Development and Characterization of a Low-Fat Mayonnaise Salad Dressing Based on Arthrospira platensis Protein Concentrate and Sodium Alginate

Featured Application

This work presents an alternative approach to produce low-fat mayonnaises with good acceptability and using clean label ingredients, useful for developing new formulations in this type of products.

Abstract

The food industry is constantly reformulating different foods to fulfill the demands of the consumers (natural ingredients and good sensory quality). The present work aimed to produce low-fat mayonnaises using 30.0, 22.5, and 15.0% oil, 1% soy protein isolate (SPI) or spirulina (Arthrospira platensis) protein concentrate (SPC), and 2% sodium alginate. The physical properties (thermal stability, rheological behavior, and particle size), the sensory attributes (appearance, texture, taste, and acceptability), the purchase probability, and amino acid availability (after a simulated digestion) were evaluated. The mayonnaises demonstrated good thermal stability (>90%) using 22.5 and 15% oil, all products showed shear-thinning behavior and a consistency index of 20–66 Pa·s. The reduction of oil from 30 to 15% increased the particle size from 6–9 µm to 10–38 µm. The most acceptable product was the formulated with SPI and 22.5% oil (8.3 of acceptability and 79% of purchase probability). Finally, the addition of proteins improved the total essential amino acids compared to a commercial product (28 and 5 mg/25 g, respectively). In summary, it was possible to obtain well accepted products with high purchase probability using low concentrations of oil and vegetable proteins.

1. Introduction

In recent decades, the demand for healthier foods has risen, boosting the interest of the food industry to reformulate a wide variety of products. This reformulation includes reducing a specific component (such as sugar or fat content) or the use of naturally derived ingredients [1], but without compromising the food safety and the sensory quality. Fulfilling the latter requirements is especially challenging in food products where fat is the primary component, for example, in ice creams, vegetable spreads (i.e., peanut butter), and salad dressings (i.e., mayonnaise), because the fat plays a crucial role as structuring agent, contributing to the development of texture [2].

Mayonnaise is a worldwide-consumed seasoning used in several types of foods, for instance, sandwiches, salads, hamburgers, hot dogs, seafood, and others. The traditional formulation contains approximately 70–80% of oil, vinegar, and spices; however, the consumers usually perceive its high oil content as potentially unhealthy [3]. This conception led to the elaboration of low-fat mayonnaise, which contains 20–40% oil, this has been possible by using thickening agents such as modified starches or gums to develop texture and physical stability in the product [4,5,6]. However, the use of the latter ingredients might have some sensory disadvantages. For instance, the texture or taste of the product could be negatively affected [7,8], and some of these ingredients (modified starches) could be perceived as potentially harmful [9].

One alternative to overcome the negative texture impact of reducing the fat content in mayonnaises is the use of Emulsion-Filled Gels (EFG). The EFGs are gel matrices in which emulsified oil droplets replace the water phase of the gel. These systems are stable against creaming, flocculation, and coalescence, have a gel-like behavior, and they can be easily produced with a wide range of biopolymers such as gums (pectin, tragacanth gum, gellan gum, alginates, carrageenan) and proteins [10,11]. To date, the application of EFGs in mayonnaise has been demonstrated to mimic the sensory properties (texture and mouthfeel) of the regular product using egg yolk (as emulsifier), tragacanth gum, and sodium alginate (as a thickening and gelling agents) [12,13], fulfilling the actual demands of the consumers (low-fat and naturally derived ingredients).

There are many sources for obtaining natural derived biopolymers, for instance, cereals, tubers, oilseeds, and algae. These latter are comprised by macroalgae and microalgae, where the first one is a common source of sodium alginate; this biopolymer is used as thickening and gelling agent, and one of the most important characteristics of this compound is the irreversible gelation when is combined with ionic salts, for example, calcium chloride [14]. In the case of microalgae, the Arthrospira platensis, commonly known as spirulina, is nowadays an important raw material to extract high valuable ingredients, because it contains unsaturated fatty acids (omega 3), pigments (phycocyanin), and a high concentration of proteins (40–50%) [15,16] with promising applications due to its emulsion capacity (approximately 60%) and stability (ranging from 80–100%) [17,18,19,20].

In our previous work [21], it was observed that the consumers of mayonnaise (in Mexico) perceived the proteins as natural ingredients and have good acceptability, hence, it was decided to produce mayonnaises using egg yolk, vegetable proteins, and sodium alginate. Therefore, the present study aimed to exploit the techno-functionality (i.e., emulsifying properties and water and oil holding capacities) of the A. platensis protein in combination with sodium alginate to produce low-fat mayonnaises (using an EFG approach) with comparable physical and sensory characteristics than a commercial low-fat mayonnaise.

