Production, characterization, and functional properties of protein concentrate from Momordica cochinchinensis seeds

Abstract

With the rising global demand for sustainable, health-conscious dietary alternatives, identifying novel plant-based protein sources remains a critical challenge. This study addresses the underutilization of Momordica cochinchinensis (Gac) seeds—an agricultural by-product—by developing the first integrated process for obtaining protein concentrate through isoelectric precipitation combined with physical separation. The research aimed to optimize the ultrasound-assisted extraction (UAE) parameters and provide a comprehensive characterization of the resulting Momordica cochinchinensis protein concentrate (MCPC). Under optimized UAE conditions (pH 11, 25% amplitude, 20 min sonication, and 1:25 material-to-solvent ratio), MCPC achieved a high protein purity of over 81%. Notably, Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) revealed a unique protein profile dominated by low-molecular-weight fractions (< 25 kDa), which correlated with significant antioxidant activity and superior functional properties compared to traditional plant proteins. MCPC demonstrated high oil-binding capacity and robust emulsifying and foaming stability, essential for diverse food matrices. Distinct physicochemical modifications observed after isolation highlight the efficiency of the UAE process in enhancing protein functionality. These findings distinguish this study by transforming a waste product into a high-value ingredient for the first time. The results underscore MCPC’s potential as a versatile, bioactive ingredient for functional foods, meat analogues, and nutritional supplements, offering a sustainable solution to the global protein transition.

Introduction

The global food industry is currently undergoing a paradigm shift toward sustainable protein sources, driven by the increasing environmental footprint of animal-based proteins and a burgeoning consumer preference for health-conscious, plant-based diets¹. Plant proteins derived from agricultural byproducts offer a compelling solution to bridge the “protein gap” while adhering to circular economy principles. Among these emerging sources, Momordica cochinchinensis Spreng (Gac) seeds—typically discarded as low-value waste after aril processing—represent a significant yet underutilized reservoir of high quality nutrients².

Momordica cochinchinensis, as known as Gac, a tropical climbing plant belonging to the Cucurbitaceae family, is widely distributed across Southeast Asia and Northeastern Australia³. While the red aril is highly valued for its oil and carotenoid content, the seeds are largely considered inedible waste in the food industry⁴. Historically, Gac seeds have been utilized in traditional Asian medicine to treat inflammatory conditions and ulcers⁵. Modern biochemical investigations have primarily focused on isolating specific bioactive peptides and small molecular weight proteins⁶, such as trypsin inhibitors (MCoTI-I and MCoTI-II⁷, MCCTI-1⁸, MCo-3 and MCo-6⁹, MCoCC-1 and McoCC-2¹⁰, and ribosome-inactivating proteins like Cochinin B¹¹, which exhibit potent neuroactive and antitumor activities¹². However, despite reports indicating that the Gac seed kernel possesses a crude protein content exceeding 40% on a dry basis², the broader application of Gac seed protein as a functional food ingredient remains critically unexplored.

A fundamental challenge in developing novel plant-based protein powders is ensuring a balance between nutritional density and techno-functional performance. Recent research has underscored that the suitability of a protein source for food systems—such as meat analogues, dairy alternatives, or functional beverages—is dictated by its amino acid profile and attributes like solubility, emulsification, and foaming stability^13,14,15. While commercial plant proteins such as soy and pea have been extensively characterized, there is a conspicuous lack of data regarding the essential amino acid composition and molecular weight distribution of the total protein fraction from Gac seeds. Furthermore, conventional isolation methods often struggle with yield efficiency and maintaining the functional integrity of the protein. In recent years, green extraction technologies, particularly ultrasound-assisted extraction (UAE), have emerged as superior alternatives to traditional alkaline soaking. UAE utilizes acoustic cavitation to disrupt complex seed matrices, enhancing mass transfer and potentially modifying protein structures to improve functionality^16,17. Recent achievements in extracting proteins from unconventional oilseeds, such as pumpkin¹⁸, watermelon¹⁹, and hemp seeds²⁰, have demonstrated that optimized UAE can significantly enhance antioxidant activity and techno-functional properties compared to traditional methods However, the systematic optimization of UAE for Momordica cochinchinensis seeds, particularly through isoelectric precipitation to produce a high-purity protein concentrate (MCPC), has not yet been reported in the literature.

Therefore, this study aims to address this scientific gap by developing an optimized protocol for producing MCPC. The objectives are to define the UAE treatments (pH, amplitude, time, and solid-to-liquid ratio) to maximize protein yield and provide a comprehensive profile of its amino acid composition, molecular weight distribution and the physicochemical and functional modifications that occur following isolation. By transforming an agricultural by-product into a value-added ingredient, this research provides a technical foundation for utilizing MCPC as a novel, bioactive protein source in the global food industry.

