Pea protein–sugar beet pectin binders can provide cohesiveness in burger type meat analogues

Methylcellulose is commonly used in meat analogues for binding ingredients. In this study, we compared the binding properties of a methylcellulose hydrogel (5% w/w) to a novel, clean-label binder based on a mixture of pea protein and sugar beet pectin (r = 2:1, 22.5% w/w, pH 6.0) with and without laccase addition in a burger type meat analogue. It was shown that the pea protein–pectin binder glued vegetable protein particles and fat mimic particles together prior to cooking and frying, thereby improving forming of the mass into burger patties. Furthermore, sensory analysis revealed that the cohesiveness of the fried burger patties was better when the protein–pectin binder was used. However, the used binder system did not affect the hardness of the burger patties indicating that the binders rather affected the coherence of the structural elements. Burgers with solid fat particles were rated better in terms of appearance as compared to emulsified fat particles, since the former were not visible. This study is useful to better understand meat analogue product design for a higher acceptance among consumers.


Introduction
Burger patties are made from minced beef, and salt and spices can additionally be added [1]. Mincing breaks down large meat pieces into smaller meat particles and leads to disintegration of the muscle membrane [2]. This cell destruction causes leaking of intercellular proteins that are tacky and stick single meat particles together. The proteinaceous exudate acts as a food glue and facilitates coherence of the mass to enable it to be formed into shapes [3,4]. Furthermore, the leaked proteins can gel upon thermal treatment, thus strengthening the overall coherence [5].
Since burger patties are a popular product and in light of meat consumption having been shown to contribute significantly to greenhouse gas emissions [6], it is not surprising that plant-based burger patties such as those from Impossible Foods and Beyond Meat were among the first commercially successful products in the growing category of meat analogues [7]. A typical bottom-up approach for plant-based burger patties is to combine key structured and functional ingredients to assemble them into a final product [7,8]. Structured ingredients are, for example, textured vegetable proteins for a fibrous sensation, fat (mimic) particles for a juicy and elastic bite. Functional ingredients include colorants for better visual appearance, as well as spices and aroma for taste tuning [9,10]. The structured components, namely, textured vegetable protein and fat pieces, do not necessarily sticky together on their own [11]. Because of this, vegan food glues are required to facilitate binding of these ingredients [12]. In plant-based burger patties, methylcellulose is frequently used to achieve this. Aqueous dispersions of methylcellulose are sticky, and the hydrophobic character of the compound facilitates a reversible gelling upon heating [13,14], thus mimicking the heat-induced transition of the protein binder in beef burger patties. However, methylcellulose is a synthetic food additive derived from cellulose and, therefore, suffers from a low consumer acceptance [10,15]. This currently fuels the search for a clean-label binder alternative.
Our earlier research showed that pea protein-apple pectin mixtures have a viscoelastic character and their stickiness 1 3 is governed by a balanced combination of adhesion and cohesion that warrants its use as food glue [12]. As such, a concentrated mixture thereof could be used to adhere the aforementioned structured ingredients in a plant-based burger patty. High stickiness was also shown for a mixture composed of pea protein and sugar beet pectin [16]. Furthermore, it was found that this mixture transitions from a semisolid and sticky material into a non-sticky solid upon use of laccase inducing crosslinks between tyrosine in pea protein and ferulic acid in sugar beet pectin [16]. This resembles the above-mentioned transition of methylcellulose in vegan burger patties and the proteinaceous binder in beef burger patties upon heating [5,13,14]. The binding mechanism of a pea protein-sugar beet pectin mixture was already shown to be superior to methylcellulose in a bacon type meat analogue [17].
Therefore, the aim of this study was to test the binding performance of a clean label pea protein-sugar beet pectin mixture (with and without laccase) in a vegan burger patty matrix and to compare it to that of methylcellulose as a benchmark. Minced textured vegetable protein was chosen as base material. Furthermore, two different fat systems were tested. One was composed of solid fat as this is often used for vegan burger patties [10] and the other consisted of an emulsified fat gel that mimics animal fat tissue [9]. We hypothesized that the pea protein-pectin binder improves forming of the uncooked mass into a burger patty due to the binder providing for a high stickiness at similar cohesiveness in fried burger patties as methylcellulose. Furthermore, we expected that laccase-induced solidification of the protein-pectin binder increases hardness in the burger patties.