2. Materials and Methods

Microalgae (A. platensis) was purchased from Galtec Algae Technologies (Guadalupe, Mexico). Soy protein isolate containing 91.9% w/w of protein was purchased from Food Technologist Trading S.A de C.V. (Atizapan de Zaragoza, Mexico). Soybean oil, salt, sugar, vinegar, and low-fat commercial mayonnaise (41% of total fat) used as reference were purchased from local groceries in Puebla, Mexico. Sodium alginate and calcium chloride were obtained from Sigma-Aldrich Co. (St. Louis, MO, USA) and J.T Baker (Phillipsburg, NJ, USA), respectively. All reagents used for the analyses (amino acid mix, o-phthaldehyde (OPA) and 9-fluorenylmethylchloroformate (FMOC)) were analytical grade obtained from Sigma-Aldrich Co. (St. Louis, MO, USA).

2.1. Spirulina Conditioning and Protein Extraction

The biomass (A. platensis) was freeze-dried, grind, and sieved using a 300 µm mesh. The freeze-dried spirulina was defatted twice with n-hexane at room temperature (for 30 min), at a ratio of 1:5 (powder: solvent w/v), then the n-hexane was decanted, and the remaining solvent was allowed to evaporate overnight at room temperature.

To obtain the Spirulina Protein Concentrate (SPC), the defatted biomass was subjected to alkaline solubilization for the protein extraction at pH 12 using a NaOH solution at 0.05 N in a ratio of 1:10 (w/v) for 1 h (at 25 °C). Afterwards, the sample was centrifuged (4500 rpm, 4 °C, 10 min), and the pH of the supernatant was adjusted to 3.5 to precipitate the proteins. The precipitate was collected by centrifugation (at the same conditions), rinsed with distilled water, and centrifuged again [17]. The protein concentrates were freeze-dried at −50 °C for approximately 48 h and stored in sealed flasks.

2.2. Protein Concentrates Characterization

2.2.1. Protein Content

The protein content was determined spectrophotometrically using the Total Protein Kit, Micro Lowry, Peterson’s Modification (Sigma-Aldrich Co., St. Louis, MO, USA). This method was selected to avoid the interaction of the polyphenolic compounds of the spirulina. First, 500 μL of solubilized protein (at pH 13 to promote full solubilization) was mixed with 50 μL a deoxycholate solution (0.15%) and followed by 50 μL of trichloroacetic acid (72%), the sample was allowed to stand for 10 min after the addition of each reagent. These solutions promote the protein precipitation, which was further collected by centrifugation (5000 rpm at 25 °C for 5 min). The protein pellet was dissolved with 500 μL of Lowry’s reagent and 500 μL of water. After 20 min, 250 μL of Folin Ciocalteau reagent was added and allowed to react for 30 min. Finally, 200 µL of the samples were transferred to a 96 wells microplate and the absorbance was measured using a UV-Vis Multiskan Sky Microplate (Thermo Fischer Scientific, Walthman, MA, USA) spectrophotometer at 720 nm. A solution of bovine serum albumin (0–1000 µg/mL) was used as standard following the same procedure. The protein content was calculated using Equation (1) and expressed as a percentage (%).

$$\mathrm{Protein} \mathrm{content} (\%)=\frac{P_s}{S_w}\times 100,$$

(1)

where P_S is the total protein quantified in the sample and S_W is the initial sample weight.

2.2.2. Water Holding Capacity (WHC) and Oil Holding Capacity (OHC)

Water and oil holding capacities (WHC, OHC, respectively) were measured by weighing 0.25 g of protein in a 50 mL Falcon tube. Then, 10 mL of distilled water or oil was added, and the sample was stirred using a vortex for 10 s every 5 min (for 30 min). Afterwards, the samples were centrifuged (at 4500 rpm, 4 °C, for 10 min), the supernatants were discarded, and the WHC and OHC were calculated by weight difference as is shown in Equation (2) [22].

$$\mathrm{WHC} \mathrm{or} \mathrm{OHC}=\frac{\mathrm{Final} \mathrm{weight}-\mathrm{Initial} \mathrm{weight}}{\mathrm{Initial} \mathrm{weight}},$$

(2)

2.3. Mayonnaise Preparation

The continuous phase, constituted by an aqueous solution added with water, vinegar, sugar, salt, and protein (soy or spirulina), was homogenized (Ultraturrax (IKA, Staufen, Germany) with the dispersed phase (egg yolk and oil) at 7500 rpm (60 s). A double emulsion was prepared by adding a solution of sodium alginate (2%) to the primary emulsion at a 1:1 ratio. Finally, for gelation of sodium alginate in the secondary emulsion, 1 mL of CaCl₂ (0.5 % w/v) was dropped in constant stirring at 7500 rpm [13]. Finally, we obtained nine different formulations of low-fat mayonnaise (

Table 1. The concentration of the ingredients in the low-fat mayonnaise.

