Bioprocessing of pineapple waste biomass for sustainable production of bioactive compounds with high antioxidant activity

The effect of temperature, moisture content and pH during solid-state fermentation (SSF) of MD2 pineapple peel with Rhizopus oryzae (MUCL 28168) was evaluated on the release of bioactive compounds with antioxidant capacity. Applying a central composite design, it was found that temperature had a significant effect (p < 0.05) on the total phenolic content and DPPH antioxidant activity while for the ABTS radical elimination activity, the factor that presented a significant effect was the pH (p < 0.05); as this factor increases, the antioxidant activity enhances. The optimal conditions for fermentation process were 80% of moisture content, pH 5.5, temperature 37.3 °C and 24 h of process to maximize phenolic content and antioxidant activity. Gallic acid, chlorogenic acid, caffeic acid and cinnamic acid were identified in the extracts by HPLC analysis. These results permit to conclude that SSF of pineapple peel is an effective bioprocess for the release of phenolic compounds with antioxidant activity.


Introduction
Tropical fruits are characterized by their appetizing flavor and aroma. In addition, these fruits contain a large number of bioactive compounds that provide health benefits for consumers [1]. In the last decade, world production of pineapple (Ananas comosus) has increased, being 25.4 million tons by 2019. The main producing countries are: Costa Rica (3328.1), Philippines (2747.86) and Brazil (2426) million metric tons [2]. This fruit is a source of antioxidants; it contains minerals, vitamins and nutrients [3,4]. It is also an important source of bromelain, a proteolytic enzyme with various biotechnological applications in the pharmaceutical, food and cosmetic industries [5]. At the agro-industrial level, most of the pineapple production is used for human consumption in fresh preparations. During processing, large amounts of wastes are generated, mainly peel, center (core) and crown, which represent approximately 25-35% of the weight of the fruit weight [6]. These by-products show great potential to be used as antimicrobial agents, antioxidants, colorants, flavorings and even as texturizing additives, because they contain bioactive compounds (mainly phenols and flavonoids), carotenoids, vitamins and fiber [7].
Singularly, in pineapple peels have been identified some compounds with antioxidant activity such as salicylic acid, myricetin, trans-cinnamic acid, tannic acid and p-coumaric acid [8]. Upadhyay et al. [9] reported antimicrobial activity thanks to the ferulic acid and syringic acids present. According to the research carried out by Li et al. [10], the main phenolic compounds identified in pineapple peels were gallic acid, epicatechin, catechin and ferulic acid. The first step to obtain the bioactive compounds present in these plant substrates is to perform extraction processes [7]. Generally, conventional extraction techniques such as Soxhlet extraction, maceration or hydrodistillation do not allow to obtain an adequate release of phenolic compounds bound to plant substrates [11,12]. However, in recent years, other techniques have been developed to obtain these compounds [13,14]. Solid-state fermentation is one of the processes that has been used on agro-industrial by-products to extract bioactive compounds [15]. In the SSF process, filamentous fungi are usually the most widely applied microorganisms, due to the enzymes they produce during fermentation, which are lignolytic and cellulolytic (they allow improve the release and extraction of biocompounds) [16,17].
Regarding pineapple by-products, there is research that reports their physicochemical characteristics as well as the bioactive compounds present in each of its by-products, which have applications in cosmetics, pharmaceuticals, and food. However, no research has been conducted that reports the extraction of bioactive compounds by SSF using Rhizopus oryzae. Hence, the objective of this research was analyze the influence of the SSF process with R. oryzae in obtaining phenolic compounds with antioxidant capacity from pineapple peel. The SSF process conditions were optimized by taking as response variables the total phenolic content (TPC) and the antioxidant activity by DDPH and ABTS.

Plant material
Pineapple (Ananas comosus) peels were provided by the IDEMA market plaza located in the city of Cali (Colombia). They were obtained at maturity grade 3 as reported in the NTC 729 (Colombian technical standard). The peel was dehydrated at 50 °C for 24 h, and worked with a particle size > 2.5 mm, then stored in vacuum polyethylene bags at room temperature.

Moisture content
To determine the moisture content, the methodology reported in AOAC 934.06 [18] was used. Briefly, 5 g of pineapple peels were weighed on an analytical balance (Boeco BAS-32 Plus, Germany), placed in petri dishes and dried at 70 °C (Universal Oven UN30 Memmert, Germany) until constant weight was reached. The measurement was carried out in triplicate.

Fat content
The fat content was measured with ether extract (by the Soxhlet system) as described in AOAC 920.39 [19]. One gram of previously dried sample was weighed in an analytical balance (Boeco BAS-32 Plus, Germany) and deposited inside the cellulose cartridges placed in the Soxhlet extractor. Next, the previously dried flat-bottomed balloons were weighed on an analytical balance (Boeco BAS-32 Plus, Germany) and this weight was recorded. The balloons were placed in the extractor and the equipment was closed. 50 mL of petroleum ether (Sigma-Aldrich, USA) was added to each sample and the refrigeration equipment was turned on. The extraction process was carried out for 8 h, during which time the ether, heated to 80 °C, was passed through the sample to extract the fat. The ether with the dissolved fat falls into the flat balloons and evaporates, being recovered by passing through the cooling coil. The extracted fat is deposited in the flat-bottomed balloons. After the extraction time has elapsed, the flat-bottomed balloons are placed in an oven (Universal Oven UN30 Memmert, Germany) for 6 h to eliminate the ether residues, after which the balloons are allowed to cool in the desiccator and weighed on the analytical balance (Boeco BAS-32 Plus, Germany). The percentage of crude fat was calculated with the following equation: where W 1 is the weight of the flat-bottomed balloon with the fat residue of the sample, W 2 the weight of the empty flat-bottomed balloon and W the weight of the sample. The fat value obtained corresponds to the % F in 100% of the dry matter.

