be up to 120% higher, up to 110% higher, up to 100% higher or up to 95% higher when compared to a protein concentrate obtained using the above-mentioned standard method. Oil-holding capacity (OHC) of proteins is described as the amount of fat absorbed per gram of protein. The interaction between proteins and lipids occurs between nonpolar side chains of proteins and the aliphatic chains of lipids through hydrophobic interactions. OHC is a key functional property of proteins when they are used as a meat extender, meat replacer for meat-based food applications, and in baked goods to provide acceptable sensory attributes in end products. We have found that the methods disclosed herein facilitate the production of protein concentrates with variable OHC values, but with consistently high WBC values. Protein concentrates with reduced OHC values are advantageous for a number of reasons, primarily relating to texture, mouthfeel and appearance. In particular, such products are useful as dairy alternatives. Products having a high OHC have a tendency to result in a greasy or overly firm texture which may not be desirable in products like soft cheeses or creamy yogurts. Therefore, having a low OHC can help maintain a smoother, more spreadable texture. The mouthfeel of products is crucial for consumer acceptance. Excessive oil binding can lead to a heavy, oily sensation in the mouth, which may detract from the overall enjoyment of the product. By reducing OHC, dairy products can maintain a lighter and more pleasant mouthfeel. Oil binding can also affect the appearance of dairy products. Too much oil can cause separation or an unattractive greasy sheen on the surface. Products with lower OHC are less likely to suffer from these issues, resulting in a more visually appealing product. In plant-based meat applications, controlling OHC is crucial for multiple reasons, primarily relating to texture, mouthfeel, flavour release and cooking performance. Plant-based meats aim to replicate the texture and juiciness of traditional meat products. Managing the OHC helps achieve a texture that mimics animal-based meat by controlling moisture retention and fat distribution. Too much oil binding can result in a dry and crumbly texture, while too little can lead to a soggy or oily mouthfeel. Having the ability to tune the OHC allows the production of a juicy and satisfying bite similar to animal meat. The OHC influences the mouthfeel and flavour release of plantbased meats. Balancing oil absorption helps create a pleasing mouthfeel without an excessive greasy sensation. Additionally, controlling oil binding can enhance the release of flavour compounds, ensuring that the plant-based meat delivers a rich and savoury taste experience comparable to animal-based counterparts. Plant-based meats also need to perform well during cooking processes such as grilling, frying, or baking. Controlling the OHC ensures that the product maintains its shape and integrity, minimizing shrinkage or crumbling during cooking. This contributes to a more visually appealing presentation and a better overall cooking experience for consumers. Similarly, in bakery applications, control of OHC can be important for the texture and moisture content of the product as well as for its shelf life. In baked goods like cakes, muffins, and pastries, managing OHC helps maintain the desired texture and moisture content. Too much oil absorption can result in a dense and greasy texture, while too little can lead to dryness and crumbliness. By having the ability to tune the OHC, bakers can achieve a tender and moist crumb that enhances the overall eating experience. Excessive oil absorption in baked goods can accelerate staling and reduce shelf life by promoting lipid oxidation. Controlling OHC helps extend the freshness and quality of products, allowing them to remain soft and flavourful for a longer period. This is particularly important for packaged baked goods sold in retail stores or consumed over multiple days. In an embodiment of the invention, the concentrate has an oil holding capacity of from about 1.5 g / g to about 3.0 g / g. In another embodiment, the concentrate has an oil holding capacity of from about 1.7 g / g to about 2.8 g / g, from about 1.8 g / g to about 2.7 g / g, from about 1.9 g / g to about 2.6 g / g, or from about 2.0 g / g to about 2.5 g / g. The protein concentrates of the invention have been found to have lower OHC values. Methods of measuring OHC are known to the skilled person, and include for example the methods described elsewhere herein. An important property of the protein concentrates of the invention is the reduced levels of off-flavours. Generally speaking, volatile components that can produce unpleasant sensations when generated by or released or retained from food raw materials and their processing are identified as off-flavours (Viana L., English M. LWT-Food Sci. Technol. 