Bioprospecting microwave-alkaline hydrolysate cocktail of defatted soybean meal and jackfruit peel biomass as carrier additive of molasses-alginate-bead biofertilizer

Optimization of microwave-alkaline hydrolysis

The nutrient analysis of both DSM and JP, as well as molasses, were broken down into several main nutritional components as shown in Table 1. More than 50% nutrient fraction of DSM consisted of protein while more than two-thirds of JP comprised of carbohydrates. Both substrates are excellent nitrogen and carbon sources respectively which fit the nutrient requirement for strain 40a proliferation. In addition, they were rich in trace minerals (Supplementary information) especially phosphorus, potassium, calcium, magnesium, and sodium. To obtain a homogenized liquid formulation, both DSM and JP were hydrolyzed concurrently through the MAH technique to extract a high amount of soluble proteins and sugars. Previous reports have successfully demonstrated the same MAH technique to extract crude proteins from chicken feathers^32,33. To the best of our knowledge, the extraction of soluble protein and sugar from the DSM-JP cocktail has yet to be reported particularly through MAH treatment.

The parameters described by Cheong et al.³², were adopted in the present central composite design (CCD) to achieve the optimal protein and sugar concentration of hydrolysate. The experimental setup and the results of CCD were presented in Table 2. It was observed from the table that the protein concentration of hydrolysate varied from 1.22 (run 10) to 1.92 mg/mL (run 7) while the sugar concentration ranged from 3.02 (run 3) to 4.88 mg/mL (run 7). The regression analysis of the full second-order polynomial model was shown in Table 3. The high F-value (protein = 82.2, sugar = 64.5) of both models suggested that they were exceptionally significant that the probability of noise interference was only 0.01%. The relatively high value of the determination coefficient, R² (protein = 0.9833, sugar = 0.9788) revealed that both models fitted to the experimental results excellently which provides a good estimation of protein and sugar concentration (P < 0.05).

In both cases, all linear and quadratic functions i.e., A, B, A², and B² were observed as significant model terms (P < 0.05), implying that both NaOH concentration and residence time played significant roles in determining the protein and sugar concentration of hydrolysate in MAH protocol. Likewise, the AB combination showed a significant interaction (P < 0.05) for both responses, thus verifying the importance of optimizing both parameters, A and B simultaneously. The polynomial models for protein (Y_P) and sugar (Y_S) concentration for the MAH cocktail of DSM and JP were regressed by considering the significant terms and thereafter expressed in Eqs. (1) and (2).

where Y_P and Y_S were the coded terms for protein and sugar concentration of hydrolysate, respectively. The regression models (Eqs. (1) and (2)) were used to predict the range of protein and sugar concentration dynamics for various levels of the selected variables. Consistent with the model equations, the maximum protein and sugar concentration was demonstrated by the contour and 3D plots shown in Fig. 1 by varying the parameter levels that were priorly determined by the CCD.

Post-analysis of quadratic models

To validate the model adequacy and investigate the reproducibility of the generated quadratic models of the CCD, the MAH procedure was recreated using the optimized parameters with a high desirability value (1.00) to achieve hydrolysate with maximum soluble protein and sugar concentration (Fig. 2). The results were interpreted by comparing the actual data and the predicted values of the previous quadratic models as illustrated in Fig. 2b. The figure shows that the protein and sugar concentration of the actual data was 5.31 and 8.07 mg/mL, respectively. These results accord with the predicted value of protein and sugar concentration i.e., 5.23 and 8.05 mg/mL, respectively, with no significant difference (P > 0.05) observed than the actual value. Interestingly, both responses showed slightly higher concentration up to 1.5% than the predicted value.

These results provide further support that the newly developed model was accurate and reliable to achieve maximum yield of protein and sugar extraction from the DSM-JP cocktail hydrolysis. Therefore, the optimized parameters for the MAH treatment of 2% (w/v) DSM-JP cocktail were finalized at NaOH concentration 0.084 M (100 mL), 8.7 min residence time at 300 W microwave power-level. The hydrolysate produced under the optimized parameters was used for the subsequent experiments.

Soluble protein and total sugar recovery

In general, DSM consists of 48–54% crude protein which largely constitutes essential amino acids such as arginine, methionine, and lysine^20,34. The protein constituents in DSM vary and mainly depend upon the processing at the crushing plant and removal of the hulls³⁵. In the present study, the total protein content of DSM and JP combined was 0.296 g/g dry mass (Table 4). With the total soluble protein yield of 531.14 mg/100 mL, the optimized MAH protocol achieved a high protein recovery of 0.896 g/g protein and 0.265 g/g dry mass. As opposed to a previous report, the protein extraction via microwave irradiation alone with water as a solvent yielded only 0.782 g/g protein even with a high (10%) biomass loading concentration³⁶. The protein recovery per dry mass DSM-JP by the optimized MAH conditions was also superior to hydrothermal (water-based) treatment as reported by Watchararuji et al.³⁷ who obtained a maximum of 0.205 g/g DSM.