Protein content
The crude protein content was determined by the Kjeldahl method according to the methodology used by [20,21] of sample was weighed on filter paper, folded and placed in the Kjeldahl digestion flask, a Kjeldahl tablet (Sigma-Aldrich, Germany) and 30 mL of concentrated H 2 SO 4 (Chem-Lab, Fisher Scientific, USA) were added. The flask was placed in the Foss Tecator digester (FOSS, Denmark) to perform the digestion process, shaken until a clear solution was obtained and allowed to cool, attached to the distillation part of the Fosskjeltec 2100 protein equipment (FOSS, Denmark). 25 mL was added to a conical flask and attached to the condenser of the distillation chamber and the outlet tube was submerged below the surface of the boric acid in the conical flask. The thistle funnel was filled with a 50% NaOH (Chem-Lab, Fisher Scientific, USA) solution and the equipment was turned on for 6 min until alkalinization was achieved with the addition of 50% NaOH (Chem-Lab, Fisher Scientific, USA). When boiling began, NH 3 was released, which moved towards the outlet tube and came into contact with the indicator boric acid (Sigma-Aldrich, Germany) for 5 min, then the NaOH tap was closed, and the boric acid mixture was titrated with 0.1 N HCl. Finally, the % crude protein was determined using %N*6.25.

Fiber content
The crude fiber content was determined by the acid and alkaline hydrolysis method AOAC 991.43 [22]. They were weighed 2 g of previously defatted sample, later they were placed in a glass of Berzelius, adding 100 mL of sulfuric acid (Chem-Lab, Fisher Scientific, USA) 0.225 N. The connection was made to the reflux equipment for 30 min from the moment it began to boil. Subsequently it was dried and filtered using linen cloth and the sample was washed three times with 100 mL of hot distilled water and the litmus paper test (Whatman R, England) was performed to verify the presence of acid. The fiber (residue left on the linen cloth) was placed in the Berzelius beaker with 100 mL of sodium hydroxide solution (Chem-Lab, Fisher Scientific, USA) 0.313 N and connected to the reflux equipment for 30 min. It was then removed and filtered through the linen and washed three times with hot distilled water and the litmus paper test (Whatman R, England) was performed to verify the presence of alkaline reaction. The excess water was wrung out by pressing the linen cloth. The linen cloth was removed from the funnel, the fiber was spread and removed with a spatula and deposited in a porcelain crucible. This crucible was placed in an oven (Universal Oven UN30 Memmert, Germany) at 100 °C for 12 h. Subsequently it was removed from the oven and allowed to cool in the desiccator, this weight was recorded. The sample was pre-incinerated in grill and introduced in muffle (Arder LM-E8, Colombia) at 550 °C for 3 h, the crucible was removed leaving it to cool in the desiccator and this weight was registered. The percentage of crude fiber is calculated with the Eq. 2:

Ash content
The ash content was carried out at 550 °C for 12 h as described by the methodology [19]. The porcelain crucibles were placed at 100 °C for 24 h in oven (Universal Oven UN30 Memmert, Germany), later they were cooled for 20 min in desiccator and this weight was registered. 2 g of sample were placed inside the crucible and placed in a muffle (Arder LM-E8, Colombia) at 550 °C for 12 h. The samples were removed from the muffle and allowed to cool for 20 min in a desiccator, this weight was recorded on an analytical balance (Boeco BAS-32 Plus, Germany). The percentage of ash was determined using Eq. 3:

Carbohydrate total
The content of carbohydrate (CHO) was calculated using the method by FAO [23] as shown in Eq. 4 [24].

Total sugars
For quantifying total sugars present in the samples, the method proposed by [25] was used, for which 250 µL of solution were taken, 250 µL of 5% phenol were added, they were centrifuged and left in an ice water bath for 5 min. Later, 1000 µL of concentrated sulfuric acid were added and everything was centrifuged, boiled for 5 min, and later, allowed to cool to room temperature and then read in a UNICO 2100 spectrophotometer at an absorbance of 480 nm. Glucose (Sigma Aldrich, Germany) was used as a standard for the calibration curve. = 100 − (crude protein% + ash% + crude fat% + crude fiber%)

Reducing sugars
The reducing sugars will be determined by the DNS method (3,5-dinitrosalicylic acid) following the methodology described by [26]; 1 mL solution was taken and 1 mL of the DNS reagent was added on an ice bath was stirred and boiled for 5 min, then the samples were cooled and 8 mL of distilled water were added. They were agitated and let it to stand for 15 min. The analysis of the samples was conducted in a UNICO 2100 spectrophotometer at an absorbance of 575 nm. For the calibrating, the curve solutions of 200-1000 mg/L were prepared using glucose (Sigma Aldrich) as standard.

Water Absorption Index (WAI) and Critical Humidity Point (CHP)
Water Absorption Index (WAI) [27] and Critical Humidity Point (CHP) [28] were determined to establish the pineapple peel was viable as a support for the SSF process. For WAI quantification, 1.5 g of peel were weight with 15 mL of distilled water, shaken for 10 min and centrifuged at 3000 rpm for 10 min. The supernatant was discarded to calculate the WAI with the weight of the remnant; these results were expressed as g gel/g dried peel. The critical humidity point was ascertained adding 1 g of peel in a thermobalance (OHAUS, USA) at 120 °C for 60 min, and the results were expressed as a percentage.

Reactivation and conservation of the microorganism
Dried spores of the microorganism Rhizopus oryzae (MUCL 28168), recognized as GRAS, were used; these were taken from the culture collection of the Laboratory of Microbiology and Applied Biotechnology (MIBIA), from Universidad del Valle. For reactivation and propagation, the spores were suspended in test tubes with 5 mL of sterile distilled water for 30 min. Test tubes with 8 mL of Potato Dextrose Agar (PDA) were prepared, and later the spore solution was seeded by depletion in the agar tubes and incubated at 30 °C for 8 days. After, 10 mL of sterile distilled water were added to the test tubes with the culture, and they were stored at 28 °C, with reseeding every 3 months for their conservation.