2021; 150:111981). The off-flavour in plant-based meat analogs is commonly described as a beany flavour, which presents characteristics of fat, grass, and earth (Ravi R, et al. Biosensors (Basel). 2019;9(2):66). More than 30 kinds of volatile substances have thus far been related to the beany flavour: predominantly aldehydes, ketones, and alcohols, with hexanal, hexanol, l-octen-3-ol, and 2-pentylfuran having the greatest impact on beany flavour (Yang L, et al. Foods. 2023; 12(5):923). The main reason for the beany flavour in soybean protein, for example, is that the unsaturated fatty acids (linoleic and linolenic acids) present in soybean are easily oxidised to form hydroperoxides under the action of lipoxygenase, which are then further degraded to form aldehydes, ketones, and alcohols, among others. n-Pentanal is a substance that originates from lipid oxidation of linoleic acid and this is also known to contribute to an odour impression often described as "green" or "milky" (W. S. U. Roland, et al., Cereal Chern. 94(1):58-65). Sweetness and bitterness are other flavours that are important to consider for protein concentrates. Bitterness is widely described as a sensory off-note in many plant products, and is often attributed to the release of low molecular weight peptides containing hydrophobic amino acids residues, particularly leucine, proline, phenylalanine, and tyrosine, as well as other substances including saponin, isoflavones, flavonoids, isochlorogenic acid and chlorogenic acid (Ben-Harb S, et al. Foods. 2022; 11(8): 1146). Sweetness was found to be increased and bitterness was found to be decreased in the concentrates of the invention. Other sensory attributes of importance include astringency. Astringency is a puckering, dry, or rough sensation in the mouth which is caused by the interaction of a nonvolatile compound with the salivary proteins and mucins in the mouth, which leads to a loss of lubrication in the mouth. It is therefore an objective of the present inventive to provide protein concentrates with an improved sensory profile, and the protein concentrates of the invention have been found to achieve this when tested in human volunteers. In particular, the protein concentrates have been found to have lower detectable levels of off-flavours compared to a protein concentrate from the same organism and obtained using the standard dehulling, milling and classification (e.g. air classification) methods mentioned hereinabove. It is therefore advantageous to reduce off-flavours as much as possible. The protein concentrates of the present invention showed particular improvements in respect of their bitter, green and beany off-flavours. Thus, in one embodiment of the invention, one or more (e.g. all) of the bitter, green and beany off-flavours of the protein concentrate are reduced compared to a protein concentrate obtained by a process comprising extrusion of a legume protein at a temperature exceeding about 115 °C. When off-flavours are present in conventional legume concentrates, this may be countered by the use of flavour masking (including using taste-masking agents), blending with other ingredients (such as sweeteners, natural fruit extract or other plant-based proteins such as rice protein or hemp protein), microencapsulation (e.g. in a liposome or microcapsule), aroma masking (such as by adding specific ingredients such as essential oils, spices, or natural aromas, which have a strong, pleasant smell), fermentation (e.g. using microorganisms break down the proteins releasing aromas and flavours, which can help mask the unpleasant taste and odour of the legume protein), and deodorisation (a process typically used to remove saponins to reduce the beany odour in some products). However, the reduction in off-flavours in the concentrates of the invention reduces the need to counter off-flavours in this way. Thus, in one embodiment of the invention, the protein concentrate contains a reduced amount of taste-masking agents, or taste-masking agents are essentially absent. Reduction of off-flavours is also beneficial for allowing the protein concentrates to be used in a broader range of recipes. Methods for measuring off-flavour characteristics are well known in the art and include organoleptic assessments, such as those described in Yanting Shen, et al., Food Chemistry 385 (2022) 132687. In tests using such methods with human volunteers tasting vegan ice cream prepared using fava bean protein concentrates according to the invention, it was found that the concentrates of the invention provide a significant reduction of typical off-pea flavours. A similar result is