Fresh papaya fruits were procured from a local supermarket located in Jilin, China. The supplied enzymes, such as protease and heat-stable α-amylase were purchased from Sigma Aldrich (St. Louis, United States). All chemicals and reagents used in this study were of analytical grade.

The proximate parameters, such as protein, fat and ash content were calculated according to the AACC (2000) methods. In case of protein analysis, nitrogen content thus obtained was multiplied by a protein conversion factor of 6.25 in order to get crude protein content as given in Eq. 2.

Assessment of the effects on morphology and microstructure of DF extracted by different extraction methods was conducted by scanning electron microscope (SEM: SU8010, Hitachi, Tokyo, Japan). Specifically, the DFs from each extraction method were fixed on the specimen holder and plated on the sputter with the gold powder and SEM monitoring was carried out at operation voltage of 5 kV and image capturing was performed at 2000× magnifications.

FTIR spectrophotometry was employed for analyzing the organic functional groups of PP and PS DFs extracted by different extraction methods. Using an FTIR spectrophotometer (Tensor 27, Bruker Daltonics Inc., Bremen, Germany), the scanning was performed at spectrum wavelength ranging from 400 to 4,000 cm⁻¹ at scan speed of <10 s for 4 scans at resolution of 4 cm⁻¹. The obtainment of spectra was carried out in Attenuated Total Reflection (ATR) mode.

The XRD patterns were measured by means of a diffractometer (7000S, Shimadzu, Kyoto, Japan) with copper Kα radiation at 40 kV (50 mA and 0.154 nm) as well as diffraction angle (2θ) ranged 4–40°C.

Thermal properties were measured in terms of weight loss, which was calculated by means of a thermogravimetric analyzer (TGA, TG/DTA 8122, Rigaku, Tokyo, Japan), following the method reported by Du et al. (2019) for evaluating the DF samples. The measurement of weight loss was carried out over an operational temperature range of 30–600°C at a heating rate of 20°C/min under nitrogen atmosphere.

DFs solutions obtained from each extraction method were subjected to viscosity measurement using a rheometer (Discovery HR-1, TA, New Castle, DE, United States). Analysis was performed by means of steel cone geometry (gap = 56 μm: diameter = 60 mm). All measurements were recorded thrice at an operational temperature of 25°C (Jiang et al., 2022b).

Monosaccharides determination through HPLC was carried out through modified method of Jiang et al. (2022b). Each extracted sample was taken in an amount of 2 mg and dissolved in 1 mL of 2 M trifluoroacetic acid (TFA) and under sealed condition for a period of 2 h at 120°C in hydrothermal reactor. Prior to HPLC analysis, a standard filter with size of 0.45 μm was employed for filtration of aqueous layer. HPLC system of Dionex Thermo Ultimate 3,000 (Dionex Co., Sunnyvale, CA, United States) equipped with diode array detector (DAD) supplied by Thermo Fisher Sci. (Waltham, MA, United States) was employed for detection of monosaccharides. “The mobile phase was prepared by mixing acetonitrile (A) and phosphate buffer solution (0.1 mol/mL; PBS, pH 6.7) in a ratio of 82:18 (v/v). The analysis was performed using a Supersil ODS2 column (5 μm, 4.6 × 250 mm²) with an injection volume of 20 μL and a flow rate of 0.8 mL/min. Detection was carried out at a wavelength of 245 nm against reference standards.

WHC and OHC were determined as per the methods described by Raza et al. (2021). Briefly, 0.5 g of each extracted sample was taken in the centrifuge tube followed by addition of 5 mL of distilled water with samples and then was subjected to incubation at 37°C for period of 1 h. After the filtration of water, the WHC of samples were measured according to following Eq. 3.

Whereby, the W₁ depicts total sample weight on dry basis and W₂ represents total sample weight after the water was filtered out.

OHC of all extracted samples were determined by preparing mixture of each sample in specified amount of 0.5 g with soybean oil in centrifuge tube followed by incubation at 37°C for 1 h interval. Samples were centrifuged at 5,000×g for 10 min interval and OHC was measured using the following Eq. 4

Whereby, the W₁ depicts total sample weight on dry basis and W₂ represents total sample weight after removal of excessive soybean oil.