Preparation of the inoculum
The inoculum was prepared following the Tane-koji technique, from steamed whole brown rice. During the process, the rice was inoculated with spores in suspension at a concentration of 1 × 10 7 spores/mL, for 75 h until complete sporulation. Later, they were dried at 40 °C for 20 min until reaching a moisture content of 12% [29]. The product obtained was ground in a blade mill (Kitchenaid BCG211) and stored in sterile containers at 28 °C. The spore count was carried out in a Neubauer chamber, and the concentration was quantified using Eq. 5:

Determination of radial growth velocity
To evaluate the speed of radial growth, 5 mm diameter discs were cut from samples of the fungus (1.8 × 10 7 spores/g), planted in Petri dishes with PDA agar for 8 days prior to the test, and placed in the center of Petri dishes with 15 mL of the previously prepared pineapple peel flour-agar-agar substrate. The invasiveness was determined by measuring the increase in the value of the radius of the colony every certain time interval.

Microstructural analysis
The morphological characteristics of fermented and unfermented peels were analyzed with a scanning electron microscope (Phenom Pro X, Thermo Scientific, USA). The peels were cut with a scalpel to obtain a transverse sample. Subsequently, they were placed on a support containing a platinum coating Auto-Fine Coater JFC-1600 (Joel, USA Inc, USA). The images were obtained with an accelerating voltage of 15 kV.

Evaluation of the solid-state fermentation conditions
The solid-state fermentation process was realized in sterile aluminum trays which were 12 × 10 × 4 cm in size. These ones were placed in a controlled temperature incubator. The samples were inoculated with Rhizopus oryzae spores and incubated in an oven at controlled temperature and 90% relative humidity conditions for 24 h. The fermentation conditions were defined by means of a Central Compound Design (DCC), to determine the effect of moisture content (%), temperature (°C) and pH on the total phenolic content and antioxidant activity (3 factors with 2 levels). For a total of 20 treatments, data obtained in the literature were taken and preliminary tests were carried out to select the number and variation of levels of the process variables in the experimental design. To obtain the extracts, 75% ethanol was used.

Total phenolic content (TPC)
The total phenolic content was determined by a spectrophotometric method with the Folin-Ciocalteau reagent, described by Avella et al. [30]. For quantification, 500 μL of the diluted extract in a ratio (1:40) was added in a test tube and 750 μL of Folin-Ciocalteau reagent (1 N) was added, shaken and left to stand for 5 min. Then, 750 μL of sodium carbonate (0.01 M) was added and left to stand for 5 min. The samples were read at 760 nm in a UNICO 2100 spectrophotometer. The calibration curve was performed with the Gallic Acid Standard (0-100 ppm).

Condensed tannin content
Total condensed polyphenols were quantified using ferric reagent and HCl-butanol [31]. The total condensed polyphenol content was expressed as milligrams of catechin equivalents per gram of dry plant material (mg QE/g).

Antioxidant activity (DPPH)
The free radical elimination activity (DPPH) of the extracts was determined following the methodology described by Ballesteros et al. [32]. For which 100 µL of each extract were taken, 2900 µL of the DPPH solution were added, agitated in a vortex and placed in the dark for 30 min. Consequently, the absorbance was recorded in a UNICO 2100 spectrophotometer at an absorbance of 517 nm. The radical scavenging activity (RSA) was calculated according to Eq. (6).
where A c is the absorbance of the control sample and A s is the absorbance of the sample.

Antioxidant capacity by ABTS
For the determination of the ABTS antioxidant activity of the extracts, the methodology applied by Mesa-Vanegas et al. [33] was used. The preparation of the ABTS + radical was carried out through oxidation of ABTS (7 mM) with potassium persulfate (2.45 mM), being left in the dark during 16 h. The ABTS + solution was diluted with 80% methanol to obtain an absorbance of 0.70 ± 0.02 at 734 nm. For the samples, 100 μL of the extract were taken and 1800 μL of the methanolic ABTS + solution was added. They were read in a UNICO 2100 spectrophotometer at an absorbance of 734 nm. The calibration curve was developing using Trolox as a standard. The ABTS + radical scavenging activity was calculated using Eq. (7).
where A 0 is the absorbance of the sample at the initial time and A t is the absorbance of the sample at 30 min.

Identification of bioactive compounds by HPLC
Extract samples were analyzed on an HPLC system (Thermo Scientific Accela 600, USA) including quaternary pump and PDA detector. They were analyzed (10 µL