Materials
High-moisture textured vegetable protein (Halsanskok Vegan Filet Piece) was provided by Nestlé (Singen, Germany) and stored at − 18 °C until further use. Soy protein isolate (Supro EX 37 HG IP) was from Solae Europe S.A. (Geneva, Switzerland) and had a protein content of 90.0% (dry basis). Rapeseed oil (61% unsaturated fatty acids, 24% polyunsaturated fatty acids, 7% saturated fatty acids) was purchased from MEGA eG (Stuttgart, Germany). Solid fat (ILLEXAO, Nestlé, Lausanne, Switzerland) was based on fractionated shea oil and had a solid fat content of 84% at 20 °C and 7% at 40 °C. Pea protein Pisane ® C9 (Cosucra, Warcoing, Belgium) had a protein content of 67.6% (N × 5.36). The sugar beet pectin (Herbstreith & Fox GmbH & CO. KG, Neuenbürg, Germany) had a molecular weight of 45.0 kDa and a 54% degree of esterification. Methylcellulose (WELLENCE™ Vege Form 183) was obtained from Danisco (Wilmington, DE, USA). Transglutaminase (ACTIVA WM from Streptoverticillium mobaraense with an enzyme activity of 114 U/g, determined by the manufacturer) was purchased from Ajinomoto Foods Europe SAS (Hamburg, Germany). Laccase from Trametes versicolor was purchased from Sigma Aldrich Co. (St. Louis, MO, USA) and had an enzyme activity of 0.9 U g −1 as determined by the manufacturer. Pepper, paprika, coriander and red beet powder were from Gewürzmüller (Salzburg, Austria).

Preparation of textured vegetable protein, fat systems, and binders
The high-moisture textured vegetable protein was thawed and chopped in a bowl chopper (K20, Seydelmann, Aalen, Germany) at stage one for 35 rounds until small textured vegetable protein (TVP) particles were obtained (Fig. 1A). An emulsified and crosslinked fat crystal network was prepared according to Dreher, Blach, Terjung, Gibis and Weiss [9] and served as fat mimic. The ratio of 12% soy protein isolate suspension to rapeseed oil was 40:60. The fat mimic was chopped into small particles with a bowl chopper at stage one for 10 rounds and will be from now on referred to as 'emulsified fat gel particles' (Fig. 1B). Solid fat served as another fat system after chopping into small particles with a bowl chopper at stage one for 50 rounds. It was labelled as 'solid fat particles' (Fig. 1C).
A previously used method [12] was used to prepare a pea protein-sugar beet pectin binder ( Fig. 1D) that had a protein to pectin ratio of 2:1, a total biopolymer concentration of 22.5% (w/w), and a pH adjusted to 6.0 using a food processor (5KSM150, Kitchen Aid Inc., St. Joseph, MI, USA). A methylcellulose hydrogel with a concentration of 5% (w/w) was prepared in a food processor (UMC 5, Stephan, Diessen, Germany) under vacuum and served as a benchmark binder (Fig. 1E).

Preparation of burger type meat analogues
The TVP particles (= 70%) were mixed manually with 10% fat mimic system (emulsified fat gel particles or solid fat particles). Then, 20% binder (pea protein-sugar beet pectin mixture or methylcellulose hydrogel) were added and mixed until the different structural elements were combined well. In addition, a spice mixture (5% red beet, 2% salt, 0.4% pepper, 0.3% paprika, 0.3% coriander with respect to the amount of binder) was incorporated. For the laccase-treated binder consisting of pea protein and sugar beet pectin, a certain amount of laccase that corresponded to 100 nkat g −1 solid substance in the binder was added [18]. Burger patties weighing 80 g were formed using a patty press maker (Hela, Ahrensburg, Germany; diameter = 8.5 cm) and from here on labelled as 'raw burger patties'.
The raw burger patties covered with aluminum foil and heated at 45 °C in a heating chamber (Reich, Urbach, Germany) 99% humidity for 4 h to accelerate laccase action. The core temperature of the burger patties reached 40 °C within 30 min. Then, an inactivation step at 95 °C for 10 min followed, during which a core temperature of 90 °C was reached. Afterward, the burger patties were cooled at 5 °C and labelled as 'cooked burger patties'. Last, the cooked patties were fried with an electric dual slide contact grill with smooth surfaces (Garland XPE12, Welbilt, Eglfing, Germany) for 130 s, where both the upper and the lower plate had a temperature of 218 °C. These patties are referred to as 'fried burger patties'. Photographic images were made during all stages of burger preparation.