Materials and methods

Materials

Gac fruits were purchased from the local market in Ho Chi Minh City, Vietnam. Partially defatted Gac seed powder was prepared as follows: Firstly, the remained pulp on the seeds was manually removed. The seeds were then washed with potable water and the kernels were manually separated from the seeds. The kernels were dried using a freeze dryer (Takudo TKD-LCD500F, Viet Nam) at 40 °C for 8–10 h to a moisture content of approximately 10%. Chemical compositions of the Gac seed powder were determined and expressed as follows (% on dried basis): 2.31% ash, 54.19% lipid, 27.11% protein, 16.39% total carbohydrate. Partially defatted Gac seed powder was prepared by mechanical pressing using a screw press (NF-100, Karaerler, Turkey) operated at 120 °C. After pressing, the solid phase was separated, dried at 40 °C to a moisture content of 5% and then ground to particles with a size less than 0.5 mm. The resulting powder was stored at 4 °C until further use for protein extraction.

All chemicals and reagents were analytical grade compounds purchased from Fisher Scientific.

Preparation of Gac seed protein concentrate

Partially defatted Gac seed flour (PDS) served as the raw material for the production of Momordica cochinchinensis protein concentrate (MCPC). Protein isolation was achieved through an integrated process of ultrasound-assisted alkaline extraction followed by isoelectric precipitation²¹.

A systematic one-factor-at-a-time (OFAT) experimental design was employed to identify the optimal treatments for maximizing protein recovery. This step-wise approach allowed for the evaluation of four critical independent variables: flour-to-water ratio, extraction pH, ultrasonic amplitude, and sonication time. For each stage, the “best response” was defined as the treatment reaching the maximum protein yield while keeping the other factors at intermediate levels. Once an optimal level of the parameter was identified, it was fixed as a constant for all subsequent optimization steps to ensure technical and energy efficiency.

The optimization sequence was conducted as follows: Flour-to-water ratio was evaluated at 1:10, 1:15, 1:20, 1:25, 1:30, and 1:35 (w/v), while maintaining pH 9.0, 30% amplitude, and 15 min sonication. Extraction pH was investigated from pH 7.0 to 12.0 using the optimized llour-to-water ratio from above step. Then, ultrasonic amplitude was tested between 20% and 45% using previous optimal values and sonication time was investigated from 5 to 30 min to finalize the UAE protocol. The detail conditions for optimizing extraction process were presented in Table S2.

For each treatment, a 10 g portion of PDS (dry basis) was suspended in distilled water. UAE was performed using a high-intensity ultrasonic processor (20 kHz, 750 W, SONICS VCX 750, USA) to facilitate the release of proteins from the complex plant matrix via acoustic cavitation. Following sonication, the suspension was stirred at room temperature for 1 h to ensure equilibrium, then centrifuged at 5,000 rpm for 20 min. The protein-rich supernatant was collected, and the protein concentration was determined using the Kjeldahl method (conversion factor of 6.25). To precipitate the proteins, the supernatant pH was adjusted to the isoelectric point (pH 4.5) using 1.0 N HCl. The resulting precipitate was recovered by centrifugation (5,000 rpm, 20 min), washed three times with distilled water to remove residual salts, and finally dried at 40 °C to obtain the MCPC powder.

Physicochemical properties of MCPC and PDS

Chemical compositions

The moisture, ash, lipid, and protein contents of Gac seed powder, MCPC and PDS were determined in this study according to standard AOAC procedures²². Moisture content was measured using an infrared moisture analyzer (Ohaus MB23, China) based on the direct drying method. Protein content was determined by the Kjeldahl method using a nitrogen-to-protein conversion factor of 6.25. Lipid content was analyzed by an automatic Soxhlet extraction system (SOXTHERM, Gerhardt, Germany) based on solvent reflux extraction. Ash content was determined by incineration of the samples in a muffle furnace at 500–600 °C until constant weight was achieved. Total carbohydrate content was calculated by difference, and total sugar content was determined according to AOAC method 974.06²².

FTIR spectra of MCPC and PDS were measured using a JASCO FT/IR-6X Spectrometer with ATR PRO ONE X. The scans were acquired at a resolution of 2.0 cm^–1 (from 400 to 4000 cm^–1).

Protein fraction of PDS

The protein composition of partially defatted Gac seeds was characterized using the sequential Osborne extraction method²³, providing a systematic overview of protein distribution based on solubility. In brief, PDS was first mixed with distilled water at a 1:10 (w/v) ratio and stirred for 2 h at room temperature. The resulting suspension was centrifuged at 5000 rpm for 10 min, and the supernatant was collected as the crude albumin fraction. The remaining residue was re-suspended in 0.5 M NaCl, followed by centrifugation, and the obtained supernatant was designated as the globulin fraction. Subsequently, the residue was extracted with 70% ethanol under continuous stirring, and then centrifuged. The supernatant was identified as the prolamin fraction. Finally, the glutelin fraction was isolated from the remaining residue using 0.1 M NaOH. The protein content of each fraction was determined using the Kjeldahl method.

Amino acid compositions of MCPC

The amino acid contents of MCPC were determined using HPLC and expressed as g amino acid/100 g protein according to AOAC (994.12)²⁴. The obtained chromatogram was represented in the Supplementary information (Fig. 1S).