			Aqueous Phase
Sample	Oil (%)	Egg Yolk (%)	Water (%)	Vinegar (%)	Sugar (%)	Salt (%)	Protein (Type; %)	Sodium Alginate (%)
A	30.0	10	22	4.5	2	1.5	NA	30
B	22.5	10	26	4.5	2	1.5	NA	34
C	15.0	10	30	4.5	2	1.5	NA	38
D	30.0	10	21	4.5	2	1.5	SPI; 1	30
E	22.5	10	25	4.5	2	1.5	SPI; 1	34
F	15.0	10	29	4.5	2	1.5	SPI; 1	38
D’	30.0	10	21	4.5	2	1.5	SPC; 1	30
E’	22.5	10	25	4.5	2	1.5	SPC; 1	34
F’	15.0	10	29	4.5	2	1.5	SPC; 1	38

SPI = soy protein isolate. SPC = spirulina protein concentrate.

).

2.4. Mayonnaise Characterization

2.4.1. Stability Analysis

The thermal stability of the samples was studied by pouring approximately 20 g of mayonnaise into a sealed Falcon flask (50 mL of capacity) and heating them in a water bath (85 °C) for 30 min. Afterwards, the samples were centrifuged at 4000 rpm for 10 min to induce phase separation and the stability was calculated with Equation (3) [23].

$$\mathrm{Mayonnaise} \mathrm{stability} \left(\mathrm{ES}\right)=\frac{\mathrm{Remaining} \mathrm{emulsified} \mathrm{layer} \left(\mathrm{cm}\right)}{\mathrm{Initial} \mathrm{emulsified} \mathrm{layer} \left(\mathrm{cm}\right)}\times 100$$

(3)

2.4.2. Particle Size Analysis

The particle size was measured using dynamic light scattering with a Bluewave Nanotrac Wave II (Microtrac Inc., Montgomeryville, PA, USA) analyzer. All samples were diluted 1:100 (v/v) with distilled water and homogenized with a Vortex at full speed for 15 s. The refractive index was set at 1.47 for oil droplets and 1.33 for water [24], the results were expressed as mean particle size (μm) and the span was calculated using Equation (4).

$$\mathrm{Span}=\frac{d_{90}-d_{10}}{d_{50}},$$

(4)

where d₉₀, d₅₀, and d₁₀ represents the diameter of the 90%, 50%, and 10% of all particles measured.

2.4.3. Rheological Characterization

The viscosity was determined at 25 °C using a Brookfield Viscometer DV II (AMETEK Inc., Devon-Berwyn, PA, USA) equipped with an LV-3 spindle at 0.25, 0.5, 1, 2, 5, 10 and 16.67 s⁻¹ for 45 s. The data were fitted to the power law model [25], this mathematical model allows a simple and easy interpretation of the flow behavior of a non-Newtonian fluid. This process was done using Microsoft Office Excel 365 solver to obtain the consistency index (K) and power law index (n) using the following expression.

$$\sigma =K{\gamma }^n,$$

(5)

afterwards, the loss of consistency index (ΔK) was calculated using Equation (6) to assess the effect of oil concentration and the vegetable proteins in the loss of viscosity of the systems.

$$\Delta K=|K_{30}-K_i|,$$

(6)

where K₃₀ is the consistency index of the sample with 30% oil and K_i is the consistency index of the sample.

2.4.4. Sensory Analysis

A total of 50 non-trained panelists were recruited in “Universidad De Las Américas Puebla”. The samples were spread on a wheat cookie and labeled using three random digits. The sensory evaluation of the mayonnaises was studied using a structured scale from 1 to 10 (1 = lowest score and 10 = highest score) for assessing the general appearance, taste, texture, and acceptability. Furthermore, the panelists were asked for their purchase intention (possible answers: Yes/No) for each sample after its evaluation; the results were expressed as percentage (%) [26].