Tests of support in solid-state fermentation
Following the methodology described in sections "Moisture content"-"Total sugars", the analysis of the raw material methodology was conducted. The results were recorded in Table 1.
According to the results reported in Table 1, the initial moisture content of the samples was on average 85%, a value that is similar to that reported by Campos et al. [35] of 82.35%, Rico et al. [36] of 83% and 84.14% of pineapple solid waste reported by Zain et al. [37].
The analyzed sample presented a fat content of (1.06 ± 0.017%), which is a similar value to that reported by Sousa et al. [38] of 0.69% who evaluated pineapple residues obtained of a fruit pulp industry. The amount of fat was not detected in the pineapple peel samples analyzed by Campos et al. [35]. For their part, Banerjee et al. [39], reported a range between 0.46 and 5.3% for this property. Selani et al. [6] reported a value of 0.6% and Rico et al. [36] presented an interval between 1.1 and 2.0%. Therefore, it can be concluded that the values obtained are consistent with the literature and may differ since it is not the same variety.
The result obtained for ashes was 2.6%, which is a high value when compared to that reported by Sousa et al. [38] of 0.53%. However, this value is within the range reported by Banerjee et al. [39] who mention values between 1.5 and 4.7% in a review of pineapple by-products, while Rico et al. [36] report ranges between 0.38 and 5.9%. On the one hand, although the variety of the fruit may influence the ash content, according to Lombardi-Boccia et al. [40] the main factor that influences the concentration of minerals in the vegetable products is the type of soil used for the culture. In contrast to the prior, Zakaria et al. [24] mention that freshest foods rarely have an ash content higher than 5%. Fruits and their juices generally contain 0.2-0.6% (w/w) of ash, while the value is higher in dried form [2.4-3.5% (w/w)].
Regarding the fiber content, values of (14.72 ± 0.03%) were found, which is a higher value than the one reported by Aparecida Damasceno et al. [41] of (4.92 ± 0.5%) and similar to the reported by Aruna [20] of 13.96 ± 0.14. According to Huang et al. [42], the fractions rich in fiber in pineapple peels are capable of converting biomass into functional foods with high added value. It is observed that this raw material presented a high content of carbohydrates (76.41%), which could be attributed to its high content of sugars (glucose and fructose), as proven by the results reported in Table 3. Similarly, this value was close to that reported by Aparecida Damasceno et al. [41] of 75.63% and Morais et al. [4].
The pH reported by Campos et al. [35] was 3.63, a similar value to the one determined (4.04 ± 0.013). This result was also close to the value reported by Selani et al. [6] of 3.86 in pineapple pomace, 4.6 ± 0.02 in pineapple peel flour [41], 4.3 in pineapple residues [43], 4.0 in pineapple residue silages [44] and 4.6 [37]. The reported differences in pH are related to the stages fruit ripening, the cultivar, difference in geographical location; in addition, the low pH value of this substrate provides a lower risk of contamination by microorganisms, enzymes or non-enzymatic reactions [41]. Regarding the protein content, the result was similar to that reported by Romelle et al. [45], who reported 5.1 g protein/100 g of dried matter in their proximal analysis of pineapple peels. For crude fiber, Rico et al. [36] reported a range between 16 and 65%, therefore the calculated value is within this interval. Furthermore, the total and reducing sugars analysis was conducted out by liquid chromatography [Reference AOAC, 18th ed: 996.04 (Modified)] (Fig. 1), finding the results that are presented in Fig. 1.
According to the results reported in Fig. 1, it is possible to observe that the predominant sugars in the sample were glucose, fructose and sucrose. Authors such as Jusoha et al. [46] and Sepúlveda et al. [47] have also reported the presence of these sugars in pineapple waste, it is important to remark that the concentration of these sugars varies according to the ripeness of the fruit.
The concentration of fructose (13.9 g/L) in the analyzed samples was higher than glucose (9.69 g/L), which coincides with that reported by Abdullah and Mat [43] who quantified values for fructose and glucose of 12.17 and 8.24 g/L, respectively. The sucrose concentration was lower compared to that of glucose and fructose, which could be attributed to the acidic pH during the drying process [37]. Abdullah and Mat [43] also mentioned that an acid pH (4.0) during this process generates the conversion of sucrose into glucose and fructose, and a color change is observed in the samples from yellow to brown, showing evidence of this forms the polymerization and degradation of sucrose. It is important to mention that the differences observed in the physicochemical composition of the analyzed variables with respect to the values reported in the literature can be attributed to the different cultivars, cultivation conditions, variety, as well as the composition of the soil and fertilization [41].
The amount of water that can absorb the solid-state fermentation support material is denominated WAI [48,49]. On the contrary, the critical moisture content refers to the water that is bound to the substrate and cannot be used by the microorganism for metabolic functions during the fermentation process [15]. The pineapple peel had WAI and CHP values of 5.61 g/g of dried peel ± 0.03 and 14.37% ± 0.08. The WAI values found were similar to or higher than those reported by Xiao et al. [50] in fermented chickpea (3.65-3.88 g/g dry material), Robledo et al. [51] for pomegranate husks (4.84 ± 0.006 g/g dry material) and pomegranate seeds (1.62 ± 0.020 g/g dry material), Buenrostro-Figueroa et al. [48] for coconut husk (12.09 g/g dry material), sugarcane bagasse (9.46 g/g dry material), candelilla stalks (3.14 g/g dry material) and corn cobs (2.97 g/g dry material) and Torres-Leon et al. [15] for mango seed (3.4-4 g/g dry material). In the case of CHP, Buenrostro et al. [49] reported a lower value in fig by-products (4.63%), while [34] and [48] reported close values in avocado seeds and coconut husk (14.33% and 16%). According to the above, the results of WAI and CHP suggest that pineapple peel is a great substrate to be used as a support in the solid-state fermentation process.

Radial growth velocity
Following the methodology proposed in section "Determination of radial growth velocity", the radial growth of the Rhizopus oryzae fungus was recorded for 2 days. The measurement of the growth rate of filamentous fungi indicates their ability to colonize some substrate used in solid fermentation processes.
The curve shows the general trend of the radial growth of the microorganism, the Rhizopus oryzae fungus grows on the substrate and has a short adaptation period (14 h) followed by a rapid growth that corresponds to the exponential growth phase. With the radius registered values, the average speed of radial growth was obtained, also considering in the calculation of this the period of adaptation of the microorganism to the environment. According to the results found, the radial growth speed of the microorganism Rhizopus oryzae was 0.77 ± 0.004 mm/h; similar value reported by Londoño [29], who used this microorganism in a SSF process with the aim of reducing tannins in sorghum.