SDS-PAGE separation of protein compositions of MCPC

SDS-PAGE experiments were performed according to the method of Laemmli (1970) using a Mini-PROTEAN 3 Cell system (Bio-Rad, USA)²⁵. The separation was carried out by running on polyacrylamide gels (7.3 cm in height) consisting of 4% stacking gel (4% acrylamide/Bis, 0.125 M Tris, and 0.1% SDS) and 12% separating gel (12% acrylamide/Bis, 0.375 M Tris, and 0.1% SDS). Electrophoresis was performed for 1.5 h at 90 V for the first 10 min, increasing to 110 V until the separation was finished using a running buffer containing 15.1 g/L Tris, 72 g/L glycine, and 5 g/L SDS. After electrophoresis, the gels were immersed in a staining solution (0.3% w/v Coomassie Brilliant Blue, 0.25% v/v EtOH, and 7% v/v acetic acid) and agitated on a shaker for 2 h at room temperature. Then gels were washed with a washing solution composed of 40% (v/v) MeOH and 7% (v/v) acetic acid.

For SDS-PAGE analysis, the protein powder was dissolved up to a concentration of 10 mg/mL in buffer solution containing 0.032 M Tris, 0.017 M SDS, 4% (v/v) 2-mercaptoethanol, 0.1% (v/v) bromophenol blue, 51% (v/v) glycerol, and H₂O. The protein sample was ultrasonicated in an ultrasonic bath for 30 min, then denatured at 95 °C in a water bath for 10 min before centrifuging at 13,000 rpm for 10 min at room temperature. The sample was run at 10-20-25-30-35 µg on a 12% polyacrylamide gel. GangNam-STAIN™ Prestained Protein Ladder (iNtRON Biotechnology DR, South Korea) was carried out at the same time as the samples.

DPPH radical scavenging activity

The antioxidant activity was evaluated using the DPPH radical scavenging assay according to the method of Brand-Williams et al.²⁶ and Molyneux²⁷, with minor modifications. The antioxidant activity was evaluated using the DPPH radical scavenging assay. In brief, samples at various concentrations were extracted with distilled water and treated ultrasonically for 15 min. Subsequently, 1 mL of the extract was mixed with 5 mL of 0.08 mM DPPH solution prepared in water and incubated in the dark at room temperature for 30 min. The absorbance of the reaction mixture was recorded at 517 nm. A standard calibration curve was generated using Trolox (0–200 µM) (as shown in Fig. 1S), and the antioxidant activity was expressed as micromoles of Trolox equivalents (µmol TE) per gram of protein.

Determination of functional properties of MCPC and PDS

The functional properties were evaluated using the MCPC obtained under the optimal extraction conditions determined in Sect.  2.2.

Protein solubility (PS)

Protein solubility of MCPC and PDS was determined according to the method described by Lawal et al. with slight modifications²⁸. Approximately 0.125 g of sample was dispersed in 50 mL of distilled water, and the pH was adjusted in the range of 2–12 using 0.1 N HCl or 0.1 N NaOH. The suspension was stirred for 30 min at room temperature and centrifuged at 4000 rpm for 10 min. The protein content in the supernatant and total protein were measured by the Kjeldahl method. Protein solubility was calculated using Eq. (1):

$$\:Solubility\:\left(\%\right)=\frac{Protein\:content\:in\:supernatant}{Total\:protein\:content\:in\:sapmle}\times\:100$$

(1)

Water holding capacities

Water holding capacity (WHC) of MCPC and PDS was determined according to according to the method described by Y. Zhang et al., with slight modifications²⁹. Approximately 0.1 g of the powder was weighed into centrifuge tubes, and 5 mL of distilled water adjusted to pH 2–12 using HCl or NaOH (0.1 N) was added. The mixture was vortexed for 3 min, allowed to stand for 30 min, and then centrifuged at 5000 rpm for 20 min. The supernatant was decanted, and the residue was weighed. Water holding capacity was calculated according to Eq. (2):

$$\:WHC\:\left(\%\right)=\frac{Weight\:of\:water\:retained\:in\:the\:sample,\:g}{Initial\:weight\:sample,g}\times\:100$$

(2)

Oil binding capacities

Oil binding capacity (OBC) of MCPC and PDS was determined according to the method described by Y. Zhang et al., with slight modifications²⁹. Approximately 0.1 g of the sample powder was placed in a centrifuge tube and mixed with 5 mL of vegetable oil. The mixture was vortexed for 3 min, allowed to stand for 30 min, and subsequently centrifuged at 5000 rpm for 20 min. After centrifugation, the supernatant was carefully removed, and the remaining residue was weighed. The oil binding capacities was then determined using Eq. (3):

$$\:OBC\:\left(\%\right)=\frac{Weight\:of\:oil\:retained\:in\:the\:sample}{Initial\:sample}\times\:100$$

(3)

Emulsifying capacity

Emulsifying capacity (EC) of MCPC and PDS was determined according to the method described by Yasumatsu et al.³⁰. Approximately 0.7 g of the material was weighed into centrifuge tubes, and 10 mL of distilled water adjusted to pH 2–12 using 1 N NaOH or 1 N HCl was added. Subsequently, 10 mL of refined soybean oil was added, and the mixture was homogenized at 5,000 rpm for 20 min to form an emulsion. Emulsifying capacity was calculated using Eq. (4):

$$\:EC\:\left(\%\right)=\frac{Ht}{He}\times\:100$$

(4)

where, Ht = height of the emulsified layer after centrifugation (mm), and He = total height of the mixture in the tube (mm).

