LWT - Food Science and Technology 149 (2021) 111809 Contents lists available at ScienceDirect LWT journal homepage: www.elsevier.com/locate/lwt Novel solid-state fermentation extraction of 5-O-caffeoylquinic acid from heilong48 soybean using Lactobacillus helviticus: Parametric screening and optimization Nelson Dzidzorgbe Kwaku Akpabli-Tsigbe a,b, Yongkun Ma a,*, John-Nelson Ekumah a,b, Juliet Osabutey c,d, Jie Hu a, Manqing Xu a, Nana Adwoa Nkuma Johnson a a School of Food and Biological Engineering, Oversea College of Education, Jiangsu University, 301#, Xuefu Road, Zhenjiang, 212013, Jiangsu, PR China b Department of Nutrition and Food Science, College of Basic and Applied Sciences, University of Ghana, P. O. Box LG 134, Legon, Ghana c Department of Early Childhood Education, University of Education, P. O. Box 25, Winneba, Ghana d Virtuous Experimental School, P. O. Box AH 106, Achimota-Accra, Ghana A R T I C L E I N F O A B S T R A C T Keywords: This study investigated the extraction of 5-O-caffeoylquinic acid (5-CQA) with increased yield and enhanced 5-O-caffeoylquinic acid (5-OCQA) antioxidant activity from heilong48 soybean (HS) under solid-state fermentation (SSF). Plackett–Burman design Solid-state fermentation and Box-Behnken design were sequentially used for screening and optimization of significant SSF conditions Soybean respectively. Screening results showed that temperature, pH, incubation time and liquid-solid ratio were the Screening Optimization significant SSF conditions that influenced 5-CQA yield, fermentation efficiency and antioxidant activity. The optimum SSF conditions obtained by Box-Behnken design were 49.90 ◦C (temperature), 7.00 (pH), 25.81 h (incubation time) and 0.67 (liquid-solid ratio). For these conditions, the experimental data obtained [5-CQA yield (11.41 ± 0.27 mg/g), fermentation efficiency (30.49 ± 1.14%), and antioxidant activity (46.13 ± 1.94 μmol AA eq/g dry sample)] were consistent with predicted values, higher than that of unfermented HS flour (RSHF), and supported by Atomic force microscopy (AFM), Fourier transform infrared (FTIR) and Scanning electron microscopy (SEM) microstructure. The results demonstrated that optimized SSF conditions significantly influenced 5-CQA yield, fermentation efficiency and antioxidant activity. This study showed that the use of optimized SSF conditions to extract 5-CQA with increased yield and enhanced antioxidant activity was efficient. Hence, this could be useful to the food and/or pharmaceutical industry in producing 5-CQA from HS. 1. Introduction Solid-state fermentation (SSF), a type of fermentation, is a cost- effective and green technique with much attention received for its The concept of extracting and processing plant bioactive components processing and biological advantages relative to submerged and liquid into useful substances/products for human utilization is profitably fermentation (Ang, Ngoh, & Chua, 2013). SSF, alternative to submerged agreeable. Hence additional research in the fields of food science and fermentation (with greater advantages than submerged fermentation in engineering, biotechnology and nanotechnology, on this subject, is various processes) is widely used for production of products with added worth looking at (Verduzco-Oliva & Gutierrez-Uribe, 2020). One most values namely enzymes, single cell protein, antibiotics, important technology in the mentioned areas/fields that cannot be left poly-unsaturated fatty acids, organic acids, aroma, biofuel and bio- out if bioactive ingredients of plant are to be harnessed for human pesticides (Bhargav, Panda, Ali, & Javed, 2005). SSF has numerous benefits is fermentation. Fermentation is an ancient biotechnology and biotechnological benefits such as higher product stability, reduced classic industrial process for improving the shelf-life, nutritional and catabolic repression, higher fermentation or volumetric productivity, organoleptic qualities of food (Magro, Silva, Rasera, & de Castro, 2019). lower demand on sterility, less effluent generation, higher concentration It also increases the release of biologically active compounds having of end-products, use of water-insoluble substrates specific microorgan- antidiabetic and antioxidative activities (Magro et al., 2019). isms, simple fermentation equipment requirement (Kapilan, 2015) as * Corresponding author. E-mail addresses: ndkakpablitsigbe@outlook.com (N.D.K. Akpabli-Tsigbe), mayongkun@ujs.edu.cn (Y. Ma). https://doi.org/10.1016/j.lwt.2021.111809 Received 9 March 2021; Received in revised form 20 May 2021; Accepted 24 May 2021 Available online 26 May 2021 0023-6438/© 2021 Published by Elsevier Ltd. N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 well as environmentally friendly and more cost- and energy-effective (Zhang et al., 2014). However, using lactic acid bacteria for extraction (Sitanggang, Sinaga, Wie, Fernando, & Krusong, 2020). It is therefore a of 5-CQA from soybean under SSF has not been investigated. As a result, promising bioconversion technology for pant products valorization into the present study sought to screen and optimize SSF parameters for high value-added products. extraction of 5-CQA with increased yield and improved antioxidative Fermentation in previous years was done to increase bioactive activity from heilong48 soybean (HS) variety using L. helviticus strain. phenolic compounds content in legumes, consequently improving their Plackett–Burman design and Box-Behnken design were sequentially antioxidant activity (Bartkiene, Krungleviciute, Juodeikiene, Vidman- used to achieve the study objective. Plackett–Burman design was used tiene, & Maknickiene, 2015). This bioprocess has been researched as an for the screening of the SSF conditions to identify the significant SSF efficient method for the extraction and production of biologically active conditions that influence 5-CQA extraction and Box-Behnken design compounds in food lately (Handa et al., 2019). Soybean (valuable used for the optimization of the significant SSF conditions obtained for legume worldwide) is a high nutritional, economic and suitable sub- extraction of 5-CQA with increased yield and enhanced antioxidant strate for SSF utilization for numerous applications to produce activity. Second-order polynomial model was developed for the value-added foods and antioxidant compounds (Correa Deza, Rodríguez extraction process and evaluated using analysis of variance for the de Olmos, & Garro, 2019). Hu et al. (2019) reported that fermentation of model credibility. Structural analyses [Atomic force microscopy (AFM), soybeans decreases antinutritional factors, lipoxygenase, urease and Fourier transform infrared (FTIR) and Scanning electron microscopy trypsin inhibitor activities. Products of fermented soybean are high in (SEM)] were performed to affirm the effectiveness of the optimized SSF antioxidative activities (Yang et al., 2019), and more attention given to conditions on the degradation of the cell wall of heilong48 soybean (HS) those with high nutrition and health benefits (Bartkiene et al., 2015). variety to release more 5-CQA. The products obtained after whole soybean SSF are directly lyophilized with no centrifugation. Thus, SSF of whole soybean is a more 2. Materials and methods cost-effective, simple technology for probiotics carrier food production. Soybean has been fermented to produce specific foods that contain HS variety was purchased from Tianxia Agricultural and Sideline phenolic antioxidants with related consumer good health and wellbeing Products and Distribution Department, China. L. helviticus LH-43 was (Handa et al., 2019). bought from Synbio Tech Inc., Taiwan. It was stored at 4 ◦C until use. One important phenolic acid with many health benefits, obtained Only analytical grade chemicals were used in this study. from soybean, but least investigated is 5-O-caffeoylquinic acid (5-CQA) (Nabavi et al., 2017; Naveed et al., 2018). 