WSC was calculated for each DF sample according to the modified method of He et al. (2020). In brief, DF sample in a specified amount of accurately weighed 0.5 g was transferred to 26 mL of measuring cylinder and then was mixed with distilled water of 20 mL. The volume was calculated prior and after the completion of standing time of 2 h and WSC was measured in terms of volumetric change in DF dry sample after water swelling as per the Eq. 5 given:

Whereby, m₀ represents DF sample dry weight, V₁ represents initial volume and V₂ indicates final volume after completion of standing time interval.

CAC values were measured as per the reported method of Jiang et al. (2022b). Each extracted sample (1 g) was mixed with 30 mL of diluted yolk solution. After the adjustment of pH to 2.0 or 7.0, samples were subjected to incubation at 37°C for 180 min. Absorbance of the supernatant was recorded spectrophotometrically at detection wavelength of 550 nm and calculated CAC values were represented in terms of mg/g as given in Eq. 6.

Where m is denoted as the mass of the constant sample weight (g); n₁ is the cholesterol concentration of before adsorption (mg/mL), while n₂ is the cholesterol concentration in the supernatant after centrifugation (mg/mL); V is volume (mL).

BAC measurement was carried out as per the method reported by Jiang et al. (2022b). Each sample in specified amount of 1 g was mixed with 30 μL of sodium cholate (pH 7.0: 1–3 mg/mL) followed by incubation at 37°C for 2 h interval. Supernatant was analyzed for absorbance through spectrophotometer (UV-1800, Shimadzu Instruments Mfg. Co., Ltd., Kyoto, Japan) at detection wavelength of 620 nm.

NIAC was assessed as per the reported method of Jiang et al. (2021) with certain modifications. Each extracted sample (1 g) was mixed with 25 mL of sodium nitrite solution (100 μmol/L) and pH adjustment was carried out to 2.0 and 7.0 in terms of simulation of the pH conditions similar to human stomach and small intestine. Then, the samples were incubated for 120 min at 37°C. N-(1- naphthyl) ethylenediamine dihydrochloride/sulfanilic acid reagent was employed for determination of nitrite ion content spectrophotometrically at detection wavelength of 538 nm.

GAC values were measured as per the method of Wang et al. (2021). Each sample (0.5 g) was mixed with 50 mL of glucose solution (conc. 50 mmol/L). Samples were continuously stirred for 2 h interval at 37°C followed by centrifugation of samples at 2,862×g for 5 min time interval. Determination of the final glucose concentration of collected supernatant was performed by a glucose assay kit (oxidaseperoxidase kit, Rongsheng Biotech Co., Ltd., Shanghai, China).

The DF samples were extracted according to the method described by Jiang et al. (2018). Homogenization and sonication treatments were performed sequentially (first the homogenization was performed followed by sonication). The obtained extract was employed for determination of antioxidant activities in later experiments.

Antioxidant activities, such as DPPH, ABTS and FRAP were assessed accorng to the methods described by Thaipong et al. (2006). A spectrophotometer was employed for measurement of the absorbance values of DPPH, ABTS and FRAP values at detections wavelengths of 515, 734 and 593 nm, respectively. The results of antioxidant activities were represented in terms of μmol trolox equivalent (TE)/g and standard curve linearity was ranged between 25 and 800 mM TE/g.

All the statistical analyses were performed by means of SPSS 20 0.0 software (SPSS Inc., IL, United States). All experiments were carried out in triplicate manner and results were represented in terms of mean ± standard deviation (SD) whereas analysis of variance (ANOVA) in conjunction with Duncan’s test were employed for assessment of the differences between the means at a significance level of p < 0.05.

The results of DF yield and proximate composition of DF fraction from PP and PS obtained by various extraction methods are provided in Table 1. In the case of PP, DF extracted by AL exhibited the highest DF yield of 84.23%. Similarly, DF yield of PS was also analyzed and it was evident from results that PS-AL fraction exhibited the highest DF yield (81.13%). The probable reason for the highest DF recovery by alkaline extraction can be ascribed to the thorough destruction of cell wall and the disruption of compact molecules within the cell wall structure when exposed to an alkaline solution. This process results in the partial dissolution of hemicellulose in IDF, consequently leading to a higher extraction yield of DF (Sun et al., 2018). The protein content of DF extracted by PP-AC and PS-AC was the highest. Ash and fat contents in DF extracted by PP-AC and PS-AC was found to be much higher. Similar results have been reported by Wang et al. (2021) for the extraction yield and proximate compositions of IDF and SDF extracted from kiwi fruit using various extraction methods.