Scanning electron microscopy (SEM)
The structural changes in the peel samples before and after the fermentation process were analyzed by SEM images. The images were observed at ×400 and ×1500 magnification. The open and porous structure of the dehydrated peel can be appreciated in Fig. 2a, where a structure suitable and conducive to mycelial attachment of the fungus can be appreciated. It is also possible to appreciate irregular morphologies with different particle sizes, rough surfaces and with large pores.
Similarly, a dense structure is observed, because the components such as pectin, lignin and hemicellulose were intact. Structures with similar morphologies have been reported by Zhang et al. [52]. On the other hand, Fig. 2b shows the adherence of the fungus to the substrate after the fermentation process and that the lignocellulosic components have undergone breaking processes.
As shown in Fig. 2b, Rhizopus oryzae demonstrated its ability to adhere to the peel using its mycelia. The network of mycelia and spores was completely distributed on the peel, which shows that this mycelium was able to adapt and successfully populate this structure. The mycelial hyphae are responsible for releasing hydrolytic enzymes to break down the cell wall and release the bound phenolic compounds. These images are similar to the observed by Zain et al. [37] in their work on solid-state fermentation of pineapple waste to produce lactic acid.

Evaluation of solid-state fermentation conditions (SSF)
Following the methodology described in section "Microstructural analysis", the assembly was conducted to evaluate the appropriate parameters of the fermentation process. The results, according to the treatment for each of the variables studied during the solid-state fermentation process (total phenol content, DPPH and ABTS radical elimination activity) are recorded in Table 2.
The highest value (83.77 mg GAE/g d.w) for the total phenolic content (TPC) was provided at a temperature of 36 °C, pH 6 and 80% moisture content. The values of condensed polyphenols (66.5 mg QE/g) determined in this study were much higher than those previously determined by Ordoñez-Torres et al. [53] in samples of mango peel subjected to ultrasound (18.21 mg QE/g) and by García et al. [54] in Ataulfo mango peel (3.7 mg QE/g). It was also possible to conclude that the value of condensed tannins was higher in the fermented sample than in the control sample (40.32 mg QE/g). For DPPH radical scavenging activity, the highest value was 59.76% inhibition, under a pH of 4.5, temperature of 30 °C and 60% of moisture content. Finally, for ABTS radical scavenging activity it was 73.66% inhibition, with a pH of 6, moisture content of 60% and temperature 36 °C, conditions similar to TPC.
The fermented samples presented higher total phenolic content compared to the control sample. The main phenolic compounds present in pineapple peels are flavonoids, derived from benzoic acids and ferulic acid [10]. These components can be found in chain form or can be part of complex structures such as hydrolyzable tannins or lignins; they can also bind to structural components of the cell wall such as lignin, cellulose and hemicellulose [55]. The most soluble phenols are found inside cell vacuoles, in free or conjugated form; while insoluble phenols are attached to cell wall structures, esterified with arabinose or galactose residues from pectic components or hemicellulose [56]. These compounds can be formed through the disintegration of the bonds between cellulose, hemicellulose and lignin [57].
In the process of fermentation of pineapple peels, the increased content of phenolic acids is mainly due to the fact that R. oryzae is an excellent producer of hydrolytic/cellulolytic enzymes such as endoglucanases, xylanases, cellobiohydrolases, which degrade lignocellulosic components, thus releasing the conjugated phenolic compounds in the cell walls by hydrolysis. Therefore, these compounds become soluble, generating an increase in the total phenolic content in the extracts obtained [58]. This is confirmed by Martins et al. [28], who report that filamentous fungi are producers of enzymes that help lignin break down, and also have two extracellular systems: the first produces carbohydrolysis (extracellular system) and the second helps degrade the phenyl rings, increasing the free phenolic content (ligninolytic oxidative system). This observed behavior is in accordance with that reported by Janarny and Gunathilake [59] who carried out SSF in four varieties of rice bran (Bw367, Bg352, Bg406 and H4). Using Rhizopus oryzae, identifying that this process increased the total phenolic content (TPC), total flavonoids (TF), total carotenoids (TC) and total anthocyanins (TAC) in two of the varieties analyzed (Bw367 and Bg352). The same behavior was observed by Leite et al. [60], who obtained higher total phenolic content in fermented samples of grape residues, olive oil and beer compared with the unfermented samples,being the maximum increase in TPC in grape residues fermented by Aspergillus ibericus (2.9 times) in regard to the unfermented sample. Similarly, Dulf et al. [13] studied the release of phenolic compounds in wild "chokeberry" cherries by SSF with Aspergillus niger, observing an increase in the extraction of phenolic compounds of 1.7 times compared to the unfermented samples. In addition, in a study carried out previously, obtained a 21% increase in the TPC of the fermented extracts compared to the control sample (without fermenting) [61]. Sari et al. [62] also concluded that the fermentation process using pineapple by-products (peel and core) with the bacteria Acetobacter xylinum enhanced the total phenolic content and antioxidant activity, arguing that this behavior could be attributed to the fact that, during the fermentation process, A. xylinum grows best in an acid medium by converting glucose into gluconate acid and fructose into acetic acid, presenting an increase in the concentration of organic acids contained in the medium, thus increasing phenolic compounds and also antioxidant activity.
In order to obtain the models that would allow to find the best relationship between the variables moisture content, temperature and pH that effectively predicted the total phenolic content, the radical elimination activity DPPH and ABTS as response variables, a design of response surface (central compound), which consists of 20 points corresponding to the combinations of the treatments of the design 2 3 , 6 central points that are constituted by a treatment, the middle point between the highest and lowest level of each variable of the process, achieving the estimation of the experimental error and 6 axial points that allow the incorporation of the quadratic terms in the model when the relationship between the response variable and the study factors require it. The p-value was calculated for each type of model, selecting the one with the highest order with significant terms (p < 0.05).
To determine the quality of the model, the correlation coefficient R 2 was found, and the root of the mean square error (RMSE) was estimated; the first is the determination coefficient and shows the good fit of the model to the data, it oscillates in a range of 0-100%; in general, the larger the R 2 , the better the model fits the data. The second indicates the absolute fit of the model to the data, how close the observed data points are to the predicted values of the model, low values indicate a better fit [63]. The results of the ANOVA analysis of variance for each of the response variables are presented in Table 3. Relatively to the total phenolic content, the results show that the linear model presents a significant effect p < 0.05 on the response variable; however, it can be observed that the quadratic model and all its terms did not present a significant effect p > 0.05, as well as the interactions. Because the p-value of the moisture-pH interaction was so high (0.737), that is, it did not exert a significant effect, it was decided to eliminate it to increase the lack of fit of the model. Therefore, for the adequate prediction of this response variable, all linear and quadratic terms and the temperature-moisture content and temperature-pH interactions were considered. This was confirmed with the R 2 value of 0.84, indicating that the selected terms explain 84.00% of the response variability and presenting a strong correlation (0.7 < R 2 < 1) [64]. The root mean square error was 5.396, which indicates that there is no significant difference between the values obtained and those predicted by the model. Correspondingly, for the% inhibition of DPPH radicals, the quadratic model presented a significant effect p < 0.05, except for moisture content (0.291). Likewise, the linear model and the interactions did not present a significant effect; however, the moisture content factor did present a significant effect in the linear model (0.044) and also the temperature-pH interaction (0.040). In this case, all the terms of the variables studied were taken into account to guarantee the quality of the model. The value obtained from R 2 was 0.8148, indicating that the selected terms explain the variability of the response in 81.48%. Similarly, the RMSE value was 3.517 demonstrating a good fit of the model to the experimental data. Finally, it was found for the% inhibition of ABTS radicals that most of the terms exert a significant effect on the response variables given that p < 0.05. The only terms that did not show a significant effect were moisture content in the linear model (0.644) and the interaction moisture content by pH (0.070); nonetheless, it is not possible to exclude them because they affect the precision of the model. Therefore, all terms are considered. The value of R 2 was 0.9604, being considered a good fit and thus, can represent the variable studied effectively.