5-CQA has tremendous 2.1. HS flour preparation application in food, pharmaceutical and cosmetic industries. It has been the focus of interest due to its putative health benefits and impact on HS variety was milled with a hammer crusher (FC160, Shanghai food quality. Due to the numerous health benefits of 5-CQA, its demand traditional Chinese medicine machinery factory, China) and further is on the rise; however, the cost of production is high, limiting its sieved into fine flour of particle size 0.25 mm. The final flour was packed availability for human benefits. The reason for this is that, existing in air-tight low density polyethene bags in weights of 150 g and stored works on 5-CQA production/extraction focused on using coffee (an (− 20 ◦C) for further studies. expensive cash crop which is not available all-year round) as the raw material. Also, conventional methods that use organic solvents (e.g. 2.2. Inoculum preparation chloroform, dichloromethane, etc.) are commonly used to extract 5-CQA. These solvents, however, are dangerous to handle and harmful L. helviticus LH-43 was activated by subculturing twice in de Man, to human health and the environment (Torres-Mancera et al., 2013). In Rogosa and Sharpe (MRS) broth at 37 ◦C for 24 h (Zhou et al., 2019). The addition, the conventional methods (that make use of dichloromethane, culture was centrifuged using RJ-TDL-50A centrifuge (Ruijiang Analyt- methanol, ethanol, acetone extraction, etc.) in extracting 5-CQA are: ical Instrument Co., Ltd., China) at 4000×g for 10 min. The supernatant time-consuming, relatively high in solvent usage, often unsatisfactory in was discarded and the bacterium cells washed in sterile saline (0.1% reproducibility and poor in the extraction of polar substances (Wia- NaCl) solution. An XB-K-250 hemocytometer (Jianling Medical Device nowska & Gil, 2019). As a result, the food, pharmaceutical, and cosmetic Co., China) was used to measure the inoculum concentration and cor- industries are lately searching for rich and cost-effective plant sources rected to 109 CFU/ml with 0.1% sterile NaCl solution. The obtained (for 5-CQA) and also efficient extraction techniques (Wianowska & Gil, suspension was used as starter culture for SSF. 2019). That notwithstanding, existing research on soybeans (a cheap crop available all-year round) are limited to isoflavones, even with the 2.3. SSF of HS variety new varieties with improved qualities. Till date, the only literature on soybean in connection with 5-CQA dates back to 1979, where it was only Sterile distilled water was added to 10 g HS flour (on dry matter reported as a source of 5-CQA (Pratt & Birac, 1979) with no further basis) to attain different moisture contents (20, 30 and 40%) in a conical work/data on content quantification or extraction. flask (250 ml). The contents were thoroughly mixed and sterilized for SSF (most efficient bioprocess) could be an efficient approach to 15 min at 121 ◦C (Li et al., 2020). After cooling to 25±2 ◦C, the mixture increase the release of 5-CQA from soybean. SSF enhances the phenolic was inoculated with 1, 3 and 5% inoculum of L. helviticus with 109 content in plant extracts via the breakage of ester bonds between the CFU/g cell population. This was followed by thorough mixing and plant cell wall and phenolics, increasing their concentration and hence culturing at different temperatures (30, 40 and 50 ◦C) in an incubator functional properties (Santos da Silveira et al., 2019). Due to the low (SPX-250, Jintanshizhongdayiqichang, China) for 0, 24 and 48 h under water availability in SSF, a limited number of microorganisms are used. static aerobic conditions. The pH of the SSF of HS variety was adjusted Filamentous fungi are considered the most desirable microbes for SSF through the addition of the 1 N NaOH or 1 N HCl to the culturing me- followed by yeasts and moulds (Santos da Silveira et al., 2019). Though, dium (Adnan, Ashraf, Khan, Alshammari, & Awadelkareem, 2017). All in nature, filamentous fungi and bacteria typically grow on solid sub- fermented samples were stored at − 20 ◦C for further investigations. strates (leaves, roots, seeds, stems and wood of plants) in symbiotic associations (Kapilan, 2015), bacteria are not considered for SSF. Some 2.4. Screening with Plackett-Burman design bacteria species (Bacillus thuringiensis, Bacillus subtilis and Lactobacillus sp.), however, have been reported for SSF (Soccol et al., 2017). Lactic Plackett–Burman design was utilized to screen the significant factors acid bacteria have been extensively utilized in soybean fermentation to (having influence on SSF process for 5-CQA extraction with increased produce soybean flour, sufu (Chinese soy-food), and soybean-milk yield) – temperature (Temp), pH, incubation time (IT), inoculation size 2 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 (IS) and liquid-solid (L-S) ratio coded A, B, C, D and E respectively, in absorbance taken at 517 nm against a blank in a UV-1600 spectropho- ranges of 30–50 ◦C, 5–7, 0–48 h, 1–5%, and 0.25–0.67. Two factorial tometer (Beijing Rayleigh analytical instrument, China). The results (− 1 and +1) design locating significant variables for the extraction by were expressed in micromoles ascorbic acid equivalents per gram of dry screening “n” variables in “n+1” experiments was used. Thirteen runs of sample (μmol AA eq/g dry sample) using ascorbic acid standard curve different combinations of independent variables A – E given by the generated under same conditions. The linear range for ascorbic acid design were investigated at high (+), mid (0) and low (− ) levels. standard was 12.50–800.00 μg/ml (r2 = 1.00). Plackett–Burman design was achieved based on a first-order polynomial model (Boateng, Yang, & Li, 2020): 2.8. Total phenolic acids determination ∑5 Y = β0 + βiXi (1) Total phenolic acids were determined by adopting the method i=1 known as the Folin–Ciocalteu phenol reagent technique reported by Haida and Hakiman (2019), with minor modification. Briefly, 1 ml each where; Y = response, β0 = model intercept and βi = linear coefficients, Xi of HS and LHFHS extracts was added to 9 ml of distilled water in = independent variables. separate test tubes. Then, 1 ml of Folin–Ciocalteu phenol reagent was added to it and the mixture was mixed thoroughly via vortex. After 5 2.5. Optimization with Box-Behnken design min, 10 ml of 7% sodium carbonate was added. Next, 4 ml of distilled water was added and the mixture was adjusted to 25 ml of final volume. The four significant variables; temperature (X1), pH (X2), incubation The reaction mixture was incubated for 90 min at room temperature, time (X3) and liquid-solid ratio (X4) selected from Plackett–Burman and the absorbance was measured at 750 nm in a UV-1600 spectro- design experiment were further subjected to Box-Behnken design photometer (Beijing Rayleigh analytical instrument, China). The total (optimization) analysis to increase the yield of 5-CQA from HS variety. A phenolic acids were expressed as milligram of gallic acid equivalents per 4-factor-3-level Box-Behnken design comprised of 29 experimental runs gram of sample (mg GAE/g sample). A standard curve for gallic acid (as was used. A second-order polynomial model was fitted to correlate the standard) in methanol was prepared using different concentrations association of each parameter to the response. The equation as adopted (100–700 μg/ml). by Wang et al. (2020) was used: ∑3 ∑3 ∑3 ∑3 2.9. Determination of 5-CQA Y = β 20 + βiXi + βiiXi + × βijXiXj (2) i=1 i=1 i=1 j=i+1 5-CQA determination was performed according to advanced pro- where; Y predicted dependent variable, intercepts, linear cedures from previous studies (Adane et al., 2019) with slight modifi-= β0 = βi, = regression coefficients, , second-order regression coefficients and cation. 