2.4.5. Amino Acid Bioavailability

To investigate the nutritional enhancement of proteins (bioavailability of amino acids) in the low-fat mayonnaise, the products with the best purchase intention with one of each vegetable protein (SPI and SPC) were selected for an in vitro gastric digestion simulation (GDS). First, artificial saliva (AS) and gastric juices (GJ) were prepared to mimic the buccal and gastric environment. The AS was prepared by solubilization of α-amylase (0.2 mg/mL) in phosphate buffer at 20 mM (pH 7) and GJ consisted of pepsin enzyme in HCL solution (0.1 M) at 3.2 mg/mL. The digestion in the mouth was simulated by homogenizing mayonnaise (25 g) and 10 mL of AS using a magnetic stirrer at 150 rpm for 3 min (at 37 °C). Afterward, the gastric digestion was simulated by adjusting the pH to 2.0 using concentrated HCl (4 M) and adding 15 mL of GJ using the same stirring conditions as above; this digestion was done for 60 min. The conditions of the experiment were selected according to the study of Rui et al. [27], which selected the appropriate food to digestive juices ratios based on the human physiology (food to AS = 2.5:1; food to GJ = 1.5:1) [28].

The sample collection was done after the buccal digestion (time 1) and gastric digestion (time 2), whereas the fresh sample was considered as time 0. For each sample, the amino acid profile was determined (time 0, 1, and 2) with the Agilent amino acid analysis protocol [29] in an Agilent 1290 Infinity II (Agilent Technologies Inc., Santa Clara, CA, USA) HPLC coupled with a diode array detector (DAD). The column used was a ZORBAX Eclipse Plus C18 (Agilent Technologies Inc., Santa Clara, CA, USA), the mobile phase A was a 10 mM of Na₂HPO₄:10 mM Na₂B₄O₇ (pH 8.4) buffer, and the mobile phase B was a solution of acetonitrile: methanol: water (45:45:10). The column temperature was at 40 °C, and the flow rate was 1.5 mL/min. The total time of the assay was 16 min, with 98% of mobile phase A (at 0.35 min), 100% of mobile phase B at 13.5 min, and 98% of mobile phase A used at 16 min. Amino acids (primary and secondary) were derivatized using o-phthalaldehyde (OPA) and 9-fluorenylmethyl chloroformate (FMOC). The amino acids detection was done at 263 and 338 nm and quantified using norvaline (internal standard) and a calibration curve (amino acid mix).

2.5. Statistical Analysis

All experiments were done in triplicate (n = 3). The data was expressed as mean with standard deviation, median with interquartile ranges, and percentage where appropriate. The statistical differences were analyzed using ANOVA (p < 0.05). Afterward, a pairwise t-test was used to determine the sample’s mean differences (p < 0.05). The relationship between mayonnaise properties and the sensory scores was determined by a correlation coefficient analysis and expressed as Correlation Index (CI). All statistical analyses were done in Microsoft Office Excel 365 using the open-source Real Statistics Resource Pack [30]. Artwork (Figures) was processed in Python 3.9 using Matplotlib (v. 3.4.2) library in Spyder (v. 5.1.1) IDE.

3. Results and Discussion

3.1. Protein Concentrates Characterization and Techno-Functional Properties

The content of protein (dry basis) found in SPC was 66.1 ± 1.5%, this concentration was slightly lower to others reports on A. platensis protein concentrates (69–75%) [19,31]. The difference in protein concentration may be due to the biomass growth conditions or the protein extraction process, in this regard, the use of pretreatments (sonication or high pressures) for cell disruption as used in [19,32] and [33] could increase the protein concentration up to 73–85% in the concentrates.

The WHC and OHC properties of SPC were 0.49 ± 0.029 g/g and 3.25 ± 0.218 g/g, respectively. The WHC observed for the SPC was very lower than previous reports (3–5 g/g) [17,20] but similar than the values found by Bleakley and Hayer [32]. In the case of the OHC, it was slightly higher than the reported by Benelhadj et al. (2.5 g/g) [17] but lower than the OHC of the SPC in [20,32], these studies demonstrated around 6–8 g/g, this enhanced OHC could be attributed to a partial unfolding of the protein structure due to the sonication pretreatment for the protein extraction.

For soy protein (control) the WHC (18.77 ± 1.091 g/g) and OHC (1.48 ± 0.126 g/g) were significantly higher and lower (p < 0.05), respectively, compared to the results of SPC. In related works [34,35,36], the OHC of soy protein (1.0–5.5 g/g) agrees with the results of this study. In contrast, the WHC determined herein was higher than previous studies (3.5–6.9 g/g) [35,36,37]. The difference in WHC may be due to the processing of the protein or drying methods.