expected to be observed for other legumes mentioned elsewhere herein, such as peas, lentils and other beans. For the avoidance of doubt, protein concentrates of the present invention may have one or more of the properties described herein. Preferences and options for a given aspect, feature or parameter of the invention should, unless the context indicates otherwise, be regarded as having been disclosed in combination with any and all preferences and options for all other aspects, features and parameters of the invention. For example, the protein concentrates of the present invention (e.g. protein concentrates obtained from peas or beans) may have an enhanced WBC (a water binding capacity of at least about 2.5 g / g), reduced levels of antinutritional factors (e.g. trypsin inhibitor activity of up to about 1 mg / g), and reduced off-flavours (e.g. lower levels of bitter, green and beany off-flavours). The protein concentrates of the invention may be manufactured using the methods described herein, particularly in the examples. We have found that the protein concentrates of the invention may be readily manufactured through extrusion of leguminous material, e.g. using a conventional industrial extruder apparatus, e.g. a high-shear extruder. Extrusion of vegetable proteins is a process in which moistened, expansile, starchy and / or proteinaceous materials are plasticised in a tube by a combination of moisture, pressure, heat, and mechanical shear. Extrusion, also often referred to as "extrusion cooking", can result in elevated product temperature within the tube along with gelatinisation of starchy components, denaturation of proteins, stretching or restructuring of tactile components, and the exothermic expansion of the extrudate. A typical extrusion system consists of a mixer (mixing raw materials), bin feeder (transferring uniformly raw material into preconditioner), preconditioner (preconditioning of raw material temperature and moisture through the injection of water or steam), extrusion barrel configuration, die, and knife. Extrusion cooking is a complex process due to its hydration, forming, cooking, unfolding, and alignment functions, in addition to providing the texturization of vegetable proteins through restructuring. During this process, proteins are extensively denatured in the presence of water and thermal energy. The extrusion barrel configuration is composed of various sections in which proteinaceous materials are structured. Briefly, the first barrel acts as a feeding zone, where the preconditioned protein material is transferred from the preconditioner into the extruder. After the first barrel, the protein material moves into the processing zone, where the amorphous protein material is converted into a colloidal dough with the addition of water and compression. The temperature of the highly moist proteinaceous dough begins to increase in a short time (2-5 sec) due to mechanical energy generated by a rotating screw. The moist dough is forced to flow along barrels through the screw(s) under high pressure and shear, which leads to vaporisation of water and transformation of the dough into a viscoelastic, plasticised mass in the extruder barrels. Finally, the viscoelastic mass is pushed through the die openings. The mass moisture partially evaporates and rapidly expands to generate aerated, cellular structure in the finished product at the die when it is exposed to ambient pressures upon exiting the die. The leguminous material that is used in the process of the invention may be supplied in any suitable form, such as in the form of a powder or granulate, including crushed granules. Such products are obtained from harvested material via conventional methods, such as dehulling, milling to form a fine powder, classifying (e.g. air classifying) to produce starch and protein concentrates. The product used in the extrusion step may be in the form of a fine powder of protein concentrate or pelletized and crushed protein concentrate. The use of a low barrel temperature during extrusion has been found to produce protein concentrates with the characteristics described herein, e.g. protein concentrates with an advantageous combination of high WBC, low antinutritional factor content and / or reduced off-flavours. In particular, these products were obtained when the extrusion processing temperature (e.g. the temperature of the leguminous material in the barrel) did not exceed about 115 °C. Thus, according to a second aspect of the invention, there is provided a process for obtaining a legume protein concentrate (e.g. as hereinbefore defined), which process comprises extrusion of a legume protein at a temperature not exceeding about 115 °C. Such processes are referred to herein as "processes of the invention". In