40 mg amount of RHSF and LHFHS samples was weighed and βii = βij interaction regression coefficients, all estimated by the model. X and dissolved in 30 ml distilled water in a 100 ml beaker separately. The = i Xj = independent factors. Overall Desirability Index (DI) was used to solution was stirred for 30 min using magnetic stirrer (model C-MAG HS ◦ select the optimized parameters from the equation (Akpabli-Tsigbe 7 S025, IKA, Germany) and heated (at 40 C) to increase the solubility of et al., 2021) below: 5-CQA in solution. The solution was filtered through double-loop qual-itative filter paper (NO. 1568, Ge Biotechnology Co., Ltd, China) to get [ ]1 3 3 rid of particles from solution. The filtrate containing 5-CQA was ∏ DI= di(yi) (3) collected and measured to obtain volume of the sample extract. The i=1 absorbance of the measured sample extract was taken using UV-1601 spectrophotometer (Beijing Rayleigh Analytical Instrument Co. Ltd, where; di = Desirability Index of response variable (0–1) and yi = China) within wavelength ranges of 190–1100 nm from which 5-CQA responses. concentration was computed against the standard solution by Beer Lambert’s Law at the maximum wavelength (λmax = 325 nm). Equations 2.6. Standard 5-CQA solution preparation (4) and (5) as adopted by Adane et al. (2019) were used to compute 5-CQA content and % 5-CQA of RHSF and LHFHS respectively: For the standard 5-CQA solutions preparation, a commercial 5-CQA (MO 63103, Sigma-Aldrich, Co., USA) was used. The method of Adane, [5 − CQAconc (mg/L)]× [total sample volume ml 2( ) ]5 − CQAcontent (mg)= Yoseph, and Kusse (2019) with slight modification was used for the [measured sample volume (ml) ]×1000 preparation of the standard solution. 1000 mg was dissolved in 1-L (4) distilled water to prepare stock standard 5-CQA solution. The solution was uniformly mixed using magnetic stirrer (C-MAG HS 7 S025, IKA, [calculated mass of 5 − CQA (mg) ]% 5 − CQA (w/w%) = × 100% (5) Germany) in the dark. Series of standard solutions (5, 10, 15, 20, 25 and [mass of sample measured (mg) ] 30) mgL− 1 were prepared from the stock solution for 5-CQA in distilled water. All measurements were done within 10 min after preparation and 2.10. Fourier transform infrared (FTIR) analysis absorbance of each series of standard 5-CQA solutions was taken immediately. The method was validated against Beer-Lambert’s law FTIR spectroscopy was applied to examine the structure of LHFHS with the series of standard 5-CQA solutions prepared. and RHSF samples according to the method described in the literature (Musa et al., 2019) with slight modification. Briefly, 1 mg from the 2.7. Antioxidant activity determination freeze-dried fermented and raw (unfermented) HS powder (control) was thoroughly mixed and ground with 200 mg of dried spectroscopic grade The antioxidant activity of the raw (unfermented) HS flour (RHSF) KBr (at 105 ◦C for 24 h) powder separately in a mortar with pestle (both and L. helviticus fermented HS (LHFHS) was determined by 2,2-diphenyl- made of agate). The resulting mixture was compacted with hydraulic 1-picrylhydrazyl radical scavenging activity using the method described machine (15 t) into a see-through (transparent) glass-like pellets of by Haida and Hakiman (2019) with slight modification. Aliquots of 1 ml thickness, 1–2 mm. The pellets were scanned in the wavenumber RHSF and LHFHS extracts were added to 2 ml of 1 mM methanolic ranging from 4000 to 400 cm− 1 with 128 scans using model Nicolet IS50 dilution of 2,2-diphenyl-1-picrylhydrazyl (1 × 10− 3 M). The mixture device (Thermo Nicolet Corporation, USA) at a resolution of 4 cm− 1. The after vortexing, was incubated in the dark for 30 min at 37 ◦C and blank (KBr pellet without test samples) used under setting parameters 3 / N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 was reported as reference spectra. appropriate proposed model should be close to 100% since it indicates a better explanation of the variability of the experimental data by the 2.11. Scanning electron microscopy (SEM) analysis proposed model. And also shows that a better correlation exists between observed and predicted values (Mintah et al., 2020). The effects of the The structure of LCFHS and RHSF samples was examined using SEM SSF parameters on the responses and the statistical significance were method described by Musa et al. (2019) with slight modification. The shown by Pareto chart (Fig. 1). This study obtained 2 limit lines namely LHFHS and RHSF samples were placed on a copper sample-holder with Bonferroni limit line (2.733) and t-value limit line (2.035). The t-values double-sided adhesive tapes and coated with a conductive layer of gold of effect above the Bonferroni limit line were considered extremely powder (about 10 nm) by using vacuum coating apparatus. Their significant, between Bonferroni limit line and t-value limit line were structures were examined with Hitachi S–3400N (Hitachi High Tech- considered significant and below t-value limit line were considered nologies, Tokyo, Japan) at 15 kV acceleration voltage. non-significant (Guo et al., 2018). These determinants were used to determine the extremely significant SSF parameters for 5-CQA extrac- 2.12. Atomic force microscopy (AFM) analysis tion from HS variety. The t-value effect of temperature and pH were above the Bonferroni limit line and positive for 5-CQA yield, fermen- The method outlined by Dabbour et al. (2020) (slightly modified) tation efficiency and antioxidant activity, suggestive that increasing of was used to determine the topography of LCFHS and RHSF samples. temperature and pH increased 5-CQA yield, fermentation efficiency and Dissolution of LHFHS and RHSF samples were done in 0.01 M saline antioxidant activity. Likewise, incubation time had a positive t-value of phosphate buffer (pH 8.0) to prepare 10 μg/ml final concentration. The effect (above the Bonferroni limit line) for 5-CQA yield and fermentation solution was heated in thermostatic water bath (50 ◦C) for 10 min and efficiency. However, liquid-solid ratio exhibited a negative t-value effect centrifuged (4000 rpm, 10 min). 5 μL aliquots of the supernatant were (above the Bonferroni limit line) for all responses (5-CQA yield, rapidly pipetted onto a newly cleaved mica substrates, placed in petri fermentation efficiency and antioxidant activity) indicative that dishes and dried in an incubator (25 ◦C) for 12 h. Multimode microscope increasing of liquid-solid ratio decreased 5-CQA yield, fermentation ef- (Bruker, Santa Barbara, CA) was used to generate the AFM images of the ficiency and antioxidant activity. samples. The lens was used in Peak ForceQNM mode with Bruker Sca- The decrease of antioxidant activity with increased liquid-solid ratio nAsyst needle at a typical spring and resonance frequency of 25.1 N/m was due to oxidation or degradation of antioxidant compounds (Kap- and 300 kHz respectively. rasob, Kerdchoechuen, Laohakunjit, Sarkar, & Shetty, 2017) dependent on the synergetic and redox interactions among the different compounds 2.13. Statistical analysis in HS, thus led to the low antioxidant activity of LHFHS sample. Simi- larly, the decrease of the 5-CQA yield at increased liquid-solid ratio was Version 11.0.5.0 Design Expert Software (STAT-EASE, Inc., USA) was due to degradation of 5-CQA (Heo, Adhikari, Choi, & Lee, 2020). Only used for the experimental designs and optimization. MINITAB v18.1 the t-value effect of pH was above the t-value limit line for total phenolic software (Minitab Inc., USA) was used to screen the variables. Accuracy acids. All the SSF parameters except inoculation size were extremely of the model was evaluated with P-test, determination coefficient (R2), significant (above Bonferroni limit-line) for 