With the results of the ANOVA table, the models were proposed for each of the response variables studied, considering the coefficients of the temperature (T), moisture content (CH) and pH components. The equations obtained were recorded in Table 4.
To properly observe the behavior of each of the variables studied, contour graphs were made (Figs. 3, 4). These graphs allow us to observe the regions in which the lowest or highest values of each of the variables occur, through a variation in the tones, which allows identifying the region of interest and predicting at what levels the values should be factors (temperature, moisture content and pH) to find the desired response values.
In Fig. 3, the contour graph for the behavior of the total phenolic content under the different combinations of the analyzed factors is observed. The blue color represents the area where the lowest TPC values occur after the fermentation process, and the green color represents the area where the highest values occur. It is possible to appreciate that the areas where the highest values are obtained are those where the pH, temperature and moisture content factors are higher.
In the case of pH, it was observed that there was an increase in the content of total phenols. This behavior is the same as that reported in the research by Chinnarajan et al. [65] who performed SSF using the fungus Chaetomium globosum, under alkaline conditions, using sugarcane bagasse and cotton seeds as substrate, with the objective of obtaining phenolic compounds. The authors reported that "alkaline stress is a determining factor in the production and release of these compounds". According to their results, there was an increase of 500 and 620% of gallic acid in cotton seeds and sugarcane bagasse. Reporting a linear correlation between the increase in pH and the amount of gallic acid due to the best conditions for the enzymatic activity generated by the effect of pH, improving the hydrolysis of the substrate and increasing the production of the compound. Likewise, when observing the interactions between pH-temperature and pHmoisture content, it is possible to see that the highest TPC occurs from a pH of 5.5, the highest value being at a pH of 6.5. This value is similar to that reported by Aziman et al., [66] who used R. oryzae in solid state fermentation in order to produce lactic acid, finding that the most adequate conditions were, temperature of 27 °C, moisture content of the 80% and pH of 6.5.  Regarding the temperature, it was observed that when it is increased, also the TPC did. It is important to mention that an enhance in temperature generate an acceleration of the movement of molecules and the dissolution of phenolic compounds, augmenting their concentration and their ability to eliminate free radicals. Correspondingly, temperature is also a determining parameter in microbial growth as well as in the development of cell wall degrading enzymes [67]. Most of the fungi grow between pH 5 and 6 [68], particularly R. oryzae grows at temperatures of 7-45 °C, with an optimum close to 37 °C [69,70] which is consistent with the observed behavior, since a range between 28 and 38 °C was analyzed, and the highest values are observed from 36 °C.
In the case of moisture, a proportionality was also observed. This parameter plays an important role during the solid-state fermentation process since microorganisms have different moisture content requirements [67]. Sufficient moisture content is necessary to achieve adequate nutrient transfer for ideal fungal growth and metabolism, substrate swelling and enzyme stability for lignocellulosic breakdown  [71]. In Fig. 6 it is possible to analyze that the higher the moisture content the TPC increased, specifically from 70% (w.b). This value is similar to that reported by Benabda et al. [72] who carried out SSF on bread residues using R. oryzae to increase the production of proteases and amylases, concluding that the best conditions were: a moisture content of 65%, 120 h of incubation and an inoculum concentration of 10 5 spores/g. In the same way, this behavior was observed by Aziman et al. [66] who evaluated a moisture content range between 60 and 80% (w.b), for the production of lactic acid by SSF, concluding that the best conditions were given to the highest evaluated moisture content. Figure 4 shows the contour graph for the behavior of the% inhibition of DPPH radicals under the different combinations of the factors studied. The light green color represents the area where the lowest inhibition values occur, while the dark green color represents the area where the highest values occur. It is possible to observe that the areas where the highest values are obtained are in the moisture content-temperature and pH-temperature interaction. The heating process can trigger the degradation of natural antioxidants present in honey gold pineapple, accelerating their oxidation. It is observed that, at higher temperature values, a higher moisture content is linked, and the inhibition percentage increases. Oxidation processes can occur faster in the presence of heat, alkaline conditions, enzymes, oxidizing agents and by iron and copper catalysts. Moreover, heating at temperatures above 60 °C for 30 min can reduce the antioxidant capacity of some fruits by 10% Sari et al. [62]. However, the ranges evaluated do not exceed the optimum of the microorganism (37 °C). It is important to mention that temperature control is one of the critical criteria in solid state fermentation systems due to the static nature of the incubation conditions. Still, an increase in the temperature can be generated due to the heat released during fungal growth, causing a physiological alteration of the microorganisms and consequently affecting the metabolism of the compounds produced as well as their production [37,73]. However, this variable cannot be analyzed independently; but, in addition to the pH and moisture content parameters that also influence the production of the compounds of interest.
Particularly from pH greater than 5.5, higher percentages of inhibition were presented. The pH of the media is an important parameter closely related to microbial growth, enzyme secretions, and the stability of phytochemicals. Furthermore, changes in pH during the solid-state fermentation process could influence the structures and concentrations of phenolic compounds, as well as antioxidant activity due to changes in the ability to donate electrons [67]. Figure 5 shows the contour plot for the behavior of the% inhibition of ABTS radicals under the different combinations of the studied factors. The light green color represents the area where the lowest% inhibition values occur, while the dark green color represents the region where there is the highest inhibition of ABTS radicals. It is possible to notice that the areas where higher values are obtained occur at higher values of temperature, moisture and pH. It is important to highlight that if the moisture content is high, the production of compounds such as acids decreases, this could be attributed to the agglomeration of the particles and the reduction of gas diffusion on the surface. In contrast, if the moisture content is very low, there will be a limitation in the transfer of nutrients, altering the stability of the enzymes, the low supply of oxygen (water) and these conditions directly affect the growth of the fungus resulting in a decrease in the production of bioactive compounds [74][75][76]. Most mushrooms need a relatively low moisture content, between 30 and 60% [67]. In the case of the pineapple peel used as substrate, the results of water retention capacity allowed to conclude that the critical moisture content was 84 ± 1.05%, for which the highest value evaluated was 85% (w.b), a value higher than that could not be analyzed because it would favor the growth of other types of microorganisms.
According to the results observed in Figs. 3, 4, 5, it is possible to analyze that as the pH increases, the antioxidant activity increases, reflecting this in higher values for the content of total phenols, activity radical DPPH and ABTS scavenger. This behavior coincides with that reported in the work carried out by Chin-Hang and Ming-Yeou [77] who evaluated the effect of the pH of the culture (3-6) on the performance of bioactive compounds and antioxidant activity of methanolic extracts of mycelia and Antrodia camphorata filtrates, observing that a low culture pH (3.0) favored cell growth, but a higher culture pH (5.0) favored the antioxidant properties, this being the pH where the maximum production of antioxidants (55%) and a DPPH radical scavenging activity of 75% at a concentration of 0.6 mg/mL of filtered methanolic extracts.