3.2. Mayonnaise Characterization

3.2.1. Stability Analysis

The physical stability of the mayonnaise is an important quality attribute, but also thermal stability for applications in hot dishes such as hamburgers, hot dogs, and grilled sandwiches, among others. The low-fat mayonnaises showed excellent physical stability (100% of stability and any syneresis was observed) during storage (30 days) at 4 °C. This result was consistent with previous reports of similar systems; for example, low-fat mayonnaises developed with A. platensis and Dunaliella proteins and starch showed a low reduction of G’ and G’’ during 60 days of storage, which was indicative of a physically stable network [38].

In the case of the thermal stability (Figure 1), this property was negatively affected by the increase in the oil concentration (p < 0.05), for instance, the stability was in the range of 81–90% at 30% of oil, where the highest value was found in D’, E’, and F’ systems. The mayonnaise formulated with 15% oil showed similar stability and presented the best stability (>94%) compared to all systems elaborated. Finally, as was expected, the commercial product did not show instability against high temperature (due to the high content of stabilizers), and it was significantly higher (p < 0.05) than all our samples.

Mayonnaises are physically stable emulsions due to the high viscosity of the system; however, the stability could be affected by ingredient interactions (proteins and polysaccharides). For instance, pea protein used in the low-fat mayonnaises decreased the product’s thermal stability (62%) [39]. In contrast, using starches or cereals flour has enhanced thermal stability (~99%) [7,40]. The products’ stability in this study agrees with gelled emulsions with sodium alginate (previously reported), which demonstrated ~95% stability at pH 2 and 4 [41]. Moreover, Yang et al. [42] and Li et al. [12] developed very similar mayonnaises to the present study (containing sodium alginate) with good thermal stability due to minor microstructural changes (observed by Confocal Laser Scanning Microscopy). However, these results cannot be compared with those obtained in the present work. A possible synergistic mechanism among the polysaccharides and proteins in our products probably helped produce systems with good thermal stability. In this regard, sodium alginate could increase the stability due to the irreversible gelation of the continuous phase, preventing the mobility of oil droplets [13,43]. Meanwhile, the WHC and the OHC could favor the interactions between the protein hydrophilic residues with the hydrogel and the hydrophobic protein surface with the oil.

3.2.2. Particle Size Analysis

The particle size in the samples increased when the oil concentration decreased and was significantly higher than the commercial product (p < 0.05) (Figure 2). All samples with 30% oil (A, D, and D’) demonstrated a mean particle size in a range of 6–9 μm; however, the presence of protein showed a slight reduction in particle size (6–7 μm). In samples with lower oil content (22.5%), there was an increase in the particle size for A, B, C, D, E, and F (~10 μm), while the systems with SPC (D’, E,’ and F’) showed the twice size compared to the samples with 30% oil. In the case of the mayonnaises with the lowest content of oil (15%), D’, E,’ and F’ showed smaller particle sizes (12 μm) in comparison to the rest of the samples where the particles increased up to 30 μm. Two possible mechanisms for this increase in particle size could occur. First, A, B, and C behavior could be due to the absence of a second emulsifier to help reduce the interfacial tension and reach smaller droplet sizes in emulsions [44]. In the case of D, E, and F, the SPI could contribute to a larger droplet size because the mayonnaise pH is close to this protein’s isoelectric point (4–4.5). Hence, when the net charge is near or equal to zero (at the isoelectric point of the protein) leads to poor emulsifying activity and the generation of larger droplet size [45].

Our products showed larger particle sizes than related studies of low-fat mayonnaises. For example, Sun et al. [46] developed mayonnaises with microparticulate Whey Protein Isolate (M-WPI), and the systems had approximately 2 µm in diameter. The difference in the particle size could be because the M-WPI have a tiny particle size (~8 µm), contributing to reducing the lipid droplets size in their products [46]. Nevertheless, the particle size determined in the samples with more than 22.5% oil showed comparable results to a previous report of low-fat mayonnaise (2–12 µm) elaborated with modified starches [7].

The span measurement indicates the width of the particle size distribution. Higher span values indicate more heterogeneous particle sizes within the system. The lowest span values (Figure 2) were found for the commercial low-fat mayonnaise and C (~0.9). In contrast, the rest of the samples showed values of 1.1–1.6, and the highest span (span = 2.7) was recorded in F’. The particle size distribution (Figure 3) in commercial mayonnaise and the samples with 22.5 and 30% oil (A, B, D, E, D’, and E’) was monomodal; on the contrary, the mayonnaises with 15% oil content (C, F, and F’) showed a bimodal or multimodal distribution. Usually, systems with high viscosity and homogeneous particle size distribution (monomodal) have been related to better stability. In contrast, multimodal distributions are usually observed in low viscosity and unstable systems [47]. In our low-fat mayonnaises, only the latter behavior was detected (higher viscosity for monomodal systems), and even though the particle size distribution (multimodal) could induce instability, the presence of the sodium alginate probably helped to prevent the oil droplet movement and their further coalescence in the products [10].