5-CQA yield and fermenta- lack of fit test and variation coefficient (CV), represented at p < 0.05, tion efficiency. Temperature, liquid-solid ratio and pH were extremely 0.01 and 0.001. All experiments were done three times and data pro- significant for antioxidant activity. The t-value of effect of inoculation cessed with MS Excel 2016 (Microsoft Corporation, USA). All graphs size was below the t-value limit line for all the responses. The results were constructed using OriginPro version 2018 (OriginLab Corporation, revealed that temperature, pH and incubation time showed significant USA). All values were reported as mean ± standard deviation. Tukeys’ positive effect on 5-CQA extraction from HS variety with the exception test was used for comparison of the means at p < 0.05. of liquid-solid ratio which exhibited a significant negative effect (Fig. 1a). This implied that increasing temperature, pH and incubation 3. Results and discussions time, increased 5-CQA extraction from HS variety while the reverse was obtained for liquid-solid ratio; increasing liquid-solid ratio, decreased In this study, a chronological optimization plan involving two 5-CQA extraction. The effect of inoculation size on 5-CQA extraction, stages/phases was used. The first phase involved screening of various however, was not significant though negative, suggestive that the effect SSF conditions and identifying those with the significant effects on of inoculation size (increasing or decreasing) on 5-CQA extraction was critical variables affecting 5-CQA extraction from HS variety using L. negligible. This showed that the four SSF parameters (temperature, pH, helviticus. The comparative importance of the various SSF conditions was incubation time and liquid-solid ratio) were the most significant factors studied with Plackett-Burman experimental design. Once these signifi- for 5-CQA extraction from HS variety, hence selected and used in the cant conditions were determined, the second phase ascertained their following experiments. combinations for best useful SSF conditions for 5-CQA extraction with Though inoculation size increased, it’s t-value of effect was not increased yield. Response surface methodology, a mathematics method significantly enough as shown by the interaction effect of Plack- based on the suitability of polynomial (quadratic) equation to an ett–Burman design matrix. Likewise, the influence of inoculation size on experimental data (Bezerra, Santelli, Oliveira, Villar, & Escaleira, 2008) the various responses was insignificant. Nonetheless, inoculation size was applied to achieve this aim. (the non-significant SSF parameter) was also studied in initial experi- ments. The results revealed that an average inoculation size of L. helvi- 3.1. Influence of SSF parameters on 5-CQA extraction ticus was adequate for maximum growth, proliferation and initially colonization of most HS substrate without overcrowding and competi- From Electronic Supplementary Table 1c, the F-values for 5-CQA tion for nutrient, resulting in high fermentation efficiency and 5-CQA yield, fermentation efficiency, total phenolic acids and antioxidant ac- extraction from HS variety. Inoculation size of 3% was therefore used tivity obtained were 18.11, 17.20, 2.80 and 20.47 respectively, indica- in further experiments which was slightly lower than that (4%) used by tive of significant model. However, the F-value for total phenolic acids Gao, Wang, Zhu, and Qian (2013) for A. oryzae. This could be due to the was relatively small and probably suggested that it was not an important differences in the microbial types. indicator for evaluating the effects of the SSF conditions for 5-CQA extraction from HS variety. The model for total phenolic acids was therefore not used for predictive purposes due to its low R2 value (29.77%). According to Handa et al. (2019), the R2 value of an 4 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Fig. 1. Pareto chart for 5-O-caffeoylquinic acid yield (a), fermentation efficiency (b), total phenolic acids (c) and antioxidant activity (d) of LHFHS sample. Pa- rameters having t-values greater than 2.035 (critical value) were regarded significant statistically. 3.2. Optimization of significant SSF conditions for 5-CQA extraction variables. using Box-Behnken design Optimization of SSF technique for 5-CQA extraction by selection of 3.3. Fitting models for extraction of 5-CQA from HS variety under SSF the best conditions is important to increase the 5-CQA yield. The sig- method nificant factors (temperature, pH, incubation time and liquid-solid ratio) chosen from Plackett-Burman design screening analysis were considered The experimental results obtained from Box-Behnken design analysis for further optimization using response surface methodology. Response (Table 1) showed that temperature, pH, incubation time and liquid-solid surface methodology is a group of mathematical and statistical tech- ratio had significant influence on the 5-CQA yield of the LHFHS sample. niques that optimizes conditions for an assured goal via establishment of The effect of the studied variables, parameter interactions, and arith- a model from analysis of problems involving one or more responses of metic significance of the model was evaluated with analysis of variance. interest affected by several variables (Tang, Zhang, & Fang, 2015). The Table 2 displayed the F- and p-values of regression coefficients for the levels of the factors used for the optimization study were set based on the response variables. The quadratic polynomial model deduced from the previous screening analysis. The experimental conditions and extraction highly significant p-values (<0.0001) of the models gave good estimates yield from HS variety using four-factor-three-level Box-Behnken design for the responses measured, indicating the fitness of the model. The R 2 were presented in Table 1. Response surface methodology explains the values for 5-CQA yield, fermentation efficiency and antioxidant activity nature of data set and makes mathematical prediction. It saves time, (0.9989, 0.9927 and 0.9996 respectively) affirmed it. In addition, the sample use, and space, hence, more favourable relative to single-factor lack of fit values (p-value = 0.8588, 0.5151 and 0.2372) for the re- optimization (Lee et al., 2012). Multiple regression analysis was per- sponses (5-CQA yield, fermentation efficiency and antioxidant activity formed on the experimental data (Table 2) to evaluate for significance. respectively) were statistically not significant, confirmatory of model The mathematical model for the SSF conditions optimization for adequacy. extraction of 5-CQA from HS variety was achieved with second-order polynomial equation through investigation of the relationships be- 3.3.1. Influence of SSF parameters on 5-CQA yield of HS variety tween the independent (process) and the dependent (responses) Table 1 showed the 5-CQA yield obtained from LHFHS under SSF, which varied from 2.00 ± 0.06–11.41 ± 0.07 mg/g. The effect of SSF 5 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Table 1 Box-Behnken design matrix with experimental design and data for the extraction of 5-CQA by SSF technique for one experimental block with Lactobacillus helviticus. Run Fermentation parameters (actual and coded values) Responsec Temperature pH Incubation time Liquid-solid 5-CQA yield (mg/ Fermentation efficiency Antioxidant activity (μmol AA eq/g dry (oC) (h) ratio g) (%) sample) X1 X2 X3 X4 1 40.00 (0) 5.00 24.00 (0) 0.67 (+1) 9.40 ± 0.04 29.50 ± 0.18 35.77 ± 2.20 (− 1) 2 40.00 (0) 5.00 24.00 (0) 0.25 (− 1) 4.14 ± 0.02 17.32 ± 0.08 49.42 ± 1.51 (− 1) 3 50.00 (+1) 6.00 (0) 0.00 (− 1) 0.46 (0) 4.12 ± 0.05 16.35 ± 0.21 46.55 ± 1.23 4 40.00 (0) 5.00 48.00 (+1) 0.46 (0) 3.21 ± 0.09 16.80 ± 0.00 48.65 ± 1.42 (− 1) 5 50.00 (+1) 5.00 24.00 (0) 0.46 (0) 