Texture analysis
Texture measurements of raw, cooked, and fried burger patties were done using an Instron universal material testing machine (Instron Model 3365, Instron Engineering Corporation, Norwood, Massachusetts, USA) equipped with a Warner Bratzler v-shaped guillotine blade. For this, 1 cm wide stripes were cut from the middle of the burger patties and placed under the guillotine blade that cut the sample with a speed of 20 cm min −1 . The maximum force (N) served as measure of hardness. The raw and cooked burger patties were equilibrated to room temperature before measurement and the fried burger patties were analyzed immediately after frying. Ten measurements were done per sample.

Sensory analysis
The fried burger patties were evaluated by 20 panelists in terms of hardness, juiciness, cohesiveness, and texture acceptance on a scale ranging from − 5 to 5 immediately after frying. The score zero referred to an ideal vegan burger patty reference as defined by each panelist. The negative and positive attributes at the end of the scale (i.e., − 5 and 5) were labelled as too soft/too hard (hardness), too dry/too juicy (juiciness), too crumbly/too chewy (cohesiveness), and worse/better (overall texture), respectively. Furthermore, the panelists evaluated the appearance of the cooked burger patties from −5 (worse) to 5 (better), where again zero referred to an ideal vegan burger patty reference.

Statistics
SPSS statistics (V27, IBM Corporation, Armonk, NY, USA) was used to conduct a one-way analysis of variance (ANOVA) using the Duncan post-hoc test with a significance level of α = 5%. An independent t test (SPSS Statistics 26.0, IBM Corp., NY, USA) was done to determine statistical differences (α = 5%), where ANOVA was not applicable.

Properties of the raw materials
The TVP pieces were minced into particles with heterogeneous shape and size (Fig. 1A). Most particles were < 1 cm as this is the target size for meat particles in beef burger patties [19]. The minced TVP particles were found to be suitable for plant-based burger patties in preliminary experiments. The emulsified fat gel particles (Fig. 1B) and the solid fat particles (Fig. 1C) were ground to similar sizes as the TVP particles. All of those structured ingredients were inherently not sticky and, therefore, could not be shaped into 1 3 burger patties without the use of a binder. A mixture of pea protein and sugar beet pectin (r = 2:1, 22.5% w/w, pH 6.0) was chosen as a binder, which had been previously found to be viscoelastic and sticky (Fig. 1D). A methylcellulose hydrogel (5% w/w) served as a benchmark binder, and was firmer and more elastic (Fig. 1E). The binders used in this study were less concentrated than in a study on a bacon type meat analogue (25% and 6% w/w) [17] to increase viscous properties and, therefore, ease incorporation of binders into the aforementioned mixture of structured and functional ingredients.

Formability of the burger type meat analogues
TVP particles and emulsified fat gel particles or solid fat particles and spices were mixed before a binder was added to ensure homogeneous distribution. During the mixing process, the emulsified fat gel particles took up some red color form the red beet powder, while the solid fat particles retained their color. This led to visible white spots on the raw burger patties with solid fat particles ( Fig. 2A).
The pea protein-sugar beet pectin binder could be incorporated into the mixture more easily as it was runnier and more viscous than the methylcellulose hydrogel (Fig. 1D,  E), i.e., the pea protein-pectin binder was able to flow upon mixing, thereby coating the different structural elements better. This led to a burger mass that was rather sticky and could easily be formed into a burger patty ( Fig. 2A). Preliminary experiments showed that a higher concentration of the binder (30%) increased stickiness, thus leading to sticking of the mass to the patty press maker, whereas a lower concentration of the binder (10%) resulted in limited cohesion causing patties to fall apart during forming. There was no detectable difference in formability when laccase was added to the pea protein-sugar beet pectin binder, most likely because the crosslinking between the pea protein and the sugar beet pectin had not occurred yet or only to a minor extent. The incorporation of methylcellulose into the mixture was more difficult and time-consuming, since the binder was more elastic and, therefore, made it harder to ensure that a homogeneous mixture was obtained. Furthermore, the burger mass was less sticky, which made the shaping of burger patties more difficult. In case of the raw burger patties with methylcellulose and solid fat particles, the coherence of the different structural elements was visibly worse ( Fig. 2A). The lower stickiness and coherence of the burger patties with methylcellulose binder was also the reason why no textural analysis of the raw burger patties with methylcellulose was possible (Table 1), since those samples fell apart during attempts to cut them into stripes before measurement. There was no significant difference in hardness (p ≥ 0.05) among the samples with pea protein-pectin binder. As stated above, laccase action had not yet taken place that could have increased hardness. Furthermore, the fat system only Fig. 2 Appearance of raw, cooked, and fried burger type meat analogues with different binders and fat systems. The insert shows the cooked burger with pea protein-sugar beet pectin with laccase and solid fat particles from the bottom side. The scale bar equals 1 cm contributed to 10% of the overall mass and, therefore, was unlikely to make a substantial difference. Overall, results that's show that the pea protein-sugar beet pectin mixture was able to glue the TVP particles and fat system particles together better than methylcellulose.