In the current study, SEM was employed to elucidate the microstructures of DF fractions derived from all extraction methods. As shown in Figure 1, DF extracted by PP-AL and PS-AL exhibited more complicated structures with lesser degree of pore loosening as compared to other samples. In PP-AC and PS-AC extracted DFs, the surface irregularity, pore loosing, disintegration of cell walls and microstructure impairment were more prominent as compared to DF extracted from PP and PS by EN and AL. Severe structural impairment of DF extracted by PP-AC and PS-AC might be ascribed to the disintegration of sheet-like structure in DF of PP and PS. The SEM images indicate that AC condition exhibited significant effect on DF microstructure. Moreover, DFs with less compactness and looser spatial structures exhibit increased surface area, which may significantly affect the adsorption capacities of bile acid, water, oil, glucose and nitrite ion. Conclusively, it may be implied that the type of extraction method may lead to alteration in the DFs properties extracted from PS and PP. These results are in agreement with the findings of Wang et al. (2021), who reported significant effects of extraction methods on the morphology and microstructural properties of DF extracted from kiwi fruit.

Figure 2A shows the transmittance percentage for all DF fractions of PP and PS obtained by various extraction methods. The stretching vibrations of two broad peaks at IR regions of 3,427 cm⁻¹ and 2,927 cm⁻¹ were visible in all DF’s samples and were indicative of the stretching vibrations of hydroxyl (-OH) group and –CH of sugar methyl as well as methylene groups, respectively. The absorption peaks existing at IR regions of 1,200–1,400 cm⁻¹ might be ascribed to the variable angle CH vibration. Furthermore, absorption peaks at these IR regions are reported in published literature as characteristic absorption peaks corresponding to saccharides in DF. Similarly, irrespective of the type of extraction method used, all DF fractions of PP and PS exhibited absorption peaks at IR regions ranged 1,000–1,300 cm⁻¹ which might be attributable to the contraction IR vibration corresponding to ester C–O, while peaks presence in IR spectral range of 700–1,000 cm⁻¹ may be ascribed to the characteristic absorption peaks of α- and β-pyran monosaccharides. The absorption peaks with weak intensities at IR spectral ranges of 3,414–3,480 cm⁻¹ (vibrational stretching of –OH groups) and 2,856–2,923 cm⁻¹ (vibrational stretching of C–H groups in case of polysaccharides polymers) in DF fractions of PS and PP might be attributed to the possible breakage of intermolecular bonds and disintegration of molecular structures after exposure to AC, EN and AL treatments. It was confirmed by the FTIR results that EN extraction caused less peak broadening in IR peaks as compared to the PP-, PS-AC, and PP-, PS-AL implying that EN extraction led to increased degree of hydrolysis followed by AL treatment in hemicellulose and pectin of PS and PP-derived DF’s samples. In all DF samples, the IR absorption peak at the region of 1,000 cm⁻¹ was indicative of the C–O stretching vibration in C–O–C linkages (typical IR peak of xylan) which implied that DF comprised of xylan hemicellulose in its structural configuration. IR peaks existence at IR spectral region of 1,660 cm⁻¹ in DF extracted by PS-AC and PP-AC corresponded to the bending or stretching vibration of aromatic lignin hydrocarbon whereas, this characteristic peak was not found in the case of DFs of PP and PS extracted by EN and AL methods, implying that EN and AL treatments led to increased degree of disintegration in lignin molecules. These results are in line with the previous studies reported by Wang et al. (2021) for DF extracted from kiwi fruit.