an embodiment of the processes of the second aspect of the invention, the temperature of the legume protein during the extrusion process does not exceed about 110 °C. For example, it may not exceed about 100 °C or may not exceed about 90 °C. Nevertheless, a sufficiently high temperature is required in order for the protein content to undergo the relevant transformations, such as protein restructuring and texturization as is described earlier. Particular processes of the invention that may be mentioned therefore include those in which the temperature of the legume protein reaches at least about 80 °C during the extrusion process (e.g. it reaches this temperature at one or more regions in the extrusion barrel). In one embodiment, the temperature of the legume protein reaches at least about 85 °C during the extrusion process. Temperature control can be achieved through various methods, such by controlling the extruder screw speed or by adjusting the rate of water addition. The extrusion process typically involves the addition of both water and leguminous starting material to the extruder. It was found that adding different amounts of water allowed the bulk density of the final product (the extrudate) to be altered: increasing the amount of water added produced a product with a higher density and less damage. "Damage" in this context may refer to popped starch, and / or macromolecular structures in starch, protein and fibers. Adjusting the proportions of water and leguminous starting material so that the process involved the addition of water in an amount of from 0.05 to 0.2 L per kg of legume powder / granulate was found to give protein concentrates with the advantageous properties described hereinabove. The protein concentrates were also found to have a moisture content of between 15 and 30%. It was found that the addition of a relatively small amount of water, e.g. an amount of from 0.05 to 0.2 L per kg, was sufficient to achieve the reduction in off-flavours noted elsewhere herein. The addition of a relatively small amount of water is advantageous because it reduces the requirement to dry the product after extrusion. The extruder used in the examples was a high-shear, single-screw extruder. Nevertheless, the skilled person would be aware that other extrusion apparatus (e.g. twin-screw extruders) could be used to product protein concentrates of the invention. Suitable process parameters, such as feed rates, screw rates, barrel dimensions, and the like, can be chosen by the skilled person as appropriate for the apparatus in use to obtain a protein concentrate having the properties described herein. Thus, in a further embodiment, the protein concentrate of the invention is obtainable from (e.g. obtained by) a process according to the second aspect of the invention. For example, the protein concentrate may be obtained by a process comprising extrusion of a legume protein and the temperature of the material does not exceed about 115 °C during extrusion. In one embodiment, the temperature does not exceed about 110 °C, about 100 °C or about 90 °C during extrusion. Additionally, the protein concentrate may be obtained by an extrusion process of the invention that includes the addition of water in an amount of from 0.05 to 0.2 L / kg. The protein concentrates obtained from that processes were found to contain trypsin inhibitor at levels far below that present in the starting material. For example, the process was found to reduce the trypsin inhibitor activity level by at least 80%. Other benefits arising from the process of the invention include the enhanced WBC, the reduction in the off-flavours in the final product, the reduction of vicine levels and the reduction of convicine levels. For example, vicine and convicine levels were each found to be reduced by around 20%. This result is surprising and contrary to that reported in the literature, for example in Tuccillo F. et al.r Food Research International 162 (2022)112036. Whilst the protein concentrate of the invention may be obtainable from (e.g. obtained by) a process according to the second aspect of the invention, the protein concentrate may alternatively (or additionally) be described as one that is not enzymatically or chemically modified. For example, the protein concentrate of the invention may be obtained by a process according to the second aspect of the invention, and at no point during that process was the legume protein subjected to any chemical processing (besides any hydrolysis or other changes that may be caused by reaction with water) or any enzymatic processing. The protein concentrates of the invention may also be advantageously obtained by a method that does not involve alkalization or acid precipitation. Such techniques are commonly used for