Influence of cooking on the burger type meat analogues
The applied heat treatment of the burger patties at 45 °C was done to accelerate laccase action that crosslinks pea protein and sugar beet pectin leading to solidification of the pea protein-pectin binder as shown before [16]. Heating at 95 °C induces enzyme inactivation, which is a typical step in the food industry prior to selling [20], and also leads to a reduction of microbial contamination, thereby improving product shelf life.
All burger type meat analogues took on a brown color with a loss of red after cooking (Fig. 2B). Herbach, et al. [21] reported a discoloration of red beet juice upon heating at 85 °C due to degradation of red-purple betalain into yellow and orange products. Furthermore, it was shown that red betanin that is present in red beet powder is oxidized into yellow and colorless products through oxidation catalyzed by laccase [22]. All burger patties lost their stickiness after the heat treatment caused by an increase in binder cohesion [12]. In the case of burger patties with pea protein-sugar beet pectin acting as binder, laccase induced crosslinks between the two biopolymers that led to solidification, i.e., higher cohesion [16]. Unexpectedly, the burger patties with no laccase addition also lost their stickiness, although no crosslinks could be formed between the pea protein and the sugar beet pectin. It is possible that some water may have evaporated during the process despite maintaining a watersaturated atmosphere and covering samples. Furthermore, the textured vegetable proteins may have taken up some water from the binder system, thereby increasing concentration of the latter. In burger patties made with methylcellulose, heating led to physical crosslinks between methylcellulose that resulted in gelling, which increases cohesion [14].
Although gelation of methylcellulose is thermos-reversible, it may be that the time for cooling was not long enough to allow for a full reversibility of the gelling process. Again, all burger patties with the pea protein-sugar beet pectin binder were coherent, while the ones with methylcellulose and in particular the burger patties with solid fat particles appeared crumblier (Fig. 2B). Interestingly, the burger patties with solid fat particles did not contain white fat spots after the heat treatment anymore, since the solid fat melted during the heat treatment and accumulated at the bottom as shown for the burger patty with the pea protein-pectin binder and the solid fat particles (Fig. 2B). From this we can conclude that the fat binding capacity of the used binders was limited. From a consumer point of view, such an accumulation of solid fat at the bottom of the burger patty is unappealing. However, the appearance of the cooked burger patties was better for samples with the pea protein-sugar beet pectin mixture without laccase, when solid fat particles were used instead of emulsified fat gel particles (Table 2). It may be that some consumers find visible fat particles in cooked burger patties unappealing. It should be noted that the panelists did not evaluate the bottom of the burger patties, where the fat accumulated. In general, the lack of red color and stickiness after cooking led to the appearance among panelists that the burger patties were not 'raw'. Therefore, they should not be compared to raw minced meat burger patties but rather pre-cooked beef burger patties.
Interestingly, there was no difference in hardness for the cooked burger patties ( Table 1), indicating that the used fat system did not make a difference and also the used binder did not significantly (p ≥ 0.05) affect hardness. Furthermore, there was no increase in hardness from the raw burger patties to the cooked burger patties (Table 1). In other words, the solidification of the binders upon cooking did not lead to an increase in hardness. It may be that the TVP particles were the hardest structural element in the patties and determined the maximum force during texture analysis, whereas the binder did not play a role. The hardness of the burgers with methylcellulose was higher when emulsified fat gel particles were used (Table 1). It may be that the melting of Table 1 Hardness of raw, cooked, and fried burger type meat analogues with different binders and fat systems Data points with different small and capital letters denote a statistical difference (p < 0.05) within each treatment (impact of binder and fat system) and within each binder and fat system (impact of treatment), respectively the solid fat particles during heat treatment led to a weakening of the hydrophobic methylcellulose acting as a binder. Bakhsh, et al. [23] reported that the maximum force during Warner Bratzler shear tests of a textured isolate soy protein burger with 1.5% methylcellulose addition was 2.41 N, which is comparable to the cooked burger patties in our study (Table 1). To compare, beef burger patties with the same concentration of methylcellulose (1.5%) were significantly harder (= 3.60 N) in the study of Bakhsh, Lee, Lee, Sabikun, Hwang and Joo [23].