3.2.3. Rheological Characterization

Regarding the rheological behavior, the low-fat mayonnaises fit well with the power law model (r² = 0.998–0.999 and RMSE = 0.284–1.442). All samples showed a power law index in a range of 0.28–0.33 (

Table 2. Rheological behavior of low-fat mayonnaise samples using different oil concentrations (%).

Mayonnaise Sample	K (Pa·sn)	ΔK	n	RMSE	r2
Commercial	43.284	ND	0.297	0.583	0.999
A	66.262	0.000	0.281	1.442	0.998
B	35.008	31.223	0.260	1.397	0.999
C	20.922	45.291	0.335	0.915	0.999
D	35.284	0.000	0.281	0.883	0.999
E	31.524	3.524	0.299	1.062	0.998
F	25.876	9.408	0.329	0.380	0.999
D’	49.552	0.000	0.308	1.324	0.998
E’	43.879	5.673	0.305	0.839	0.999
F’	35.013	14.539	0.308	0.284	0.999

K (consistency index), n (power law index), RMSE (Root Mean Square Error), r² (Coefficient of Determination).

), indicating a shear-thinning fluid (pseudoplastic fluid) behavior (power law index < 1). The results herein agree with previous reports of low-fat mayonnaises where this parameter was 0.28–0.52 [25,48].

On the other hand, the consistency index (K) represents the viscosity of the sample; the highest K values were found in samples with 30% oil (66–35 Pa·sⁿ), while the minimum values (20–35 Pa·sⁿ) were determined in the mayonnaises elaborated containing 15% oil (Table 2). Uribe-Wandurraga et al. [38] prepared low-fat mayonnaises (30% oil) using Chlorella, Spirulina, and Dunaliella microalgae, demonstrating higher K (74–94.5 Pa·sⁿ) than those found in this work, probably because these products also contained 4% starch, which increased the viscosity of the product. However, the K values of our samples agree with previous reports of low-fat mayonnaises (K = 31–49 Pa·sⁿ) elaborated with different gums such as xanthan gum, guar gum, and corn-dextran [25,48]. Like the behavior observed herein, Park et al. [8] also determined that the consistency index of low-fat mayonnaise can reduce from 90–51 and 87–23 Pa·sⁿ when the oil content is reduced from 54 to 38%. This could be explained by the role of oil particle-particle interactions in developing the viscosity of mayonnaise [49]. Hence, if the oil content is low, the number of oil particle-particle interactions will be reduced, and the system’s viscosity will be lower.

It was found that the use of vegetable proteins, the WHC, and the OHC showed a negative correlation (CI = −0.936, −0.732, and −0.571, respectively) with the ΔK (loss of consistency index), indicating that proteins helped to mitigate the loss of viscosity. For instance, the reduction of oil content (from 30 to 15%) in samples with SPI (D, E, and F) and SPC (D’, E,’ and F’) results in minor losses of consistency index (ΔK = 3–15 Pa·sⁿ) in comparison to samples containing only sodium alginate (ΔK = 31–45 Pa·sⁿ). This was probably because the techno-functional properties of proteins might increase the number and strength of inter and intramolecular interactions [50,51].

The viscosity showed to be positively affected by the oil content (p < 0.05); as can be seen in Figure 4, the viscosity of the systems showed to reduce at higher shear stress; moreover, the viscosity at zero shear rate tends to reduce with reduced oil concentrations. The highest viscosities at zero shear rate were obtained in samples A (175 Pa·s) and D’ (125 Pa·s), while the lowest values were found in the D (~100 Pa·s). In summary, the samples A, D’ and E’ were demonstrated to have similar rheological characteristics (consistency index, viscosity, and power law index) to the commercial product.

3.2.4. Sensory Analysis

The results obtained from the sensory evaluation of the low-fat mayonnaises are shown in Figure 5; in addition, the relationship between the sensory scores and the products’ characteristics (oil concentration, presence of protein, and viscosity) was determined using a correlation matrix (Figure 6). It is noteworthy that sample A was not included in the correlation matrix because it was detected as an outlier for the correlation between K with the taste score and the acceptability of the product (the full correlation matrix can be consulted in Figure S1).