the formation of protein isolates (which have higher protein contents) because the flavour profile is considered to be good, however the protein concentrates of the invention have been found to have a good flavour profile without the need for such methods. The protein concentrates of the invention are useful in food and drink products, particularly products for human or animal consumption. Such food and drink products are referred to herein as a "food compositions". Protein concentrates of the invention have particular utility as a sustainable food source, and can be used in food compositions as a replacement to animal protein, i.e. they may be meat or dairy alternatives and so may be suitable for vegans. The protein concentrates of the invention are gluten-free and so may be useful as replacements for gluten-containing protein concentrates. Still further, the protein concentrates of the invention may be useful as material for animal feed, including pet food, all of which are included within the term "food compositions". Such food compositions may contain the protein concentrates of the invention in an amount of from 1% to 99% by weight of the food composition. Thus, one aspect of the invention relates to a food composition containing a legume protein concentrate of the invention. These food compositions may contain other ingredients such as a fat, an oil, a carbohydrate, a fibre source, an additional protein source, an emulsifier, a salt, a flavouring, a colourant, and / or a preservative. Examples of food compositions that may contain a protein concentrate of the invention include flour, ice cream, yogurt, cheese, plant-based meat, cake, pastry, muffin, protein drink or plant-based milk drink. The concentrates of the invention are also suited for foods and drinks that are intended to have a high protein content, such as sports nutrition products (e.g. protein bars, powders and ready-to-drink beverages), and protein-fortified cereals and snacks (e.g. breakfast cereals, crackers, chips and extruded snacks). As is described elsewhere herein, a primary characteristic of the concentrates of the invention is the low amount of off-flavours that are present. Nevertheless, the concentrates have other properties that make them particularly suited for use in food compositions, including good texture, mouthfeel, appearance, flavour release, cooking performance, moisture content and shelf life. The concentrates of the invention may have the advantage that they provide a sustainable alternative food source with a good flavour profile. The concentrates may also have properties that make them particularly suited for use in food, such as a low content of antinutritional factors, good water binding capacity, and tunable oil holding capacity. The processes described herein may have the advantage that they may be more efficient than, use fewer resources than, be less energy intensive than, be more readily scaled than, provide more consistent products than, and / or have other useful characteristics over, processes known in the prior art for making protein concentrates. Wherever the word "about" is employed herein in the context of dimensions (e.g. values, temperatures, relative humidities, sizes, weights and timeframes etc.), amounts (e.g. relative amounts (e.g. numbers or percentages) of individual constituents in a composition or a component of a composition and absolute amounts), deviations (from constants, degrees of degradation, etc.) it will be appreciated that such variables are approximate and as such may vary by ± 10%, for example ± 5% and preferably ± 2% (e.g. ± 1%) from the numbers specified herein. The invention is illustrated by the following examples in which: Figures 1 to 5 show the viscosity-temperature measurements for Samples 1 to 5, respectively; Figure 6 shows the viscosity-temperature measurements for a standard fava bean reference sample; Figure 7 shows sensory data for two test samples and a standard fava bean reference sample; and Figure 8 shows sensory data for ice creams made using two test samples and a standard fava bean reference sample. Examples Example 1 - Manufacture of Protein Concentrates Raw materials Fava beans (Tiffany variety) were sourced from a farm in the EU, and held in a storage silo (with temperature measurement) for approximately one year following harvest. Apparatus An INSTA-PRO Model 600 JR Dry Extruder with Top Feeder was used. Method The harvested fava beans were processed to form a granulate by dehulling the beans, milling the beans to form a fine powder, and processing using air classification to produce a protein concentrate. The fine powder was conveyed to conditioner, where the powder was mixed with water to form a homogeneous dough. The moisture content was increase from 