5.33 ± 0.02 20.68 ± 0.07 38.51 ± 2.34 (− 1) 6 50.00 (+1) 6.00 (0) 24.00 (0) 0.67 (+1) 10.71 ± 0.03 30.81 ± 0.11 44.83 ± 1.37 7 30.00 (− 1) 6.00 (0) 0.00 (− 1) 0.46 (0) 2.00 ± 0.06 16.46 ± 0.25 22.19 ± 2.78 8 40.00 (0) 6.00 (0) 0.00 (− 1) 0.25 (− 1) 5.37 ± 0.02 21.56 ± 0.08 37.44 ± 1.81 9 30.00 (− 1) 6.00 (0) 48.00 (+1) 0.46 (0) 5.46 ± 0.02 22.09 ± 0.07 48.95 ± 2.48 10 40.00 (0) 6.00 (0) 24.00 (0) 0.46 (0) 4.45 ± 0.01 18.17 ± 0.04 36.19 ± 2.89 11 40.00 (0) 6.00 (0) 24.00 (0) 0.46 (0) 4.17 ± 0.03 18.04 ± 0.11 36.48 ± 0.91 12 30.00 (− 1) 5.00 24.00 (0) 0.46 (0) 2.90 ± 0.05 16.02 ± 0.21 47.00 ± 1.42 (− 1) 13 30.00 (− 1) 6.00 (0) 24.00 (0) 0.25 (− 1) 6.13 ± 0.09 24.19 ± 0.36 48.21 ± 1.48 14 50.00 (+1) 6.00 (0) 48.00 (+1) 0.46 (0) 3.10 ± 0.01 16.96 ± 0.04 24.68 ± 2.18 15 40.00 (0) 6.00 (0) 0.00 (− 1) 0.67 (+1) 6.38 ± 0.02 22.39 ± 0.08 29.63 ± 2.03 16 30.00 (− 1) 7.00 24.00 (0) 0.46 (0) 8.02 ± 0.03 27.00 ± 0.11 25.00 ± 2.14 (+1) 17 40.00 (0) 5.00 0.00 (− 1) 0.46 (0) 2.35 ± 0.07 16.83 ± 0.28 36.94 ± 3.22 (− 1) 18 40.00 (0) 6.00 (0) 24.00 (0) 0.46 (0) 4.15 ± 0.07 19.00 ± 0.28 36.37 ± 1.30 19 40.00 (0) 6.00 (0) 48.00 (+1) 0.67 (+1) 10.96 ± 0.03 32.49 ± 0.11 37.29 ± 2.21 20 50.00 (+1) 7.00 24.00 (0) 0.46 (0) 5.48 ± 0.04 19.33 ± 0.19 32.59 ± 2.49 (+1) 21 40.00 (0) 6.00 (0) 48.00 (+1) 0.25 (− 1) 3.00 ± 0.02 15.05 ± 0.07 35.90 ± 2.30 22 40.00 (0) 6.00 (0) 24.00 (0) 0.46 (0) 4.39 ± 0.05 18.05 ± 0.21 36.10 ± 2.94 23 40.00 (0) 7.00 24.00 (0) 0.67 (+1) 11.41 ± 0.07 31.10 ± 0.31 32.33 ± 2.42 (+1) 24 40.00 (0) 6.00 (0) 24.00 (0) 0.46 (0) 4.49 ± 0.08 17.24 ± 0.35 36.11 ± 4.54 25 50.00 (+1) 6.00 (0) 24.00 (0) 0.25 (− 1) 4.47 ± 0.04 17.80 ± 0.18 24.60 ± 3.24 26 40.00 (0) 7.00 48.00 (+1) 0.46 (0) 6.39 ± 0.03 25.13 ± 0.12 25.65 ± 3.69 (+1) 27 30.00 (− 1) 6.00 (0) 24.00 (0) 0.67 (+1) 8.91 ± 0.05 28.80 ± 0.23 22.29 ± 4.27 28 40.00 (0) 7.00 24.00 (0) 0.25 (− 1) 7.28 ± 0.03 26.10 ± 0.15 24.63 ± 4.56 (+1) 29 40.00 (0) 7.00 0.00 (− 1) 0.46 (0) 4.56 ± 0.01 17.09 ± 0.04 31.91 ± 4.66 (+1) X1 = Temperature; X2 = pH; X3 = Incubation time; X C4 = Liquid-solid ratio, : Data were average values (x3). conditions on 5-CQA yield from the LHFHS was shown in Fig. 2. On the sample. Handa et al. (2019) reported similar effect of liquid-solid ratio in basis of the p-values from the results, liquid-solid ratio was the notice- their study studies which was also on production of bioactive com- able most important parameter positively and significantly (p < 0.0001) pounds under SSF. influencing 5-CQA yield (Table 2). 5-CQA yield of the LHFHS sample The interactions between all the parameters influenced (negatively increased with increasing liquid-solid ratio (to 0.67 maximum). The next or positively) 5-CQA yield significantly. Thus, as incubation time and obvious parameter was pH, which had significant positive effect on 5- temperature increased, the 5-CQA of the LHFHS decreased (Fig. 2b). CQA yield of LHFHS sample. Temperature, however, was positive but However, as liquid-solid ratio and incubation time, pH or temperature its effect was insignificant. Fig. 2c, obviously showed that the liquid- increased, the 5-CQA of the LHFHS also increased (Fig. 2; c, e and f). solid ratio–temperature interaction tremendously increased the 5-CQA Similarly, as both pH and incubation time increased, the 5-CQA of the yield of LHFHS sample positively. This implied that collectively LHFHS increased (Fig. 2d). Also, as temperature decreased and pH increasing liquid-solid ratio and temperature increased the 5-CQA yield increased, the 5-CQA of the LHFHS increased (Fig. 2a). L. helviticus of the LHFHS sample. Though pH had lower positive F-value than that of possesses cinnamoyl esterase enzymes (temperature specific) which liquid-solid ratio, its effect was significant on the 5-CQA yield. hydrolyze ester bonds resulting in the release of 5-CQA (Aguirre Santos, The results showed that 5-CQA yield increased with increasing pH Schieber, & Weber, 2018). This suggests that the observed decrease in (Fig. 2a, d and e). Organic acids production and scale up are influenced yield of the 5-CQA as a function of increases in incubation time and by pH (Yazid, Barrena, Komilis, & Sánchez, 2017). pH influences mi- temperature, was due to inhibition of the activity of cinnamoyl esterase crobial growth, proliferation and substrate colonization. It also enhances enzymes, whereas the increase in 5-CQA yield realized (as temperature the efficiency of SSF, hence the observed increased yield of 5-CQA of decreased and pH increased) was as a result of the creation of optimum LHFHS obtained with increasing pH. Relative to liquid-solid ratio and conditions for cinnamoyl esterase enzymes to efficiently hydrolyze the pH, incubation time had the lowest positive but significant effect on the ester bonds of HS variety to release more 5-CQA. Increased liquid-solid 5-CQA yield. 5-CQA yield of the LHFHS sample increased with ratio, incubation time and pH created optimum conditions for L. helvi- increasing incubation time. The interaction between incubation time ticus proliferation, resulting in degradation of the cell walls of the HS and liquid-solid ratio positively increased the 5-CQA yield of LHFHS variety to release considerable quantity of 5-CQA into solution which 6 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Table 2 Analysis of variance, regression analysis and optimal conditions for 5-CQA extraction from HS variety by SSF using Lactobacillus helviticus. Source 5-CQA yield (mg/g) Fermentation efficiency (%) Antioxidant activity (μmol AA eq/g dry sample) F-value p-value F-value p-value F-value p-value Model 870.13 <0.0001*** 136.58 <0.0001*** 2825.68 <0.0001*** Linear X1 = A: Temperature 0.24 0.6336 NS 32.31 <0.0001*** 5.68 0.0318* X2 = B: pH 1346.07 <0.0001*** 165.70 <0.0001*** 11397.43 <0.0001*** X3 = C: Incubation time 290.13 <0.0001*** 64.47 <0.0001*** 435.76 <0.0001*** X4 = D: Liquid-Solid ratio 4037.12 <0.0001*** 570.53 <0.0001*** 524.60 <0.0001*** Interactions AB 399.06 <0.0001*** 92.39 <0.0001*** 1247.62 <0.0001*** AC 324.25 <0.0001*** 15.31 0.0016** 11410.88 <0.0001*** AD 193.41 <0.0001 42.88 <0.0001*** 10276.70 <0.0001*** BC 15.20 0.0016** 39.58 <0.0001*** 1558.14 <0.0001*** BD 20.63 0.0005*** 31.33 <0.0001*** 2199.41 <0.0001*** CD 780.36 <0.0001*** 167.67 <0.0001*** 408.40 <0.0001*** Quadratic A2 34.15 <0.0001*** 16.32 0.0012** 20.70 0.0005*** B2 245.58 <0.0001*** 52.16 <0.0001*** 0.13 0.7284 NS C2 363.19 <0.0001*** 20.63 0.0005*** 16.15 0.0013** D2 3714.12 <0.0001*** 585.76 <0.0001*** 77.97 <0.0001*** Fitting statistics Lack of fit 0.45 0.8588 NS 1.08 0.5151 NS 2.17 0.2372 NS R2 0.9989 0.9927 0.9996 Adjusted R2 0.9977 0.9855 0.9993 Predicted R2 0.9956 0.9664 0.9982 Adeq. Precision 105.274 37.401 166.000 C.V. % 2.22 3.01 0.64 Standard Dev. 0.12 0.64 0.23 Optimization equations 5 − CQA yield (mg /g) = 4.33 − 0.018X1 + 1.32X2 + 0.61X3 + 2.28X4 − 1.24X1X2 − 1.12X1X3 + 0.87X1X4 + 0.24X2X3 − 0.28X2X4 + 1.74X3X4 + 0.29X 21 + 0.77X 22 − 0.93X 23 + 2.98X 2 4 Fermentation efficiency (%) = 18.1 − 1.05X1 + 2.38X2 + 1.49X3 + 4.42X4 − 3.08X1X2 − 1.25X1X3 + 2.1X1X4 + 2.02X 2 2 22X3 − 1.79X2X4 + 4.15X3X4 + 1.02X1 + 1.82X2 − 1.14X3 + 6.09X 2 4 Antioxidant activity (μmol AA eq /g dry sample) = 36.25 − 0.16X1 − 7.01X2 + 1.37X3 − 1.5X4 + 4.02X1X2 − 12.16X1X3 + 11.54X1X4 − 4.49X2X3 + 5.34X2X4 + 2.3X3X4 − 0.41X 21 − 0.032X 22 − 0.36X 2 2 3 − 0.79X4 *, ** and *** denote significance at p < 0.05, p < 0.01 and p < 0.001 respectively while NS denotes not significant. Fig. 2. Contour and response surface plots showing interactive influence of temperature, pH, incubation time and liquid-solid ratio on the 5-O-caffeoylquinic acid yield of LHFHS sample. 