Analysis of the control sample with respect to the treatments under SSF
The control sample presented a total phenolic content of 30.0 ± 0.93 mg GAE/g d.m, which is a value similar to that reported by Benites et al. [78] who performed the analysis on lyophilized pineapple peel powder. Golden variety, finding a value of 34.39 ± 0.89 mg GAE/g d.m. Similarly, it is close to that reported by Kuppusamy et al. [79] who carried out polyphenol extraction in pineapple peels using different solvents: water, methanol and ethanol, registering 68.3 ± 1.2, 28.6 ± 1.5 and 22.3 ± 0.6 mg GAE/g d.m, respectively. Lourenço et al. [80] carried out the encapsulation of pineapple peel extracts that were obtained using a water: ethanol ratio (20:80 w/w), finding a TPC of 29.33 mg GAE/g d.m, value close to reported. The lowest value observed in the literature was that of Azizan et al. [81] of 10.73 mg GAE/g d.m in MD2 variety pineapple peel extracts. However, the inhibition percentage of DPPH radicals was 37.02%, which is similar to the value found in this work of 37.12%. Regarding the ABTS + radical elimination activity, the control sample presented (5.64 µmol/g d.m), which is a similar value to that reported by Kuskoski et al. [82] who evaluated this property in lyophilized pineapple products, obtaining a value of 3.4 µmol/g d.m Authors such as Martínez et al. [83] reported a lower value (1.7 ± 0.2 µmol/g d.m) in dehydrated pineapple by-products (peel and core).
The results obtained by Duhan et al. [84] and Chawla et al. [85] show that fermentation processes enhance antioxidant activity in different substrates. Saharan et al. [86] found a positive correlation between antioxidant activity and phenolic content through a comparative analysis where they evaluated the effect of fermentation on phenolic compounds, antioxidant activity and flavonoids in commonly used cereals. In the same way, Sadh et al. [87] analyzed that there was an increase in antioxidant activities and phenolic content when the fermentation process was applied, and at the same time some rheological and enzymatic properties increased.
In order to evaluate whether the solid-state fermentation process presented a statistically significant difference (p < 0.05) with respect to the control sample (unfermented peel), a single factor ANOVA test was executed for each of the response variables (total phenolic content, DPPH and ABTS elimination activity) and, as a factor, the treatments carried out. According to the results, if there was a statistically significant difference for each variable with respect to all the treatments since (the p-value = 0.000). For all response variables, the two treatments in which the highest values were obtained were selected, which were T7 and T8, the first corresponds to (T = 36 °C, MC = 60% and pH 6) and the second (T = 36 °C, MC = 80% and pH 6), finding statistically significant differences (p < 0.05) (Fig. 6a, b).
As can be seen in Fig. 6a, there was a significant difference between the control sample and treatments 7 and 8 (p < 0.05), observing that the solid-state fermentation process enhanced the total phenolic content in 106.7 and 176, 2% respectively. This comportment was comparable to that observed by Santos et al. [88] who used SSF in dehydrated granadilla seeds using Aspergillus niger, finding that the total phenolic content increased 236.5% compared to the unfermented samples and using 80% acetone as extraction solvent. In addition, Feitosa et al. [89] performed SSF on moringa leaves using A. niger. Their results showed an increase of 136.4% in the total phenolic content and 783.1% in the total flavonoid content. In the study performed by Schmidt et al. [90], they described that R. oryzae improved the content of phenolic compounds by 110%. The content of chlorogenic acid, vanillin and p-hydroxybenzoic acid enhanced during the fermentation process and the highest increase was quantified in ferulic acid (764.7 mg/g d.m) after 120 h. The researchers mentioned that the fungus Rhizopus oryzae is a producer of some enzymes that break the cell walls of the rice bran causing the release of ferulic acid.
In the case of Fig. 6b, it is possible to observe that there were statistically significant differences (p < 0.05) with respect to the control sample, and that the inhibition percentages were higher in the fermented samples. There were not any significant differences between treatments 7 and 8 (p > 0.05), but in terms of the control sample (p < 0.05). The enhance in antioxidant compounds after the fermentation process is closely related to the action of the enzymes produced by the fungi, which convert the phenolic compounds present into phenolic aglycones, which have higher antioxidant activity [91]. Thus, the increase in phenolic compounds and consequently the antioxidant activity can be attributed to the hydrolytic enzymes produced by R. oryzae, which act on the peels and increase the access of hydroxy functional groups to phenolic compounds. This behavior improves the number of phenolic groups and thus the antioxidant capacity [92]. To optimize the process, the desirability function (d) was applied, which is represented in graphs. The d value is equivalent to the degree of desirability of the response, values close to 1 indicate that the response is desirable. To apply this function, we intended to maximize the value of total phenolic content and antioxidant activity (DPPH and ABTS). The optimization values found were: a temperature of 37.3 °C, moisture content of 85% and pH 5.5, obtaining a value for the total phenolic content of 52.70 mg GAE/g d.m, 61.46% of inhibition for DPPH and 77.39% for ABTS.
Temperature is a very significant parameter in the control of the SSF process, since it influences the growth, germination, sporulation and metabolism of the fungus. Therefore, an optimal control of the temperature of the microorganism will provide maximum production of the products of interest [29,93]. The growth temperature varies between the species and the substrate in which they are growing, resulting for this study a temperature of 37.3 °C, which matches with the optimal growth temperature of R. oryzae. Furthermore, authors such as Kitpreechavanich et al. [94] reported that this microorganism has the Fig. 6 a Comparison of the total phenolic content in a control sample with respect to the two best treatments (T7 and T8). b Comparison of the DPPH and ABTS radical elimination activity in a control sample with respect to the two best treatments (T7 and T8). The means with different capital letters within a column (a-c) were significantly different (p < 0.05) ability to resist and grow up to 45 °C while at 34 °C they observed the highest mycelial growth. Huang et al. [95] evaluated the effect of temperature on lactic acid synthesis using R. arrhizus and R. oryzae on potato starch samples, showing an increase in hydrolysis of the sample at 40 °C. For its part, moisture content is also an influencing factor in the process, and for this case the optimum value was 85%, which corresponds to the maximum absorption content that the substrate can withstand, according to the support tests carried out. Adequate moisture content helps microbial growth, but too much could block oxygen penetration and hinder the process [96]. In the case of pH, a maximum value is observed for the three variables at 5.5 and then a decrease; this is also an important factor in the process, because the changes generated during the growth of the microorganism will directly affect the synthesis and stability of the products created. Ferreira et al. [97] found that the optimal activity pH of most of the enzymes produced by this fungus is between 4 and 5.5, which coincides with the optimization value. Moreover, the results of this research highlight the ability of the Rhizopus genus strains to synthesize active enzymes at alkaline pH.