9-10 % to approximately 15 %, following which it was passed through a pelletization die where temperature was increased from 50 to 65 °C. The resulting granules were cooled to 15 °C and stored for processing. The granulate was milled through a 6 mm sieve before being fed into the extruder at a rate of around 300 kg / hour. Water was added to the extruder at a rate ranging from 15 to 60 L / hour. The temperature of the extruder barrel was held at a set value in the range 90-110°C. Five different samples were obtained using the following conditions: Sample no. Highest temperature in the Barrel Screw speed Moisture content 1 110-115 °C 20 Low moisture 2 110 °C 20 Increased moisture 3 100 °C 18 Increased moisture 4 100 °C 18 Reduced moisture 5 98 °C 18 Reduced moisture Temperature control was achieved by regulating screw speed and the addition of water. Results It was observed that the product's bulk density changes (rapid volume expansion) could be controlled through water addition. When water addition was higher, the extrusion product had a higher density and less damage compared to samples with smaller water addition. The samples were analysed for water activity and moisture immediately after extrusion. Sample no Water activity Moisture content (%) 1 0.8465 19.78 2 0.8573 23.33 3 0.8988 25.07 4 0.8042 17.43 5 0.9086 26.6 In order to have comparable results in further characterisations, the samples were standardised to contain 15 % moisture by drying at 40 °C until a 15% humidity, and then ground to a homogeneous particle size distribution. Samples 1 to 5 were characterised and compared to each other and against reference samples (see Examples 2 to 8). Samples 1 to 5 differed from the references in almost all chemical physical characteristics analysed. Products (ice cream) prepared with the experimental samples also differed from products prepared with reference samples. For all samples, characteristic off-flavours were drastically reduced. Example 2 - Water Binding Capacity The water binding capacity of various test samples was examined using the following method. Materials The test materials included the five samples in Example 1 and a standard fava bean concentrate. The standard fava bean concentrate was obtained by dehulling harvested fava beans, milling the resulting beans, and then air classifying the powder to form a protein concentrate and a starch concentrate. The WBC of both Sample 5 and the standard fava bean concentrate were measured at different pH values, with either HCI or KOH used to adjust the pH of the protein solutions. Method 1. lg of legume protein was weighed and transferred to a cuvette. 2. 9ml of distilled water was added to the cuvette containing the protein. 3. The mixture was stirred thoroughly using the small pipette and the cuvette was shaken vigorously. 4. The mixture was centrifuged at 3000 rpm for 30 minutes. 5. The sample was left to rest for 30 minutes. 6. The supernatant was removed and the pellet containing the sediment was weighed. For the measurements at different pH, either HCI or KOH added in step 2 to decrease or increase, respectively, the pH of the protein solutions. WBC measurements were made at pH values of 5, 6.4 and 8. Water binding capacity was calculated as follows: [Water binding capacity] = [weight of sediment] I [weight of legume protein powder] This method is described in Delaney Madeline Webb, by B., &Professor Sajid Alavi, M. (2021). Physicochemical properties of pea proteins, texturization using extrusion, and application in plant-based meats. Results Sample Water binding capacity (g / g) 1 3.59 2 2.63 3 3.42 4 3.68 5 3.58 Standard fava bean concentrate 1.9 WBC increased in all samples compared to reference sample, the increase was ranging from 38.47 % (sample 2) to 93.68 % (sample 4). PH Water binding capacity (g / g) Extruded sample 5 Standard fava bean concentrate 8 4.37 1.75 6.4 3.72 1.98 5 3.09 3.25 For the extruded sample, increasing the pH increased the WBC. In contrast, for the native fava bean protein, increasing the pH decreased the WBC. Example 3 - Oil Holding Capacity The oil holding capacity of various test samples was examined using the following method. Materials The test materials included the five samples in Example 1 and the standard fava bean concentrate of Example 2. Method 1. lg of legume protein was weighed and transferred to a 10 mL cuvette. 2. 9ml of rapeseed oil was added to the cuvette containing the protein. 3. The mixture was stirred thoroughly using the small pipette and the cuvette was shaken vigorously. 4. The mixture was centrifuged at 3000 rpm for 30 minutes. 5. The sample was left to rest for 30 minutes. 