7 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 contributed to the observed increased 5-CQA yield of the LHFHS sample. efficiency, even though there was a slight increase after the mid-point All the quadratic terms of the model significantly influenced 5-CQA (about 45.00 ◦C) (Fig. 3c). Thus, fermentation efficiency decreased yield of the LHFHS sample. A second-degree quadratic equation gener- with increased temperature (Fig. 3a and b). This was clearly shown by ated from regression analysis was used to examine the association be- the perturbation plot. Temperature affects microbial growth. High tween the SSF parameters and response variables. The statistically temperatures kill microorganisms (inhibiting their activity) while opti- insignificant (p > 0.05) term (temperature), was removed from the mum temperatures enhance microbial proliferation and thus their effi- model to obtain a better fit model. The regression equation describing ciency, leading to production of desired products. The inverse the effectiveness of SSF in coded variables for achieving the maximum relationship between temperature and fermentation efficiency of L. 5-CQA yield of LHFHS sample was: helviticus observed in the present study could be that most of L. helviticus 5 − CQA (mg / g)= 4.33+ 1.32X2 + 0.61X3 + 2.28X4 − 1.24X1X2 − 1.12X1X3 + 0.87X1X4 + 0.24X2X3 − 0.28X2X4 + 1.74X3X4 + 0.29X21 + 0.77X 2 2 − 0.93X 2 3 + 2.98X 2 4 (6) were killed as temperature increased. Handa et al. (2019) observed similar decreasing effect of temperature on bioactive compounds pro- duction using SSF. 3.3.2. Influence of SSF parameters on fermentation efficiency of L. Both the interactions between all the SSF parameters and their helviticus on HS variety quadratic terms significantly influenced the fermentation efficiency of L. From the results (Table 1), the obtained fermentation efficiency of helviticus in the LHFHS sample (Table 2). Specifically, as liquid-solid the LHFHS sample using SSF was within the range of 15.05 ± 0.07 to ratio and temperature, pH or incubation time increased, the fermenta- 32.49 ± 0.11%. The maximum fermentation efficiency was attained at tion efficiency of L. helviticus also increased (Fig. 2; c, e and f). Likewise, temperature of 40.00 ◦C, pH of 6.00, incubation time of 48.00 h, and as temperature decreased and pH or incubation time increased, the liquid-solid ratio of 0.67. All the SSF parameters (temperature, pH, in- fermentation efficiency increased (Fig. 2; a and b). Also, as both pH and cubation time and liquid-solid ratio) investigated significantly influ- incubation time increased, the fermentation efficiency increased enced fermentation efficiency. However, only pH, incubation time and (Fig. 2d). Suggestive that optimum conditions for L. helviticus prolifer- liquid-solid ratio had positive effect on the fermentation efficiency. ation were attained when liquid-solid ratio and temperature, pH or in- Temperature exhibited a general decreasing effect on the fermentation cubation time increased (also as both pH and incubation time Fig. 3. Contour and response surface plots showing interactive influence of temperature, pH, incubation time and liquid-solid ratio on the fermentation efficiency of L. helviticus in LHFHS sample. 8 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Fig. 4. Contour and response surface plots showing interactive influence of temperature, pH, incubation time and liquid-solid ratio on the antioxidant activity of LHFHS sample. increased), which in turn increased the activity of L. helviticus resulting and liquid-solid ratio on antioxidant activity of the LHFHS sample were in the observed high fermentation efficiency. This scenario was same as, found to be extremely significant while that of temperature was signif- temperature decreased and pH or incubation time increased and resul- icant (Table 2). The results revealed pH to have the most significant ted in similar result mentioned above. The predicting quadratic equation effect on antioxidant activity but with a decreasing activity (Fig. 4a, showing the effectiveness of the optimized SSF parameters in obtaining d and e). The antioxidant activity of LHFHS sample steeply decreased as high fermentation efficiency of L. helviticus in the LHFHS sample, written pH increased; confirmed by the perturbation plot. Adebo, Njobeh, in coded variables was: Adebiyi, and Kayitesi (2018) reported similar results and attributed it to rearrangement of the phenolic structures caused by the acidic environ- ment of the fermentation process as the pH increased, leading to Fermentation efficiency (%)= 18.1 − 1.05X1 + 2.38X2 + 1.49X3 + 4.42X4 − 3.08X1X2 − 1.25X1X3 + 2.1X1X4 + 2.02X2X3 − 1.79X2X4 + 4.15X3X4 + 1.02X21 + 1.82X22 − 1.14X 2 3 + 6.09X 2 4 (7) 3.3.3. Influence of SSF parameters on antioxidant activity of HS variety phenolic compounds undergoing self-polymerization and/or interacting Auto-oxidation of food components is prevented by antioxidants. with other macromolecules like amino acids and starch, hence reducing Antioxidants also neutralize the excess free radicals produced in the their extractability. According to Hur et al. (2014), during fermentation, human body (Hur, Lee, Kim, Choi, & Kim, 2014). Numerous fermented antioxidant activity is influenced by pH changes through changing of the products have high antioxidant activity, hence very useful in this regard. phenolic compounds contents and structure. Similar results (decreased Considering that, fermentation of food materials is a valuable technol- antioxidant activity at increased pH) were reported by Ruenroengklin ogy for improvement of the antioxidant activity of food products. In et al. (2008) also. accordance with this, the antioxidative activity of LHFHS sample was Furthermore, all the interactive terms showed positive, extremely evaluated with 2,2-diphenyl-1-picrylhydrazyl radical scavenging activ- significant (p < 0.0001) effect on antioxidant activity. The results also ity method. The effect of four SSF parameters on the antioxidant activity found three quadratic terms out of four to exhibit significant effect on of LHFHS sample was investigated. The effects of pH, incubation time antioxidant activity. The quadratic effect of pH on antioxidant activity 9 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 was not significant, though exhibited a positive influence. The antioxi- dant activity obtained for the LHFHS from the study (under the exper- imental SSF conditions) ranged from 22.19 ± 2.78 to 49.42 ± 1.51 μmol AA eq/g dry sample. The lowest antioxidant activity (22.19 ± 2.78 μmol AA eq/g dry sample) was obtained under temperature = 30.00 ◦C, pH = 6.00, incubation time = 0.00 h and liquid-solid ratio = 0.46 experi- mental SSF conditions, whilst the highest antioxidant activity (49.42 ± 1.51 μmol AA eq/g dry sample) was attained at temperature = 40.00 ◦C, pH = 5.00, incubation time = 24 h and liquid-solid ratio = 0.25 SSF conditions. This implied that at higher pH and liquid-solid ratio, and lower temperature and incubation time, SSF conditions may not be desirable for the production of LHFHS with highest antioxidant activity. The regression equation for explaining the effectiveness of the optimized SSF conditions for producing LHFHS with maximum antioxidant activity after the removal of insignificant term was written in coded variables as follows: 3.3.4. Verification and validation of predictive model Fig. 5a. FTIR spectra of RHSF and LHFHS samples. All the dependent variables/responses (5-CQA yield, fermentation efficiency and antioxidant activity) investigated were experimented Fig. 5b. 5-O-caffeoylquinic acid yield (a), antioxidant activity (b), SEM and AFM micrographs of RHSF and LHFHS samples. 