Foaming capacity and stability

Foaming capacity (FC) and foam stability (FS) of MCPC and PDS were determined according to the method described by Deng et al. with slight modifications³¹. Approximately 1 g of the powder was dispersed in 50 mL of distilled water, and the pH was adjusted from 2 to 12. The suspension was homogenized for 3 min using a high-speed homogenizer and then quickly transferred into a 100 mL graduated cylinder to record the volume before and after homogenization. Foaming capacity was calculated immediately at 0 min, while foam stability was determined after 120 min at 25 °C according to the Eqs. (5) and (6):

$$\:FC\:\left(\%\right)=\frac{V-{V}_{0}}{{V}_{0}}\times\:100$$

(5)

$$\:FS\:\left(\%\right)=\frac{{V}_{2}}{{V}_{0}}\times\:100$$

(6)

where, V₀ = initial liquid volume (mL), V = volume immediately after homogenization (mL), and V₂= volume after 120 min (mL).

Gelation ability

The minimum gelation concentration of MCPC and PDS was determined following the method described by Sathe et al.³², with slight modifications. Sample solutions were prepared at protein concentrations of 2%, 4%, 6%, 8%, 10%, 12%, 14%, 16%, 18%, and 20% (w/v) in 5 mL of distilled water. The test tubes containing the protein solutions were heated in a boiling water bath at 95–100 °C for 30–60 min, then immediately cooled under running tap water, followed by further cooling at 4 °C for 2 h. The minimum gelation concentration was defined as the lowest protein concentration at which the sample did not flow or slide when the test tube was inverted, indicating the formation of a self-supporting gel. This value was determined experimentally by observing the gel strength and flow behavior of each sample upon inversion.

Statistical analysis

All experiments were conducted in triplicate, and the results are expressed as mean ± standard deviation. Microsoft Office Excel 2019 was used for data processing and graphical presentation. Analysis of variance (ANOVA) was performed using Minitab Statistical Software 22 to evaluate differences among treatments, and Tukey’s multiple comparison test was applied for pairwise comparisons at a 95% confidence level.

Results and discussion

Protein fractions of PDS

The fractionation of proteins in PDS was conducted to establish a molecular basis for the isolation process. The protein components of Gac seeds were isolated and quantified following the sequential extraction procedure based on their solubility in different solvent systems. As shown in the fractionation results, the proteins were isolated in the following order: Albumin (water-soluble) and Globulin (salt-soluble) accounted for 12.41% and 27.68% of the total protein content, respectively. Prolamin (ethanol-soluble) represented a minor fraction at 9.66%, while 2.17% of the protein remained non-extractable. Notably, Glutelin, which is soluble in dilute alkaline solutions (NaOH-soluble), was identified as the predominant protein fraction, accounting for 48.08% of the total protein content. These results indicate that glutelin and globulin serve as the primary storage proteins in Gac seeds. Interestingly, the present findings contrast with those reported for another Momordica species, Momordica charantia, where albumin was identified as the dominant fraction (49.3%) and prolamin was absent³³. This distinct protein profile underscores the necessity of a species-specific approach for protein recovery. Specifically, the predominance of alkaline-soluble glutelin in Momordica cochinchinensis provides a clear scientific rationale for the selection of ultrasound-assisted alkaline extraction as the most effective methodology to maximize yield in this study. Furthermore, understanding these specific fractions is essential for predicting the techno-functional behavior of the resulting protein concentrate in potential food applications.

Ultrasound-assisted extraction of protein from partially defatted Gac seed

Effect of solid-to-solvent ratio on protein extraction yield

The effect of solid-to-solvent ratio on extraction yield was evaluated at five levels under fixed conditions (pH = 9; ultrasound amplitude = 30% and ultrasound time: 15 min) and presented in Fig. 1.

Fig. 1

Effect of solid-to-solvent ratio on protein extraction yield. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

As illustrated in Fig. 1, the extraction yield increased progressively with higher solvent-to-material ratios, rising from 1:10 to 1:25 and reaching a maximum of 48.42%, which was significantly higher (p < 0.05) than those obtained at lower ratios. This enhancement can be attributed to improved diffusion of proteins into the solvent, reduced protein re-precipitation in a more diluted medium, and more efficient propagation of ultrasonic waves^34,35. However, a further increase in the solvent ratio to 1:30 and 1:35 resulted in a slight decline in yield, likely due to excessive dilution, which decreased ultrasonic energy density and cavitation efficiency³⁵. Similar observations have been reported in studies on protein extraction from other sources. Nguyen et al. found that protein yield from defatted peanut (Arachis hypogaea L.) increased up to a 1:25 ratio before stabilizing³⁶, while Dong et al. reported notable improvements in protein extraction from rapeseed (Brassica napus L.) as the solvent ratio increased from 1:10 to 1:25³⁷. Based on these findings, a material-to-solvent ratio of 1:25 was identified as the optimal condition for protein extraction in this study.