Identification of bioactive compounds by HPLC
Measurements were performed on the samples (fermented pineapple peel, 60:40 ethanol:water). To identify the phenolic compounds, a comparison between spectra and retention times of standards was performed. Employed the area method, quantification was performed at a wavelength of 280 nm. However, some compounds could not be quantified because they were below the limit of quantification (BQL) ( Table 5). Overall, the major phenolic groups in the peel were composed of phenolic acids. Eight peaks were identified in the chromatogram, five of which were quantifiable: gallic acid, chlorogenic acid, catechin, epicatechin and p-coumaric acid. These compounds are similar to those reported by Li et al. [10], who identified in methanol extracts of pineapple peels (gallic acid, epicatechin, catechin, and ferulic acid). Nonetheless, two of the reported compounds were not quantified in this work (catechin and ferulic acid). Yahya et al. [98] identified in ethanolic extracts of pineapple peels: catechin, gallic acid and quercetin, compounds identified in this analysis. Yapo et al. [99], identified eight phenolic compounds, all belonging to the family of hydroxybenzoic and hydroxycinnamic acids. In the work by Campos et al. [35] where there were found the compounds present in liquid residues of pineapple peels, they reported chlorogenic acid, caffeic acid, syringaldehyde and ferulic acid. The compounds identified are consistent with four of the five compounds quantified by Lourenço et al. [100], who established: gallic, chlorogenic, caffeic, p-coumaric and ferulic acid in samples of pineapple peel extracts with water and 80% ethanol.

Conclusions
The research results reveal the potential of Rhizopus oryzae to generate enzymes that enable the hydrolysis of phenolic compounds bound to the cell wall of pineapple peel, increasing antioxidant activity. Optimal process conditions were a moisture content of 85%, pH 5.5, temperature of 37.3 °C for 22 h. The solid-state fermentation process increased the TPC in all treatments with respect to the control sample (unfermented peels) in which there was an increase of 176.2%. HPLC analysis allowed the identification and quantification of five phenolic compounds in pineapple peel: gallic, chlorogenic, caffeic, epicatechin and p-coumaric acids, which have multiple health benefits due to their antimicrobial, antioxidant and anticancer activity. The pineapple peel extract presents a wide field of application that includes food, pharmaceutical and cosmetic industries due to presence of the bioactive compounds identified.