TGA curves for all DF samples are depicted in Figure 2B. TGA curve is usually described in terms of three stages. During the initial stage at temperature range of 30–210°C, all DF samples irrespective of extraction method showed decreasing trend with respect to weight loss in conjunction with devolatilization occurrence at 120°C. In the case of both PP and PS, the highest weight loss was observed in case of DF samples by AC method followed by AL and EN methods with corresponding increase in the temperature. During initial stage of TGA analysis, the loss percentage in weight might be ascribed to the possibility of absorbed water evaporation from the DF samples extracted by various methods. It was also implied from the TGA curves that 210–400°C temperature range was for the second stage of TGA analysis and maximum degree of weight loss was occurred during this stage. This enhanced degree of weight loss during second stage might be attributable to the possible elevated degradation of pyrolytic polysaccharides mainly comprising of pectic polysaccharides and hemicelluloses. The weight loss observed in DFs extracted by PS-EN and PP-EN was relatively slower and proceeded in gradual manner, whereas weight loss was higher in the case of AL and AC methods during third stage for both PS and PP, which implied that DF’s from PS-EN and PP-EN exhibited relatively more thermal stability in comparison with those of other samples. The lower thermal weight loss observed for enzymatic extracted DF as EN method caused less polymer chain cleavage when compared to other methods of extraction. This may probably the main reason for higher thermal stability of enzymatic extracted fibre in such a way that enzymes had not degraded the main polysaccharides of extracts compared to acidic and/or alkaline treatments. The lower thermal weight loss observed for enzymatically extracted dietary fiber compared to acidic and alkaline extracted dietary fiber can be attributed to several factors related to the extraction methods and the nature of the fibers obtained. Here are some probable reasons, such as (1) enzymatic extraction often involves milder conditions compared to acidic or alkaline extraction. The use of enzymes is more specific and selective, allowing for the targeted breakdown of cell wall components without causing extensive degradation or modification of the fibers; (2) enzymatic extraction is less likely to disrupt the structural integrity of dietary fibers. Acidic and alkaline conditions can lead to chemical modifications and structural breakdown of fibers, making them more susceptible to thermal degradation; (3) acidic and alkaline extractions may leave residues of the chemicals used in the process, which can contribute to thermal weight loss during subsequent analysis. Enzymatic extraction typically involves fewer chemicals and, therefore, results in lower chemical residues that might contribute to thermal degradation (Huang et al., 2018; Ma et al., 2022; Jiang et al., 2023). These results were in agreement with the previously reported findings of Wang et al. (2021) for DF extracted from kiwi fruit.

The results of diffractograms from DF extracted by each method are demonstrated in Figure 2C. Generally, all the DF samples of PP extracted by various methods exhibited characteristics main signal peaks at 2θ angles of 14–15°, 21.5°, 22.32°, 23.5–24.5° and 30–31°. It has been reported in the published literature that signal peaks around 14–15° and 21–23.5°might be due to the presence of cellulose crystals. Cellulose exhibits crystalline configuration owing to Van der Waals forces and hydrogen bonding linkages in adjacent molecules. Moreover, the amorphousness in DFs exists because of non-crystalline cellulose, hemicellulose and lignin (Kulasinski et al., 2014). The peak existing at 2θ angle range of 21.2–23.5° was quite broader in DF extracted by PP-AC as compared to sharper peak in the case of PP-AL. PP-AL DFs showed existence of another broader peak at 34.23–37.5°whereas this peak intensity was much lower in case of PP-AC. PP-EN DFs showed two split peaks emergence at 38.23–32.51°. Regarding PS samples, DF extracted by AC (PS-AC) exhibited large peak with broad shape at 14.12–27.5° as compared to PS-AL, which showed very low signal peak intensity at same 2θ angle. PS-AL DFs exhibited sharp peak with medium intensity at 34.19–37.25°. For PS-EN DFs, small peaks emergence was noted at range of 25.28–39.53°. As far as the extractions methods are concerned, the peak positions and peak width were affected by AL and EN extraction methods, which implied that both AL and EN extraction methods resulted in significant disruption in the crystalline structures of PP- and PS-DF. Moreover, both AL and EN methods led to emergence of new signals peaks and consequently new crystalline formations in structural configurations of DF from EN and AL methods as compared to that was obtained by AC extraction method. The irregularity in peaks of DF extracted by PS-EN, PP-EN, PS-AL and PP-AL might be because of denaturation effects of enzymatic hydrolysis and NaOH solution. Similar results have been reported by Ma and Mu (2016) whereby authors have carried out the research regarding DF extraction from deoiled cumin and have reported the denaturation effects caused by enzymatic hydrolysis.