Effect of pH on protein extraction yield

The variation in protein extraction yield as a function of solution pH was investigated using the optimized solid-to-solvent ratio identified in the previous step. Specifically, the conditions were fixed at a solid-to-solvent ratio of 1:25, while the ultrasound amplitude and time were maintained at intermediate levels (30% and 15 min, respectively) based on preliminary range-finding trials. This systematic approach ensures that the effect of pH is evaluated under a stabilized material-to-liquid environment (Fig. 2).

Fig. 2

Effect of pH on protein extraction yield. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

The extraction yield increased steadily as the pH rose from 7 to 11, reaching a maximum of 76.46% at pH 11, with significant differences observed among treatments (p < 0.05). This enhancement is likely due to improved protein solubility under alkaline conditions, where a greater deviation from the isoelectric point (pI) enhances electrostatic repulsion, promotes protein unfolding, and disrupts associations with polysaccharides and lipids, thereby facilitating protein release^38,39. Beyond this point, at pH 12, no significant improvement was observed compared to pH 11, which may be attributed to excessive denaturation or partial hydrolysis leading to reduced solubility and protein aggregation³⁹. Similar pH-dependent patterns have been reported for other oilseeds. Celik et al. observed that protein extraction from defatted sour cherry seeds increased with pH, reaching a maximum yield of 76.39% at pH 12, although pH 10 was identified as optimal to balance extraction efficiency and structural preservation⁴⁰. Therefore, pH 11 was determined to be the optimal condition for protein extraction from Gac seeds. At this level, the process achieved maximum recovery efficiency while maintaining the structural stability of the proteins. This stability was confirmed by the retention of characteristic peptide backbone vibrations in the FTIR spectra and the preservation of distinct molecular weight distributions in SDS-PAGE profiles, indicating that the alkaline environment did not induce significant protein hydrolysis.

Effect of ultrasound conditions on protein extraction yield

The effect of ultrasound amplitude and ultrasound time on extraction yield was evaluated and presented in Fig. 3.

Fig. 3

Effect of ultrasound amplitude (A) and ultrasound time (B) on protein extraction yield. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

Figure 3A illustrates the effect of ultrasound amplitude on protein extraction yield. Increasing the amplitude from 20% to 25% significantly enhanced the yield from 43.82% to 75.49% (p < 0.05), attributed to intensified cavitation and cell wall disruption that facilitated protein release. At 30% amplitude, the yield reached a plateau, indicating saturation of the extraction process. This trend aligns with the observations of Villalón-López et al., who reported a similar threshold effect when ultrasound energy density increased from 132 to 150 kJ/L⁴¹. Beyond 30%, further increases in amplitude (30–45%) led to a pronounced decrease in yield, likely resulting from excessive cavitation that generated intense shock waves, shear forces, and localized heating, promoting protein denaturation and fragmentation⁴². Moreover, the formation of excessive cavitation bubbles may have hindered efficient energy transfer and elevated the bulk temperature, thereby diminishing extraction efficiency⁴³. Comparable findings were reported by Sert et al., who observed that moderate ultrasound amplitude (50%) enhanced protein recovery from pumpkin seed press cake, while excessive amplitude levels induced denaturation and yield reduction⁴⁴. Since no significant difference was detected between 25% and 30% amplitudes, 25% amplitude was selected as the optimal condition for subsequent experiments, ensuring high extraction efficiency while minimizing energy consumption and the potential for protein degradation.

As shown in Fig. 3B, increasing the sonication time from 5 to 20 min significantly enhanced the protein extraction yield, reaching a maximum of 79.37% at 20 min (p < 0.05). This improvement can be attributed to intensified cavitation effects, where repeated bubble formation and collapse generate strong shear forces that disrupt Gac seed cell walls, thereby facilitating protein release and improving solubilization in the alkaline medium⁴¹. However, extending the sonication time beyond 20 min (20–30 min) resulted in a notable decline in yield. Prolonged sonication may lead to the formation of free radicals, structural degradation, or protein aggregation due to the exposure of hydrophobic regions following protein unfolding, resulting in insoluble complexes^16,45. Similar behavior has been reported in other plant sources. Fatima et al. observed that ultrasound-assisted protein extraction from Moringa leaves increased up to 20 min before declining with longer treatment times¹⁶. Likewise, Zhang et al. reported that protein yield from Dictyophora rubrovolvata mushroom caps rose between 10 and 25 min, then decreased beyond 25 min due to protein aggregation¹⁷. Therefore, a sonication time of 20 min was identified as the optimal condition for protein extraction from Gac seeds, providing high recovery efficiency while minimizing the risk of protein denaturation and aggregation.

Physicochemical properties

Chemical composition of MCPC

The proximate compositions of PDS and MCPC were analyzed and are presented in Table 1.

Table 1 Chemical composition of PDS and MCPC.