The appearance of all mayonnaises containing SPC (D’, E,’ and F’) was significantly lower (p < 0.05) in comparison to the rest of the samples, probably due to the green color of the SPC (Figure 7); therefore, proteins with creamy-like color such as from amaranth, oat, or rice could be considered for further reformulations. The rest of the products demonstrated identical evaluations (p > 0.05) compared to the commercial product. This sensory attribute was demonstrated to positively correlate with the oil concentration, the WHC, and the consistency index (+0.29, +0.49, and +0.47, respectively). This result probably because these factors enhance the thickness of the system and mimic the appearance of the mayonnaise. In contrast, the presence of proteins had a negative correlation (CI = −0.46) with the appearance, and the same trend was found for the OHC (CI = −0.84). In this regard, the negative effect of proteins and the OHC could result from the low scoring in the mayonnaises containing SPC.

The texture of the samples B, D, D’, E, E’, and F’ demonstrated similar scoring to the commercial product (p > 0.05). Conversely, the A, C, and F scored significantly lower than the commercial mayonnaise (p < 0.05). The main contributor to desired texture was the oil concentration (CI = +0.58), followed by the OHC of proteins (CI = +0.15). The K was negatively correlated with this sensory attribute (CI = −0.42), suggesting that even if the hydrogel formed provided similar viscosity to a commercial low-fat mayonnaise, using only sodium alginate did not mimic the lubricity and the smooth mouthfeel expected for this product [52]. However, the techno-functionality of the proteins could develop these textural properties [53].

All samples containing SPC (D’, E’, and F’) generally showed an acceptability score higher than 7. However, these values were significantly lower than the commercial mayonnaise (p < 0.05).

One of the decisive factors was the product’s appearance; nevertheless, the panelists commented that these samples with SPC could be directed toward producing flavored mayonnaises, such as jalapeño, coriander, and avocado taste. The other samples showed similar acceptability with the commercial mayonnaise (p > 0.05), except for the A sample. The most correlated characteristic with the acceptability was the appearance and the taste scores (CI = +0.78 and +0.75, respectively). Additionally, WHC and the consistency index (K) (CI = +0.49–+0.63, respectively) contributed to the acceptability mainly because they enhanced the products’ appearance and taste. In addition, the texture score showed an inverse relation with the acceptability; this could suggest that the consumers are more interested in the product’s appearance and taste than the texture.

The sensory evaluation of our products demonstrated better results compared to similar formulations. For example, Yang et al. [42] developed mayonnaises with an emulsion-filled gel approach using sodium alginate and 30% oil; these products showed acceptability around 4.9–6.9. On the other hand, comparable acceptability values (7–8.5) in microparticulate low-fat mayonnaises (20–40% oil reduction) using whey protein isolate are reported by [46].

Finally, the E, C, D, and the D’ showed similar or higher purchase (>65%) intention in comparison to the commercial low-fat mayonnaise (68%), indicating that these selected products have promising opportunities in the market (Figure 8). In contrast, the rest of the samples showed a lower probability of being purchased by the consumers.

3.2.5. Amino Acid Bioavailability

For evaluating the essential amino acids (EAA) provided by the proteins used, it was decided to study the bioavailability of EAA in the sample with higher purchase intention containing soy protein (sample E) and SPC (sample D’). Furthermore, these were compared to the EAA bioavailability in the commercial product (

Table 3. Essential amino acid profile for E, D’ and Commercial low-fat mayonnaises.