6. The supernatant was removed and the pellet containing the sediment was weighed. Oil holding capacity was calculated as follows: [Oil holding capacity] = [weight of sediment] / [weight of legume protein powder] This method is described in Delaney Madeline Webb, by B., &Professor Sajid Alavi, M. (2021). Physicochemical properties of pea proteins, texturization using extrusion, and application in plant-based meats. Results Sample Oil holding capacity (g / g) 1 2.07 2 1.80 3 2.66 4 1.97 5 1.96 Standard fava bean concentrate 2.13 The majority of the samples showed significant loss in OBC compared to the reference sample. Sample 1 had a loss of 2.81 %, samples 4 and 5 had losses of 7.5 % and of 7.9 %, and the biggest loss was recorded in sample 2 with 15.49 %. Sample 3 showed an increase in OBC of 24.88 %. Example 4 - Viscosity Analysis Viscosity analyses were done for all samples. A 5% solution of standard fava protein concentrate was used as reference. Method 5 % solutions in water were prepared and kept at room temperature at least 30 minutes before measurement. Measurement was done on Viscosity analyser RM100 Plus (Lamy Rheology instruments), with 3 points heating until 95 °C, holding at 95 °C and cooling to 24 °C, while continuously measuring viscosity in mPa.s. Results Results for Samples 1 to 5 are shown in Figures 1 to 5, respectively. Results for the standard fava bean reference sample are shown in Figure 6. In the reference sample the starting viscosity was at least 4 times higher than the samples that have gone through extrusion. This indicates better solubility for extruded samples, which was also noticed while doing application tests. In the reference sample protein denaturation and starch swelling begins over 75 °C which rests in an increased viscosity. These results indicate that proteins have been partly or fully denatured and the starch has lost its crystallinity. In general, there is a significant difference between the reference sample and the extruded samples, with the extruded samples having improved dispersibility compared to the reference sample. Example 5 - Microscopy Analysis Samples 1 and 5 from Example 1 were analysed by dispersing the sample in oil, marking it with iodine, and observing it under a light microscope. There are noticeable differences between the samples. Sample 1 had a more bonded (blue) amylose to protein agglomerates (yellow), whereas in sample 5 such a difference was not so noticeable. This may be due to the processing parameters: sample 1 had a higher extrusion temperature than sample 5. More disulfide bonds are created between starch and proteins in higher temperatures. Example 6 - Antinutritional Factors The trypsin inhibitor activity, and levels of vicine and convicine present in a protein concentrate both before and after extrusion were assessed. The test material was Sample 5 from Example 1 (both before and after extrusion). Method Trypsin inhibitor activity was measured using test method NEN-EN-ISO 14902:2001. Vicine and convicine levels were determined using HPLC. In this method, 4.5 ml of Milli-Q water was used to extract vicine and convicine from 0.1 g-samples with added 1.6 mg uridine (Sigma Aldrich, St. Louis, MO, USA) used as the internal standard. After vortexing the mixtures and letting them stand for 15 min, they were centrifuged (10 min, 9,600 g) and the supernatants were collected and boiled for 5 min. The supernatants were analyzed by HPLC with PDA (at 273 nm) and the results were calculated according to the method described in Pulkkinen M., et al. Food Research International, 76(1), 2015, pl68-177. Results Antinutritional Factor Before extrusion After extrusion Trypsin inhibitor activity 2.92 mg / g <0.5 mg / g* Vicine 7920 mg / kg 6260 mg / kg Convicine 5130 mg / kg 4070 mg / kg *The detection limit is 0.5 mg / g Example 7 - Sensory Evaluation A sensory assessment of the of extruded protein samples with a panel of 5 human volunteers (2 males and 3 females) was performed. The organoleptic assessment was conducted based on a method mentioned in an article by Shen et al. 2022. A 4% solution was prepared with each sample, and the panelists were asked to rate the strength of different flavour characteristics on a scale ranging from 1 (no discernible attributes) to 10 (very pronounced presence). The results are shown in Figure 7. The results showed that the extruded samples had significantly less off-flavours, which is beneficial for developing new recipes without the off-note of pea and without the addition of masking agents. Among the extruded protein samples, sample 5 had less beany, green, astringent, and bitter notes compared to sample 4. From a sensory perspective, sample 5 was considered to have the least amount of off-flavours. Example 8 - Food Product Extruded proteins were tested in a vegan ice cream application. Products We prepared 3 different vegan ice