The analytical results revealed that the PDS contained up to 44.09 g of protein per 100 g, comparable to other oilseeds known for their high protein content following oil removal. When compared with common oilseed residues, Gac seed residue exhibited a protein content equivalent to or exceeding that of soybean meal (25.7–52.4%), peanut meal (40–50%), and sunflower meal (29–43.4%)^46,47. These findings suggest that Gac seed residue is not merely a by-product but a valuable protein-rich resource. The high protein content of both PDS and MCPC demonstrates the feasibility of extracting protein for industrial applications. Thus, Gac seed can be considered a promising novel source of plant-based protein.

Comprehensive characterization of the defatted seed and its protein concentrate further revealed notable alterations in physicochemical and functional properties following protein isolation.

Amino acid compositions

The amino acid composition plays a crucial role in determining the biological and nutritional properties of proteins. The amino acid profiles of albumin and glutelin are presented in Table 2.

Table 2 Amino acid compositions of MCPC.

Analysis revealed that Gac seed protein contains both essential (EAAs) and non-essential amino acids (NEAAs), with glutamic acid (141.9 mg/g) and aspartic acid (72.4 mg/g) being the most abundant⁴⁸. Arginine was also present in high concentration (103.2 mg/g), suggesting potential health-promoting benefits⁴⁹. Among the essential amino acids, leucine (67.9 mg/g) and valine (39.0 mg/g) were relatively abundant, while isoleucine (37.8 mg/g) and lysine (27.3 mg/g) were identified as limiting amino acids. This composition aligns with previous reports indicating that lysine, methionine, and cysteine are typically the most restricted amino acids in plant-derived proteins⁵⁰. Comparable amino acid distribution patterns have been observed in various plant protein sources, where methionine and lysine commonly serve as the first limiting amino acids.

Protein molecular characterization

The protein molecular characterization of MCPC was determined using SDS-PAGE analysis. The obtained results were presented in Fig. 4.

Fig. 4

SDS-PAGE profile of Momordica cochinchinensis seed protein concentrate (MCPC). Lane M: Molecular weight marker; Lanes 1–5: MCPC loaded at concentrations of 10, 20, 25, 30, and 35 µg, respectively.

SDS-PAGE analysis showed that the Gac seed protein bands were mainly distributed in the range of 20 to 45 kDa, with the most abundant fraction below 25 kDa. This range corresponds to previously reported trypsin inhibitors such as MCCTI-1⁸, MCoCC-1, MCoCC-2¹⁰, MCoTI-I and MCoTI-II⁷, MCo-3 to MCo-6⁹. Additional bands at 25–30 kDa were consistent with the ribosome-inactivating protein Cochinin B (28 kDa) described by Chuethong et al. (2007)⁹. Furthermore, protein fractions in the range of 30–40 kDa were also detected. According to the study of Balachandran et al., it was mentioned that most of the wheat sieve fluid protein has a weight in the range of 10–40 kDa, which is also consistent with the study of the protein mass of Gac seeds⁵¹.

DPPH radical scavenging activity

DPPH Radical Scavenging Activity of MCPC was measured and expressed as Trolox equivalent antioxidant capacity (TEAC) and presented in Table 3.

Table 3 Antioxidant capacity of MCPC.

As shown in Table 3, TEAC values increased with protein concentration, ranging from 9.05 ± 0.20 µg Trolox/mL at 50 µg/mL to 23.63 ± 0.19 µg Trolox/mL at 200 µg/mL, demonstrating a clear dose-dependent relationship. The progressive increase in TEAC values indicates that MCPC exhibits measurable antioxidant potential, likely attributable to its amino acid composition—particularly the presence of tyrosine, tryptophan, methionine, and cysteine residues capable of donating electrons and scavenging free radicals⁵². Although the TEAC values obtained were lower than those reported for hydrolyzed soybean protein (~ 30–40 µmol Trolox/g protein)⁵³ and sunflower seed protein (~ 28 µmol Trolox/g)⁵⁴, the observed concentration-dependent trend aligns with findings from other plant-derived antioxidant peptides, such as those from mung bean protein hydrolysates⁵⁵.

Overall, these results suggest that MCPC possesses moderate but notable antioxidant activity relative to commonly studied plant proteins. Given that Gac seeds are typically underutilized byproducts, their exploitation as a protein source with antioxidant potential presents a valuable opportunity for incorporation into functional foods and nutraceutical formulations.

FTIR spectroscopic analysis of MCPC

The FTIR spectrum of Momordica cochinchinensis protein concentrate (MCPC) was analyzed to evaluate the impact of the isolation process (UAE and isoelectric precipitation) on the chemical structure and functional groups. The spectrum exhibited characteristic absorption bands typical of proteinaceous materials, confirming the high efficiency of the isolation treatments.

Fig. 5

FTIR spectra of MCPC and PDS.