10 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Antioxidant activity (μmol AA eq / g dry sample) = 36.25 − 0.16X1 − 7.01X2 + 1.37X3 − 1.5X4 + 4.02X1X2 − 12.16X1X3 + 11.54X1X4 − 4.49X2X3 + 5.34X2X4 +2.3X X − 0.41X23 4 1 − 0.36X 2 3 − 0.79X 2 4 (8) under the predictive, optimized SSF conditions for 5-CQA extraction absorption peak at approximately 918 cm− 1 in RHSF sample, shifted to with increased yield and improved antioxidant activity to verify the approximately 904 cm− 1 in LHFHS sample (between 800 and 1000 model reliability. Temperature = 49.90 ◦C, pH = 7.00, incubation time cm− 1) was mainly caused by the β-glycosidic linkage between cellulose = 25.81 h, and liquid-solid ratio = 0.67 were the optimized SSF con- and hemicellulose sugar units (Loow & Wu, 2018). The small sharp ditions obtained for all the three responses. Under these optimized SSF absorption peak of RHSF sample at approximately 1063 cm− 1 indicating conditions, both the predicted responses (5-CQA yield = 11.38 mg/g, C–O stretching vibration of cellulose, hemicellulose and lignin (Loow fermentation efficiency = 30.48% and antioxidant activity = 46.12 μmol et al., 2017) was intensified (shifted to approximately 1076 cm− 1) in AA eq/g dry sample) and experimental responses (5-CQA yield = 11.41 LHFHS sample, suggesting that the lignin, hemicellulose and cellulose ± 0.27 mg/g, fermentation efficiency = 30.49 ± 1.14% and antioxidant structures were exposed after the degradation of the cell walls of HS activity = 46.13 ± 1.94 μmol AA eq/g dry sample) were compared. The variety in the SSF. This possibly led to the release of more 5-CQA which results of the experimental responses compared very well with that of gave the high content obtained. For the 1200–1400 cm− 1 range, the the predicted responses. This proposed that the Box-Behnken design absorption bands at approximately 1237 cm− 1, result of aryl–alkyl model obtained for the SSF of HS variety to extract 5-CQA with C–O–C ether bond of lignin (Fakayode et al., 2020) and 1344 cm− 1, increased yield and improved antioxidant activity was efficient. result of C–H bending vibrations (Li, Wei, Xu, Xu, & He, 2018) in RHSF Furthermore, the desirability of the model was 0.92. Jarpa-Parra et al. were shifted to 1239 and 1347 cm− 1 in LHFHS, respectively. Regarding (2014) reported 0.6–0.8 (composite desirability) as a satisfactory value; the 1400–1600 cm− 1 range, the bands (in RHSF sample) at approxi- thus, desirability index of 0.92 obtained is highly satisfactory. mately 1400 cm− 1 due to symmetrical CH2– groups bending of cellulose (Loow & Wu, 2018), 1442 cm− 1 due to C–H bending vibrations and − 1 3.4. Comparison of 5-CQA yield and antioxidant activity of fermented 1545 cm due to in-plane C–C aromatic vibration (Li et al., 2018) were and unfermented HS variety shifted to approximately 1410, 1454 and 1546 cm − 1 respectively in LHFHS sample. The alteration of the band from 1545 (in RHSF sample) − 1 After the model validation, the experimental values of 5-CQA yield to 1546 cm (in LHFHS sample) justified that the SSF released signif- (11.41 ± 0.27 mg/g) and antioxidant activity (46.13 ± 1.94 μmol AA icant quantities of phenolics (Loow et al., 2017). The absorption peak at approximately 1656 cm− 1 in the interval of 1600–1700 cm− 1 eq/g dry sample) obtained under the optimized SSF conditions were providing compared to that (5-CQA yield = 1.63 ± 0.53 mg/g and antioxidant information on the O–H bending of absorbed water (Li et al., 2018) in activity = 10.98 ± 0.21 μmol AA eq/g dry sample) of the unfermented RHSF sample was shifted to approximately 1655 cm − 1 in LHFHS sample. HS flour (RHSF). From the results, the optimized SSF conditions gave This region was the most significant carbonyl absorption region as re- higher 5-CQA yield and antioxidant activity than the extraction from the ported by Li et al. (2018). It was evident from the results that the SSF unfermented sample (Fig. 5b; a and b). The high 5-CQA yield and significantly impacted on the structure of the HS variety. antioxidant activity of the LHFHS sample obtained from the SSF model was due to the activities of L. helviticus through enzymes production 3.6. SEM of the RHSF and LHFHS samples which degraded the cell walls of the HS variety and broke the bonds between 5-CQA and other biomolecules (proteins, oligosaccharides, Comparison of the SEM images of RHSF and LHFS samples was done etc.). This resulted in the release of more free 5-CQA which gave the high to examine morphological changes after SSF. The micrographs revealed 5-CQA yield with the improved antioxidant activity than the extraction that the granules of RHSF sample were spherical, small, fused (some), from unfermented sample as observed. The results were in agreement scattered and had consistent structure and smooth surface (Fig. 5b; c). with other studies which stated that hydroxycinnamic acids mostly exist However, in LHFHS sample, there was more noticeable granular struc- in linked-form with cell walls (Santos da Silveira et al., 2019) and ture degradation, changing it to a more loosened, irregular, rough sur- breaking of the bond between 5-CQA and oligosaccharides and/or faced with pits, and agglomerated granules (Fig. 5b; d). This suggested polysaccharides through enzymatic degradation increases the content of that there was degradation of starch and amino acids components, free 5-CQA (Su, Cheng, Hsiao, Han, & Yu, 2018). Additionally, Taylor caused by the SSF, affirming the release of more, free 5-CQA, hence the and Duodu (2015) stated that lactic acid bacteria metabolic activity high content with improved antioxidant activity obtained. The results during fermentation process involves various enzyme activities that conformed to the results of Adebo, Njobeh, Mulaba-Bafubiandi et al. influence the food chemical constituents, especially phenolic com- (2018). pounds, hence determining their fate in fermented products. 3.7. AFM of RHSF and LHFHS samples 3.5. FTIR of the RHSF and LHFHS samples AFM does not only image surfaces of biological structures in their The alterations in the chemical structure of HS variety after SSF were native environment and oligomeric states, but visualizes conformational observed with FTIR. The structural changes of the samples (RHSF and changes in the structures too. Based on this, it was applied in this study LHFHS) were examined by the positional changes of the peaks of lignin, to examine the alterations in the topographic images of the samples hemicellulose and cellulose. Consistent with the spectral trends, the (RHSF and LHFHS) after SSF. Fig. 5b; e and f depicted the AFM images of sharp peaks observed at approximately 1056, 1157 and 1745 cm− 1 RHSF and LHFHS samples. Structural variations were observed between (between 900 and 1800 cm− 1), corresponding to C–O stretching vibra- the two samples. The topographic image of RHSF particles was smaller tion of cellulose, hemicellulose and lignin (Loow et al., 2017), C–O in size and scattered with very few big ones. While that of LHFHS par- ester groups stretching vibration of hemicellulose and C–O–C asym- ticles was loosed and big sized with irregular particles. The number of metrical stretching vibration of hemicellulose and cellulose (Fakayode LHFHS particles was smaller than that of RHSF. Similarly, the particle et al., 2020) respectively were same for both samples (Fig. 5a). The heights of LHFHS sample were shorter than that of RHSF with minor 11 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 scattering and micropores. The results conformed to that reported by components and microstructure of ting (a Southern African food) from whole grain Dabbour et al. (2020). The results showed that the SSF significantly sorghum. Food Bioscience, 25, 118–127. https://doi.org/10.1016/j.fbio.2018.08.007 Adebo, O. A., Njobeh, P. B., Mulaba-Bafubiandi, A. F., Adebiyi, J. A., Desobgo, Z. S. C., & changed the structure of the HS variety, resulting in the extraction of the Kayitesi, E. (2018). Optimization of fermentation conditions for ting production 5-CQA with increased yield and improved antioxidant activity. using response surface methodology. Journal of Food Processing and Preservation, 42 (1), 1–10. https://doi.org/10.1111/jfpp.13381 Adnan, M., Ashraf, S. A., Khan, S., Alshammari, E., & Awadelkareem, A. M. (2017). Effect 4. Conclusion of pH, temperature and incubation time on cordycepin production from Cordyceps militaris using solid-state fermentation on various substrates. CyTA - Journal of Food, In this study, a novel model for the extraction of 5-CQA with 15(4), 617–621. https://doi.org/10.1080/19476337.2017.1325406 Aguirre Santos, E. A., Schieber, A., & Weber, F. (2018). Site-specific hydrolysis of increased yield and enhanced antioxidant activity from HS variety was chlorogenic acids by selected Lactobacillus species. Food Research International, 109, obtained. The model combined screening with quantitative multivariate 426–432. https://doi.org/10.1016/j.foodres.2018.04.052, 2017. calibration analysis based on SSF technique. Screening of SSF conditions Akpabli-Tsigbe, N. D. K., Ma, Y., Ekumah, J.