Influence of frying on the burger type meat analogues
Similar to beef burger patties, consumers expect that frying in the pan or on a grill leads to some textural and optical changes of the plant-based burger patty [24]. The burger patties developed a visible crust after frying (Fig. 2C) due to water evaporation of the burger patties and the related increase in solid matter [25] as well as due to polymerization caused by Maillard reactions [26]. This was responsible for their higher hardness (p < 0.05) compared to the raw and cooked burger patties as determined by texture analysis (Table 1). An exception to this was the burger patty with the pea protein-sugar beet pectin binder and the emulsified fat gel particles that was not statistically different (p ≥ 0.05) from the raw or cooked matrix. This sample was also significantly softer than the burger patty with methylcellulose and emulsified fat gel particles (Table 1). In general, the standard deviation of the textural analysis was rather high due to the heterogeneous structure of the samples. Interestingly, the hardness as evaluated by sensory analysis showed that the sample with the pea protein-sugar beet pectin binder and the emulsified gel particles scored − 0.2, which was close to ideal (i.e., zero) ( Table 2). All other samples had lower scores, indicating that they were softer. The discrepancy between textural and sensorial analysis in terms of hardness may arise from the fact that in textural analysis hardness is measured as the highest force, while panelist taking part in sensory analysis tend to evaluate hardness not only as the first bite but also as a measure of how well the sample falls apart and disintegrates, which also falls under the definition of cohesiveness [27,28]. All burger patties with the pea protein-sugar beet pectin as a binder had a better score in cohesiveness (closer to zero) than patties made with methylcellulose ( Table 2). This means that they were evaluated as less crumbly. The better cohesiveness of the fried burger patties with the pea protein-pectin binder agrees with the previously mentioned trend that those patties could be formed into burger patties easier (Section "Formability of the burger type meat analogues"). In other words, the pea protein-pectin binder facilitated the sticking of the different structured ingredients together better in the raw state and after frying which may be due to stronger bonds between them as facilitated by the binder [28]. A possible explanation for this could be that the hydrophobic methylcellulose cannot adhere well to the amphiphilic TVP particles that made up most of the bulk material. On the other hand, the pea protein-sugar beet pectin mixture has a higher ability to thermodynamically adhere to those particles and bind them together [29].
The burger patties with pea protein-sugar beet pectin without laccase and solid fat particles had the highest juiciness score, which may be due to melting of the solid fat during frying. However, such trends could not be confirmed for the other burger patties with different binders. Interestingly, there were no differences in overall acceptability of the texture indicating that the cohesiveness was only a minor factor in the evaluation of this criteria.

Conclusions
A novel binder based on pea protein and sugar beet pectin was runny and sticky and, therefore, suitable to glue different structured ingredients together in a plant-based burger as hypothesized. Furthermore, those burger patties were more cohesive and less crumbly than the same burgers made with methylcellulose as binder, thus proving our initial hypothesis. Interestingly, the solidification of the pea protein-sugar beet pectin binder did not increase hardness of the burgers, indicating that this property is rather governed by the structural elements instead of the binder. The study gave insights into the binding mechanism in meat analogue products and may be of interest for product developers.