As shown in Fig. 5, a broad, intense peak was observed at approximately 3277 cm^-1, corresponding to the amide A band (N-H stretching vibration coupled with hydrogen bonding). The breadth of this peak suggests an extensive network of hydrogen bonds within the protein matrix, which is crucial for structural stability. The small shoulder near 2925–2960 cm^-1 (C-H stretching) indicates the presence of aliphatic side chains of amino acids. A prominent peak at 1618–1645 cm⁻¹ represents the Amide I band, primarily originating from C = O stretching vibrations¹⁶. This band is a sensitive indicator of protein secondary structure. The peak position suggests a significant presence of β-sheet and random coil structures, likely induced by the acoustic cavitation during ultrasound-assisted extraction and subsequent pH-induced precipitation. The band at 1500–1545 cm^-1 (Amide II) arises from N-H bending and C-N stretching vibrations¹⁶. The intensity and position of this band in MCPC, compared to raw seed flour, indicate that the isolation variables (pH 11 and sonication) promoted the unfolding of the protein globular structure, exposing more functional groups and potentially enhancing surface hydrophobicity. The absorption at 1230–1245 cm^-1 (Amide III) further confirms the complex vibrational modes of the peptide backbone¹⁷. A peak near 1050–1070 cm^-1 may be attributed to C-O stretching, possibly related to residual carbohydrates or glycosylated protein fractions remaining in the concentrate. The mechanical shear forces generated by sonication led to the partial denaturation of the protein. This is evidenced by the shift in the Amide I and II peaks toward lower wavenumbers compared to native seed proteins, reflecting a transition from more ordered (α-helix) to more flexible (β-sheet/random coil) configurations. The exposure to alkaline pH (pH 11) followed by rapid acidification to the isoelectric point (pH 4.5) significantly altered the electrostatic environment, promoting protein-protein interactions and the formation of the concentrate.

Determination of functional properties of MCPC

Protein solubility

Solubility is a crucial functional characteristic of proteins, as it strongly influences their water-holding capacity, emulsifying properties, and gelling behavior, thereby determining their suitability for various applications. The protein solubility of PDS and MCPC as a function of pH was investigated, and the results are presented in Fig. 6.

Fig. 6

Effect of pH on protein solubility of PDS and MCPC. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

As shown in Fig. 6, the protein solubility of PDS and MCPC varied significantly with pH (p < 0.05). The lowest solubility was recorded at pH 4 (20.98%), corresponding to the isoelectric region, whereas solubility increased under both acidic (57.84% at pH 2) and alkaline conditions, reaching a maximum of 74.46% at pH 12. Across all pH levels, MCPC consistently exhibited higher solubility than PDS.

The minimum solubility at pH 4 reflects protein aggregation at the isoelectric point⁵⁶. As the pH deviates from this point, increased ionization of charged groups enhances electrostatic repulsion and exposes hydrophilic residues, thereby improving solubility⁵⁷. The rise in solubility at pH 2 is consistent with observations in soybean protein, where strong positive surface charges promote protein–water interactions. Similar pH-dependent trends have been reported for walnut⁵⁸, almond, and rambutan seed proteins^59,60, indicating that Gac seed protein follows a common plant protein behavior—exhibiting minimal solubility near the isoelectric point and enhanced solubility under both acidic and alkaline conditions.

Water holding capacity and oil adsorption capacity

Water-holding capacity (WHC) is a crucial functional property of proteins, as it strongly affects the texture, viscosity, and mouthfeel of various food products such as soups. Figure 7 illustrates the influence of pH on the WHC of Gac MCPC and PDS.

Fig. 7

Water holding capacities of PDS and MCPC. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

The WHC decreased from 369.42% at pH 2 to 188.06% at pH 4, then increased progressively, reaching 495.63% at pH 12 (p < 0.05) (Fig. 7). The lowest WHC observed at pH 4 corresponds to the protein’s isoelectric point, where diminished electrostatic repulsion favors protein aggregation and reduces the availability of water-binding sites. Under alkaline conditions, protein unfolding and the ionization of carboxyl groups expose additional hydrophilic sites, thereby enhancing water retention⁶¹. Similar pH-dependent behavior has been observed for mung bean protein isolate⁶². These findings highlight the potential of Gac seed protein for use in bakery, processed meat, and plant-based meat products that require high water-holding capacity⁶³.

Oil-binding capacity (OBC) is a key functional property that reflects the interaction between proteins and lipids, primarily influenced by the presence of hydrophobic amino acid residues and the protein’s ability to form a structural matrix that entraps oil. The results revealed that MCPC exhibited a significantly higher OBC (166.23%) compared to PDS (155.01%). This improvement can be attributed to the purification process, which eliminated non-protein impurities and increased the exposure of hydrophobic groups on the protein surface, thereby enhancing the affinity and interaction with oil molecules⁶⁰. The higher OBC observed in MCPC suggests its strong potential for use in lipid-rich food systems such as sausages, baked goods, processed meats, and meat analogs, where proteins function as flavor carriers and contribute to texture enhancement. Therefore, MCPC can serve not only as a valuable nutritional protein source but also as a functional ingredient, providing an effective means of valorizing Gac seed by-products that are typically underutilized. The oil-binding capacity of MCPC can be explained by its amino acid composition, particularly the presence of hydrophobic residues such as leucine, isoleucine, valine, phenylalanine, alanine, proline, and methionine. These hydrophobic amino acids promote protein–lipid interactions through nonpolar side chains, thereby enhancing oil retention. In addition, the predominance of glutelin, a protein fraction typically enriched in hydrophobic amino acids, further contributes to the high OBC observed in MCPC.

Emulsifying capacity

The emulsifying capacity of MCPC was evaluated under varying pH conditions and compared with that of PDGS, as presented in Fig. 8.