creams according to the following recipes: Recipe 1 - a standard ice cream recipe based on a mix of fava bean and brown pea protein concentrates was used as a reference. This recipe contained Alomix IC, which is an ice cream solution mix that itself contains the pea / bean protein concentrate along with a taste masking flavour. Recipe 2 - same as Recipe 1 except that protein sample 4 from Example 1 was used instead of the fava bean and brown pea protein concentrates. The masking flavour present in the reference sample was also excluded. Recipe 3 - same as Recipe 2 except that protein sample 5 from Example 1 was used instead of protein sample 4. Ingredients Recipe 1 Recipe 2 Recipe 3 Alomix IC 10% 10%* 10% ** Sugar 18% 18% 18% Water 60% 60% 60% Coconut fat 12% 12% 12% *Alomix IC with substitution of native fava bean / pea protein with extruded protein sample 4, no masking flavor ** Alomix IC with substitution of native fava bean / pea protein with extruded protein sample 5, no masking flavor. Preparation 1. Combine Alomix IC with water at a temperature of 50°C and raise the temperature to 60°C. 2. Let the sample rest for 30 minutes at this temperature. 3. Introduce coconut fat and sugar into the mixture at 60°C and continue heating until the temperature reaches 90°C. 4. Afterward, cool the mixture down to 20°C. 5. Homogenize the mix for 2 minutes for 1200 RPM followed by 1500 RPM for 2 minutes. 6. Chill the mixture in ice cream machine until it reaches -10°C. 7. Store it in a freezer at -18°C. Analysis Sensory evaluation was performed using a method analogous to Example 7. Results The results (Figure 8) showed that Recipes 2 and 3 did not have any typical pea off-flavours despite the absence of a masking flavour. Recipe 2 had a more grainy texture compared to the ice creams of Recipes 1 and 3.

Claims

1. A legume protein concentrate having a protein content of about 40% to about 70% by weight, wherein the concentrate exhibits trypsin inhibitor activity at up to about 2 mg / g, and the concentrate has a water binding capacity of at least about 2.0 g / g-2. The protein concentrate according to Claim 1, wherein the water binding capacity is at least about 2.5 g / g.

3. The protein concentrate according to Claim 1 or Claim 2, wherein the concentrate exhibits trypsin inhibitor activity at up to about 1 mg / g, such as up to about 0.5 mg / g.

4. The protein concentrate according to any one of the preceding claims, wherein total amount of antinutritional factors in the concentrate is up to about 20 mg / g.

5. The protein concentrate according to any one of the preceding claims, wherein the protein content is 50% to 60% by weight.

6. The protein concentrate according to any one of the preceding claims, wherein the protein concentrate further comprises vicine in an amount less than about 7 mg / kg and / or convicine in an amount less than about 5 mg / kg.

7. The protein concentrate according to any one of the preceding claims, wherein the protein concentrate is obtained from fava beans, peas, lentils, or chickpeas.

8. The protein concentrate according to any one of the preceding claims, wherein the protein concentrate is not chemically or enzymatically modified.

9. The protein concentrate according to any one of the preceding claims, wherein the protein concentrate is obtained by a process comprising extrusion of a legume protein and the temperature of the material does not exceed about 115 °C during extrusion, optionally wherein said temperature does not exceed about 110 °C, about 100 °C or about 90 °C during extrusion.

10. The protein concentrate according to Claim 9, wherein the extrusion process includes the addition of water in an amount of from 0.05 to 0.2 L / kg.

11. The protein concentrate according to Claim 9 or Claim 10, wherein one or more of the bitter, green and beany off-flavours of the protein concentrate are reduced compared to a protein concentrate obtained by a process comprising extrusion of a5 legume protein at a temperature exceeding about 115 °C.

12. A food composition containing a legume protein concentrate as defined in any one of Claims 1 to 11, optionally wherein the food is a flour, ice cream, yogurt, cheese, plant-based meat, cake, pastry, muffin, protein drink or plant-based milk drink.1013. A process for obtaining a legume protein concentrate as defined in any one of Claims 1 to 11, said process comprising extrusion of a legume protein at a temperature not exceeding about 115 °C.15 14. The process according to Claim 13, wherein the maximum temperature duringthe extrusion is about 110 °C, about 100 °C or about 90 °C.

15. The process according to Claim 13 or Claim 14, wherein the extrusion process includes the addition of water in an amount of from 0.05 to 0.2 L / kg.27

Images