The viscosity of all DF samples was measured as a function of shear rate and the results are demonstrated in Figure 2D. In the case of PP-DF samples, all the extractions methods showed similar rheograms, and DF extracted by PP-AL showed slightly decreasing trend in viscosity. However, PP-AL exhibited slightly lower viscosities as compared to those of PP-AC and PP-EN. Similarly, for PS samples, the rheograms of PS-AC and PS-AL exhibited less degree of curvature in comparison with that of PS-EN. Moreover, it is worthwhile to note that charge-transfer complexes (CTC) and hydrogen bonding linkages present in the DF’s polymer chains may affect the viscosity parameter (Jiang et al., 2023). DF extracted by PP-AC and PS-EN showed a high degree of linearity in rheograms as compared to other DF samples and differences in the viscosities of PS and PP samples extracted by AL and EN methods. This may be ascribed to the structural modifications, greater interactions of particles, alterations in cluster structures of DF, and breakage of partial branching, thereby increasing the apparent viscosity (Jiang et al., 2022b). Moreover, EN hydrolysis and NaOH solution caused destruction of the molecular structures owing to breakage of partial branching patterns and reduction of internal molecular linkages and tightness (Bi et al., 2020). These results were in line with the previously reported study of Jiang et al. (2022b) for DF extracted from Sanchi flower.

The results of DF samples monosaccharide composition by determined by HPLC are shown in Table 2 and Figure 3, exhibiting the molar ratio of monosaccharides in DF derived from PP and PS by various extraction methods. It was found that all DF samples consisted of mannose, rhamnose, glucuronic acid, galacturonic acid, glucose, galactose, xylose and arabinose. The dominant monosaccharides of DF derived from PP and PS samples were found to be glucose, galacturonic acid and galactose. Previous studies have reported that the main monosaccharides of DF are glucose, galactose and arabinose from DF of calamondin pomace and sweet potato (Huang and Ma, 2019; Neela and Fanta, 2019). Glucose is mainly derived from starch and cellulose, and the content of DF extracted by PP-AC and PS-AC glucose is relatively high, which may be because the AC environment promotes the hydrolysis of cellulose in the cell wall. In both DFs from PS and PP, the higher molar ratio of galactose might be ascribed to the enhanced hydrolysis of cell wall cellulose after exposure to AC and AL extraction methods. Moreover, the increased galactose in DF extracted by PP-AC and PS-AC could be the main contribution to the increased SDF content, thereby affecting the water holding capacity of PP and PS DFs. The DF derived PP samples had higher galacturonic acid than those of DF derived PS samples, indicating that PP contains more pectin, which had good physical and functional properties. Galacturonic acid is the main component of pectin. However, higher xylose observed in DF derived from PS samples, indicating a higher presence of xylan hemicellulose. The relatively high content of rhamnose in DF extracted by PP (AC and AL) and PS (AC and AL) indicates that SDF contains more pectin. Pectin always contains varying amounts of neutral sugars, such as D-galactose, L-rhamnose, L-arabinose, and D-xylose. Some of these neutral sugars are constituents of side chains to the galacturonan backbone (Flutto, 2003). The amounts of mannose, galacturonic acid and glucuronic acid in PP- and PS-AC decreased compared to the PP- and PP-EN amount, while the glucose amount increased, which was possibly due to the breakage of glycosidic bond by AC extraction, as well as the formation of glucose (Wang et al., 2021). Based on our results, the different extraction methods had impacts on PP- and PS-derived DFs.

All DF samples were analyzed for their DPPH, ABTS and FRAP activities and results are demonstrated in Table 3. Comparatively, among all extraction methods, the PP and PS DFs samples extracted by AC exhibited the highest DPPH, ABTS and FRAP values. Moreover, DF derived from PP had higher antioxidant activities than those of DF derived from PS samples. According to relevant literature, the DPPH free radical scavenging ability of DF is related to the content of phenolic substances and the structure of phenolic acids in raw materials (Esposito et al., 2005; Guo and Beta, 2013). In addition, some studies have found that the reducing power of the crushed fiber was improved, which may be due to the change of the fiber structure caused by the shear force during the crushing process, resulting in the release of some phenolic substances (Zhu et al., 2015). Here, the high reducing power of DF extracted by PP-AC and PS-AC may be due to the release or exposure of some phenolic substances in PP and PS caused by the treatment of AC during the extraction process.