Fig. 8

Emulsifying capacity of PDS and MCPC. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

As shown in Fig. 8, the emulsifying capacity reached its minimum at pH 4 (25.93%) and increased significantly under both acidic (pH 2) and alkaline conditions, attaining a maximum of 73.35% at pH 12 (α = 0.05). The reduced emulsifying activity near pH 4 corresponds to the isoelectric point, where diminished surface charge promotes protein aggregation and restricts adsorption at the oil–water interface⁶⁴. In contrast, at alkaline pH, increased negative surface charges enhance electrostatic repulsion and induce protein unfolding, thereby improving interfacial coverage and stabilizing emulsion formation^65,66. These results demonstrate the potential of Gac seed protein as a natural emulsifying agent for use in acidic or alkaline food systems, such as beverages and sauces.

Foaming capacity and stability

The foaming capacity and stability of MCPC and PDS at different pH levels are shown in Fig. 9.

Fig. 9

Foaming capacity (a) and foaming stability (b) of MCPC and PDS. Different letters denote significant differences between treatment levels at 95% confidence level (p < 0.05).

The lowest foaming capacity (17%) was recorded at pH 4, corresponding to the isoelectric point (pI), where diminished electrostatic repulsion promotes protein aggregation and restricts their migration to the air–water interface⁶⁷. In contrast, foaming capacity increased markedly under both acidic (pH 2) and alkaline (pH 8–12) conditions, likely due to enhanced net surface charges, improved solubility, and partial protein unfolding that exposes hydrophobic regions, facilitating adsorption and film formation at the interface⁶⁸. The highest foaming capacity (175%) was observed at pH 12, suggesting that strong alkaline conditions promote extensive protein denaturation and the exposure of hydrophobic and sulfhydryl groups, leading to the formation of flexible, cohesive protein films that stabilize the foam structure⁶⁹.

Foam stability is an essential functional attribute that influences both the sensory characteristics and technological performance of foamed food products. As shown in Fig. 9b, the foam stability of Gac seed protein varied significantly with pH (p < 0.05). A pronounced decline was observed from pH 2 (60.5%) to pH 4 (3.3%), corresponding to the isoelectric point, where reduced electrostatic repulsion facilitates protein aggregation and results in weak interfacial films. Beyond this point, foam stability increased markedly with rising pH, reaching a maximum of 76.03% at pH 12. This improvement can be attributed to enhanced protein charge, better dispersion, and the formation of more cohesive interfacial films stabilized by hydrophobic interactions and hydrogen bonding⁷⁰. These observations are consistent with the findings of Khalid et al., who reported similar pH-dependent behavior in cowpea protein isolates⁷¹.

Gelation ability

Gelation ability is a critical functional property of proteins, representing their capacity to associate and form a three-dimensional network structure upon exposure to heat or environmental stimuli. This property not only determines the suitability of proteins for various food applications but also directly influences their sensory characteristics and functional value. In this study, the gelation ability of protein extracted from Gac seeds was evaluated (as shown in Fig. S3) and compared with that PDS (as shown in Fig. S4) to assess its potential as a novel functional ingredient. The results revealed a clear distinction between the two samples. Gac seed protein demonstrated pronounced gel formation at 2%, while the PDS required a higher concentration (6%) and exhibited only weak. This indicates that MCPC possesses a strong capacity to form intermolecular networks, suggesting robust interactions between protein molecules under thermal treatment. The superior gelation performance of MCPC at low concentrations highlights its potential applicability in the food industry, particularly in products that require a stable gel matrix, such as sausages, surimi, and plant-based meat analogs.

Conclusion

This study establishes a technical framework for the valorization of Momordica cochinchinensis (Gac) seeds—an underutilized agricultural by-product—by developing an optimized ultrasound-assisted extraction and isoelectric precipitation protocol. The research successfully identified the optimal processing parameters (solid-to-solvent ratio of 1:25, pH 11, 25% amplitude, and 20 min sonication) required to maximize protein recovery while preserving structural integrity. The resulting Gac seed protein concentrate (MCPC), with a purity exceeding 81%, is characterized by a unique protein profile dominated by glutelins and globulins, and a significant abundance of low-molecular-weight fractions (< 25 kDa). Techno-functional evaluations demonstrate that MCPC possesses superior oil-binding capacity and emulsifying activity, highlighting its high potential as a stabilizing agent in lipid-rich and emulsion-based food systems, such as meat analogues and functional dressings. While the pH-dependent solubility profiles broaden its applicability in alkaline-processed formulations, its moderate foaming capacity suggests that its use in highly aerated products may require further modification. Future studies should prioritize the determination of biological indices, such as the Protein Efficiency Ratio (PER) and in vitro protein digestibility, to fully validate the nutritional quality of MCPC for high-value supplements. Assessing the performance of MCPC within complex food matrices will be essential to understand the sensory and textural contributions of this novel plant-based ingredient. In summary, this work provides a scientific foundation for the sustainable transformation of Gac seed waste into a functional, bioactive protein ingredient, contributing to the circular economy and the global shift toward plant-based nutrition.