The observed improvement of protein recovery in the present study might be attributed to the use of NaOH solvent in the MAH treatment. Since NaOH is more polar than water, it is more effective at dissolving complex protein molecules in the biomass. For instance, Lu et al.³⁸ managed to obtain a maximum 0.723 g/g protein recovery from microalgal residue when a concentrated 7.9% NaOH solvent was used in the extraction. In addition, the MAH technique used in this investigation employed a low solid to liquid ratio (1:50 w/v). This allowed a great driving force of mass transfer during hydrolysis in which hindered biomass clumping, thus enhanced biomass dispersion, and improved protein and sugar extraction³⁸. These results corroborate the findings of Cheong et al.³², where they successfully achieved more than 70% protein recovery using a low ratio of chicken feather biomass to NaOH (1:50 w/v) in the microwave-savinase hydrolysis experiments.

In terms of sugar extraction, the total sugar content of DSM-JP was almost double the protein content, which was 0.535 g/g dry mass (Table 4). With the total soluble sugar yield of 806.52 mg/100 mL, the optimized MAH protocol exhibited a high sugar recovery of 0.754 g/g sugar and 0.403 g/g dry mass. The high sugar content in the hydrolysate might be contributed by the degradation of various complex carbohydrate macromolecules in the JP biomass i.e., pectin—7.52% (heteropolysaccharide), cellulose—27.75%, and starch—4.12% (homopolysaccharide)³⁹. Pectin is degraded into rhamnose, uronic acids, and neutral sugars e.g., d-galactose, d-glucose, or l-arabinose in alkaline pH through alkaline demethoxylation (saponification) and depolymerization (β-elimination) particularly when assisted with heat radiation⁴⁰. Additionally, β-elimination produces unsaturated uronides, which leads to the occurrence of non-enzymatic browning⁴¹. These mechanisms might explain the colour transformation of the DSM-JP hydrolysate solution from dull yellowish-brown into dark brownish red during the MAH process.

Apart from the other popular physicochemical treatments used in many studies to accomplish successful protein and sugar hydrolysis, such as hydrothermal, acid, and oxidative treatments, the alkaline treatment also facilitates access to the most recalcitrant structures of polysaccharides and their reactivity. On the one hand, they can remove hemicellulose, pectin, proteins, and extracts, while also shortening cellulose fibers and increasing crystallinity⁴². Most of all, the microwave heat irradiation used in this investigation, loosened the cell wall matrix, causing parenchymal cells to sever, resulting in skin tissue opening⁴³. As a result, the NaOH solvent was able to permeate the skin tissues which led to greater solvent-tissue contact, hence improving the extraction efficacy.

Alterations of the chemical structure of raw and hydrolyzed DSM-JP

This section delves deeper into the structural and chemical changes made to the DSM-JP biomass and hydrolysate produced. An FTIR spectroscopy analysis was performed on the raw, residue, and hydrolysate of DSM-JP as illustrated in Fig. 3a. In comparison to the DSM-JP residue, the raw biomass had a low transmittance, indicating the structural rigor of the original biomass. Strong bonds, such as hydrogen bonds, disulfide bonds, and peptide bonds, reinforce the protein and carbohydrate structure of the raw DSM-JP. Since the raw DSM-JP is densely packed, it obstructs light transmission in the FTIR analysis. The high transmittance of the residue and hydrolysate, on the other hand, indicates that the treatments had caused severe structural damage to the original biomass³².

The strengths of the vibrating bonds and the masses of the vibrating atoms were shown to have the greatest impact on the vibrational frequencies of the bonds⁴⁴. According to the protein conformational reports^45,46, the common band at 1600–1700 cm⁻¹ was frequently used as a protein indicator, which was associated with amide I presence (primarily C=O stretching vibration of the polypeptide backbone, combined with C–N stretching, C–C–N deformation, and N–H bending modes in plants). As shown in the FTIR spectra, the peak of amide I in DSM-JP residue and hydrolysate was shifted from 1635 to 1641 cm⁻¹ and 1642 cm⁻¹, respectively, and the intensity increased from 6.45 to 19.19% and 19.79%, respectively, indicating an increase in the length of the peptide dipole moment and demonstrating polypeptide backbone motion exerted by the MAH treatment⁴⁷.