Amino Acid (mg/25 g Product)	D’	E	Commercial
	Time 0
His	ND	1.740 ± 0.068 a	ND
Thr	2.613 ± 0.096 a	2.266 ± 0.059 b	2.191 ± 0.157 b
Val	1.566 ± 0.098 a	1.102 ± 0.043 b	0.656 ± 0.076 c
Met	0.774 ± 0.155 a	0.621 ± 0.131 a	0.511 ± 0.054 a
Trp	8.149 ± 1.478 ab	7.533 ± 0.535 b	9.278 ± 0.106 a
Phe	2.120 ± 0.177 b	2.537 ± 0.083 a	ND
Ile	1.142 ± 0.300 a	1.059 ± 0.048 a	ND
Leu	3.400 ± 0.091 a	2.699 ± 0.187 b	2.007 ± 0.069 c
Lys	1.729 ± 0.108 a	1.506 ± 0.170 a	ND
∑ EAA	21.496 ± 1.203 a	21.066 ± 1.328 a	14.645 ± 0.464 b
	Time 1
His	ND	ND	ND
Thr	2.451 ± 0.162 a	2.217 ± 0.212 a	1.520 ± 0.033 b
Val	1.302 ± 0.071 a	1.032 ± 0.077 b	0.709 ± 0.004 c
Met	0.914 ± 0.149 a	0.753 ± 0.024 a	0.628 ± 0.005 b
Trp	10.624 ± 0.095 a	7.731 ± 0.997 b	10.966 ± 0.492 a
Phe	3.029 ± 0.190 a	2.435 ± 0.021 b	ND
Ile	1.331 ± 0.107 a	1.186 ± 0.038 a	ND
Leu	3.176 ± 0.478 a	2.825 ± 0.463 a	1.527 ± 0.064 b
Lys	2.303 ± 0.930 a	1.501 ± 0.093 a	ND
∑ EAA	25.132 ± 2.181 a	19.684 ± 1.929 b	15.352 ± 0.601 c
	Time 2
His	ND	ND	ND
Thr	3.219 ± 0.164 a	3.038 ± 0.131 a	2.481 ± 0.499 a
Val	1.326 ± 0.141 a	1.274 ± 0.115 a	0.610 ± 0.166 b
Met	0.850 ± 0.146 a	0.779 ± 0.024 a	ND
Trp	13.222 ± 0.048 a	13.354 ± 1.291 a	2.471 ± 0.301 b
Phe	3.587 ± 0.630 a	3.113 ± 0.352 a	ND
Ile	1.355 ± 0.140 a	1.094 ± 0.187 a	ND
Leu	3.714 ± 0.235 a	3.324 ± 0.065 a	ND
Lys	1.992 ± 0.191 a	2.076 ± 0.105 a	ND
∑ EAA	29.265 ± 1.695 a	28.052 ± 2.270 a	5.562 ± 0.966 b

ND = Not Detected. E = mayonnaise with soy protein isolate and 22.5% oil. D’ = mayonnaise with spirulina protein concentrate and 30% oil. ∑ EAA = sum of Essential Amino Acids. Different letters in the same row indicates significant difference among samples (p < 0.05).

). Control samples (not subjected to digestion) type E (soy) contained all essential amino acids; in the case of sample D’ (spirulina), only His was not detected. The bioavailability profile showed a nutritional improvement compared to commercial samples, which lacked four essential amino acids (His, Phe, Ile, and Lys). The total EAA (∑ EAA) in low-fat mayonnaises showed around 21 mg/25 g, which overcomes (p < 0.05) to the ∑ EAA of commercial products (14.645 ± 0.465 mg/25 g).

After the mouth digestion (time 1), there was an increase in the concentration of Val, Trp, Phe, Ile, and Lys in the D’ sample. Similarly, sample E slightly increased Trp, Phe, Ile, and Leu, but the His was not detected in this sample. Probably the amylase and environment conditions (pH) promoted hydrogel hydrolysis (artificial saliva) formed by sodium alginate gum, releasing the protein and thus increasing its availability [27,54]. In this step, the highest ∑ EAA found was in D’ (25.132 ± 2.181 mg/25 g), followed by E (19.684 ± 1.929 mg/25 g), remaining the lowest in the commercial low-fat mayonnaise (15.352 ± 0.601 mg/25 g).

After gastric digestion (time 2), all EAA increased, standing out samples E and D’ (made with plant proteins). Only three EAA (Thr, Val, Trp) remained in commercial samples, while E and D’ samples were complete (except for His). The ∑ EAA of the commercial product (5.562 ± 0.966 mg/25 g) was significantly lower (p < 0.05) in comparison to our low-fat mayonnaises (∑ EAA = 28–29 mg/25 g). The increase of EAAs after gastric digestion could be explained by the disruption of the peptide bonds induced by the pepsin activity (proteolytic), provoking the release of a higher amount of EAAs from the low-fat mayonnaise [55,56].

This digestion phase is essential because es followed by the ileal digestion and absorption of amino acids; hence, our products could provide a higher amount of available EAAs. Even though the concentrations of EAAs in the mayonnaises did not fulfill the FAO requirements [57], the formulations proposed herein could slightly contribute to EAA intake in the diet with approximately one tablespoon.

4. Conclusions

The present study demonstrated that the emulsion-filled gels could be used to formulate low-fat mayonnaises. These products showed good physical stability, similar rheological characteristics to commercial products, and an increase in bioavailability of essential amino acid content (after buccal and gastric digestion). Moreover, it was possible to obtain well-accepted products with vegetable proteins and low oil concentration. Specifically, the purchase intention demonstrated that new products could be developed with the following formulations: soy protein isolate with 22.5–30% oil or A. platensis protein concentrate with 30% oil. Further research could be conducted on using different proteins with high WHC and creamy-like color. These could enhance the taste and appearance, which were the most relevant attributes for the final acceptability of the product.