This finding is consistent with that of Tian et al.⁴⁸ who reported a similar band shift of the aqueous layer of soybean protein hydrolysate (ASPH) spectrum after an extended duration of ultrasound treatments. Further characterization revealed that the secondary structures of the ASPH were significantly altered where the β-sheet content was increased but the α-helix and β-turn content was reduced as compared to the control. This might explain the shift in the amide I absorptions in this study since this frequency region is a sensitive protein fingerprint to indicate changes of the protein secondary structures e.g., α-helix, β-turn, β-sheet, and random coil⁴⁹. The breakdown of hydrogen bonds inside peptide molecules may be responsible for the alterations in protein aggregates of DSM-JP. According to Guzmán-Ortiz et al.⁵⁰, high-temperature treatments trigger alterations in secondary, tertiary, and quaternary protein structure, exposing polypeptides and open internal peptide bonds, making them more easily hydrolyzed.

Similarly, both amide II (approx. 1500–1600 cm⁻¹) and amide III bands (approx. 1100–1500 cm⁻¹), which correlate to N–H bending and chemical groups in protein side chains^51,52, exhibited frequency alterations as compared to the raw DSM-JP. The amide II band of the raw DSM-JP was detected at 1539 cm⁻¹, but the corresponding bands were not detectable for both residue and hydrolysate. Whereas the amide III band for both residue and hydrolysate showed medium absorptions at 1275 cm⁻¹ and 1282 cm⁻¹, respectively, which notably shifted from 1242 cm⁻¹ recorded in the raw DSM-JP spectrum. These differences reflected a modification in the secondary structure of DSM-JP proteins, which concur with the frequency vibrations demonstrated in the amide I band resulted from the MAH treatment.

The band in the high-energy region is commonly attributed to a large abundance of O–H groups in carbohydrates and lignin⁵³. The most prominent bands at 3000–3500 cm⁻¹, which correspond to hydrogen-bonded O–H stretching of hydroxyl groups deriving predominantly from cellulose and hemicelluloses, were observed in all spectra^39,54. The broad, strong peak of the raw DSM-JP at 3345 cm⁻¹ was observed shifted to 3400 cm⁻¹ for both DSM-JP residue and hydrolysate. In addition, the presence of carbonyl bands at 1630–1650 cm⁻¹ and 1740–1760 cm⁻¹ respectively, in the raw DSM-JP, indicate the presence of free and esterified carboxyl groups prior to hydrolysis⁵⁵. The absence of shoulder peak at 1744 cm⁻¹ after the MAH treatment suggests that the acetyl and uronic ester groups of hemicelluloses, as well as the ester bond of lignin, might be broken down during the process. This is confirmed by the absence of absorption bands at 1455 and 1539 cm⁻¹ which indicates the lack of C=C aromatic ring of lignin²⁵. Therefore, the breakdown of non-soluble polysaccharides of DSM-JP could be a major factor, causing the release of abundant soluble sugars in the hydrolysate produced.

The free amino acid and sugar profiles of the hydrolysate are presented in Fig. 3b, c, respectively. The composition of each amino acid ranged from the highest at 1.10 mg/mL (proline) to the lowest at 0.15 mg/mL (cysteine). The amino acid yield (5.14 mg/mL) in this investigation was much lower as compared to the enzymatic hydrolysis (endopeptidase) of DSM demonstrated by Liu et al.²¹ who obtained more than 25.6 mg/mL. The differences between these findings are most likely due to the possibility that these amino acids were still present in the form of intact protein and peptides in the residue, or that the amino acids were further degraded to low molecular weight carboxylic acids such as formic acids, acetic acids, propionic acids, etc. as a result of the excessive microwave irradiation^21,37.

The sugar profile as illustrated in Fig. 3c shows that the DSM-JP hydrolysate was predominantly comprised of sucrose (4.08 mg/mL), followed by mannose (3.12 mg/mL), xylose (0.54 mg/mL), and glucose (0.03 mg/mL). The ample composition of reducing sugars such as mannose and xylose in the hydrolysate suggested that they might be produced from the hemicellulose breakdown of biomass which was consistent with the previous FTIR spectra analysis. The scarcity of glucose content, as well as the other undetectable monosaccharides e.g., fructose and galactose, in the hydrolysate, could be ascribed to the occurrence of thermal-alkaline degradation of reducing sugars during the MAH treatment⁵⁶. The condensation of reducing sugar (the early stage of the Maillard reaction) such as glucose, with a free amino group e.g., lysine, results in protein glycation reaction, producing the Amadori products which eventually degraded into furfurals, reductones, and fragmentation products i.e., carbonyl and hydroxycarbonyl compounds⁵⁷.

Biocompatibility of DSM-JP hydrolysate and production of SJMo-Alg bead fertilizer

The formation of inhibitory compounds/ by-products such as acetic acid, hydroxy acids, dicarboxylic acids, and phenolic compounds has been linked to the use of mild alkaline treatment in the removal of resistant lignin and hemicellulose from agro-waste³¹. As a precautionary measure, the DSM-JP hydrolysate was primarily detoxified with activated carbon to remove the possible inhibitory compounds formed during the MAH treatment prior to use as growth media. The effects of detoxification on the soluble protein and sugar content in the hydrolysate were illustrated in Fig. 4.

According to the Fig. 4a, even with varying percentages of AC (2.5% and 5% w/v) employed, both protein and sugar concentrations in the hydrolysate were minimally impacted (P > 0.05). A marginal sugar loss of 2% and 4.4% were detected in 2.5% and 5% AC treatment, respectively. These results are consistent with the data obtained by Zhang et al.⁵⁸. In addition, the biocompatibility assay of DSM-JP hydrolysate on strain 40a as presented in Fig. 4b showed that both raw and detoxified hydrolysate could support the growth of strain 40a with no significant difference (P > 0.05) observed on cell counts during the 3-day incubation. As a result, it can be concluded that the raw DSM-JP hydrolysate requires no detoxification and can be utilized directly as a nutritional medium additive for strain 40.

The nutrient utilization dynamic (protein, glucose, reducing sugar, and total sugar) during the growth of strain 40a in the neutralized DSM-JP hydrolysate was further evaluated as illustrated in Fig. 4c. Contrary to expectations, the protein content was gradually declined while the glucose content was dramatically inclined. These findings show that strain 40a only used protein hydrolysate sparingly for cell development and rarely utilized glucose. This counterintuitive outcome differs from our previous observations in which strain 40a grew optimally with the supplementation of DSM protein and glucose as the most favourable nitrogen and carbon source, respectively⁵⁹. Surprisingly, the total sugar gradually decreased until day 3, reaching a high of 48% reduction, whereas glucose and reducing sugar increased up to 169% and 37%, respectively. This result implies that strain 40a might favour non-reducing sugars (e.g., sucrose) over monomeric sugars when DSM-JP hydrolysate was used as growth media.

The strong, inverse correlation (r =  − 0.97) between the reduction of total sugar and the increase in glucose and reducing sugar implies that their accumulation could be due to the breakdown of non-reducing sugar. Due to the abundance of sucrose in the hydrolysate, strain 40a tends to consume sucrose over other reducing sugars e.g., mannose and xylose. The breakdown of sucrose into its constituent monosaccharides, glucose, and fructose might explain the increasing pattern of both glucose and reducing sugar levels in the medium. Beisel and Afroz⁶⁰ described that bacterial cells, in general, are assumed to form a preference hierarchy as they begin with nutrients that are more quickly catabolized before moving on to others. Bacteria are assumed to abandon their pickiness and eat whatever is available when only poor nutrients are accessible, or the nutrients are present in low amounts.

The DSM-JP hydrolysate was then transformed into an alginate bead through encapsulation of strain 40a using sodium alginate and molasses (SJMo-Alg bead 40a). Molasses was incorporated in the bead formulation as a polycationic supplementary nutrient and mineral supply for strain 40a, as well as to improve the porosity of sodium alginate⁹. The representative image of SJMo-Alg bead 40a and the SEM image of a single bead are presented in Fig. 5. The beads appeared dark brownish red, mostly due to the colour combination of DSM-JP hydrolysate and molasses, shiny surface, almost spherical, and about 3–5 mm in size. Prior to further examination, the beads were aseptically transferred and stored in a refrigerator in a tightly sealed 1 L Schott bottle.

Performance of SJMo-Alg bead 40a in P and K solubilization

The performance of SJMo-Alg bead 40a in P and K solubilization was compared with the free cells of strain 40a (Fig. 6). As presented in the figure, the P release achieved the highest point after day 3, which was 512.16 µg/mL and 490.81 µg/mL for free cell and encapsulated 40a, respectively. Both free cell and encapsulated 40a demonstrated no significant difference (P > 0.05) in the solubilized P and pH in the NBRIP media. Likewise, the solubilized K and pH in the Aleksandrov media of both free cell and encapsulated 40a showed no significant difference (P > 0.05) during the 3-day incubation. The highest solubilized K achieved was 78.34 µg/mL and 73.88 µg/mL for free cell and encapsulated 40a, respectively. Both NBRIP and Aleksandrov media turned to acidic pH on day 1 and were maintained around pH 4 until the end of the experiment. These results were coherent with those observed in our prior research on P and K solubilization activities of strain 40a⁵. These results provide further support for the hypothesis that the alginate encapsulation does not deteriorate or compromise the biological cellular activities of strain 40a, hence suitable as a biofertilizer candidate particularly for soil available P and K amelioration.