-N., Osabutey, J., Hu, J., Xu, M., et al. (2021). Two-step optimization of solid-state fermentation conditions of heilong48 by Plackett-Burman design showed that temperature, pH, incubation soybean variety for maximum chlorogenic acid extraction yield with improved time and liquid-solid ratio contributed significantly to 5-CQA yield, antioxidant activity. Industrial Crops and Products, 168, 113565. https://doi.org/ fermentation efficiency and antioxidant activity. Box-Behnken design 10.1016/j.indcrop.2021.113565 Ang, T. N., Ngoh, G. C., & Chua, A. S. M. (2013). Development of a novel inoculum was successfully used to set the optimized SSF conditions; temperature preparation method for solid-state fermentation-Cellophane film culture (CFC) = 49.90 ◦C, pH = 7.00, incubation time = 25.81 h and liquid-solid ratio technique. Industrial Crops and Products, 43(1), 774–777. https://doi.org/10.1016/j. = 0.67 under which maximum 5-CQA yield of 11.38 mg/g, fermentation indcrop.2012.08.022 efficiency of 30.49% and antioxidant activity of 46.12 μmol AA eq/g dry Bartkiene, E., Krungleviciute, V., Juodeikiene, G., Vidmantiene, D., & Maknickiene, Z. (2015). Solid state fermentation with lactic acid bacteria to improve the nutritional sample were obtained. The experimental and predicted data compared quality of lupin and soya bean. Journal of the Science of Food and Agriculture, 95, very well (desirability index, 0.92), making the model acceptable. 1336–1342. https://doi.org/10.1002/jsfa.6827 Compared to extraction from unfermented HS variety, the extraction Bezerra, M. A., Santelli, R. E., Oliveira, E. P., Villar, L. S., & Escaleira, L. A. (2008). Response surface methodology (RSM) as a tool for optimization in analytical under the optimized SSF conditions gave higher 5-CQA yield with chemistry. Talanta, 76(5), 965–977. https://doi.org/10.1016/j.talanta.2008.05.019 enhanced antioxidant activity. The study showed that extraction of 5- Bhargav, S., Panda, B. P., Ali, M., & Javed, S. (2005). Solid-state fermentation systems - CQA with increased yield and enhanced antioxidant activity from HS an overview. Chemical and Biochemical Engineering Quarterly, 22(1), 49–70. https:// doi.org/10.1080/07388550590925383 variety under optimized SSF conditions was efficient, affirmed by Boateng, I. D., Yang, X., & Li, Y. (2020). Optimization of infrared-drying parameters for Fourier transform infrared, Scanning electron microscopy and Atomic Ginkgo biloba L . seed and evaluation of product quality and bioactivity. Industrial force microscopy micrographs. This study could help the food/phar- Crops and Products, 1–10. https://doi.org/10.1016/j.indcrop.2020.113108 Correa Deza, M. A., Rodríguez de Olmos, A., & Garro, M. S. (2019). Solid state maceutical industry to produce cost-effective high 5-CQA yield with fermentation to obtain vegetable products bio-enriched with isoflavone aglycones enhanced antioxidant activity from HS variety. Further studies on the using lactic cultures. Revista Argentina de Microbiología, 51(3), 201–207. https://doi. bioactivities of 5-CQA from HS variety are therefore recommended. org/10.1016/j.ram.2018.04.006 Dabbour, M., Alenyorege, E. A., Mintah, B., He, R., Jiang, H., & Ma, H. (2020). Proteolysis kinetics and structural characterization of ultrasonic pretreated CRediT authorship contribution statement sunflower protein. Process Biochemistry, 94, 198–206. https://doi.org/10.1016/j. procbio.2020.04.018 Fakayode, O. A., Aboagarib, E. A. A., Yan, D., Li, M., Wahia, H., Mustapha, A. T., et al. Nelson Dzidzorgbe Kwaku Akpabli-Tsigbe: Conceptualization, (2020). Novel two-pot approach ultrasonication and deep eutectic solvent Methodology, Software, Validation, Formal analysis, Investigation, Data pretreatments for watermelon rind delignification: Parametric screening and curation, Writing – original draft, Writing – review & editing, Visuali- optimization via response surface methodology. Energy, 203, 117872. https://doi. org/10.1016/j.energy.2020.117872 zation. Yongkun Ma: Supervision, Resources, Project administration, Gao, Y., Wang, C., Zhu, Q., & Qian, G. (2013). Optimization of solid-state fermentation Funding acquisition. John-Nelson Ekumah: Investigation. Juliet with Lactobacillus brevis and Aspergillus oryzae for trypsin inhibitor degradation in Osabutey: Writing – review & editing. Jie Hu: Investigation. Manqing soybean meal. Journal of Integrative Agriculture, 12(5), 869–876. https://doi.org/ 10.1016/S2095-3119(13)60305-6 Xu: Investigation. Nana Adwoa Nkuma Johnson: Investigation. Guo, N., Song, X. R., Kou, P., Zang, Y. P., Jiao, J., Efferth, T., et al. (2018). Optimization of fermentation parameters with magnetically immobilized Bacillus natto on Ginkgo seeds and evaluation of bioactivity and safety. Lebensmittel-Wissenschaft und Declaration of competing interest -Technologie- Food Science and Technology, 97, 172–179. https://doi.org/10.1016/j. lwt.2018.06.046 Haida, Z., & Hakiman, M. (2019). A comprehensive review on the determination of The authors declare that they have no known competing financial enzymatic assay and nonenzymatic antioxidant activities. Food Sciences and interests or personal relationships that could have appeared to influence Nutrition, 7(5), 1555–1563. https://doi.org/10.1002/fsn3.1012 the work reported in this paper. Handa, C. L., de Lima, F. S., Guelfi, M. F. G., Fernandes, M. da S., Georgetti, S. R., & Ida, E. I. (2019). Parameters of the fermentation of soybean flour by Monascus purpureus or Aspergillus oryzae on the production of bioactive compounds and Acknowledgment antioxidant activity. Food Chemistry, 271, 274–283. https://doi.org/10.1016/j. foodchem.2018.07.188 Heo, J., Adhikari, K., Choi, K. S., & Lee, J. (2020). Analysis of caffeine, chlorogenic acid, The authors are grateful for the support provided by the Zhenjiang trigonelline, and volatile compounds in cold brew coffee using high-performance Key Research and Development Program (Modern Agriculture) [grant liquid chromatography and solid-phase microextraction—gas chromatography-mass number NY2020020]. spectrometry. Foods, 9, 1–2. https://doi.org/10.3390/foods9121746, 1746. Hu, Y., Piao, C., Chen, Y., Zhou, Y., Wang, D., Yu, H., et al. (2019). Soybean residue (okara) fermentation with the yeast Kluyveromyces marxianus. Food Bioscience, 31, Appendix A. Supplementary data 100439. https://doi.org/10.1016/j.fbio.2019.100439 Hur, S. J., Lee, S. Y., Kim, Y.-C., Choi, I., & Kim, G.-B. (2014). Effect of fermentation on the antioxidant activity in plant-based foods. Food Chemistry, 160, 346–356. https:// Supplementary data to this article can be found online at https://doi. doi.org/10.1016/j.foodchem.2014.03.112 org/10.1016/j.lwt.2021.111809. Jarpa-Parra, M., Bamdad, F., Wang, Y., Tian, Z., Temelli, F., Han, J., et al. (2014). Optimization of lentil protein extraction and the influence of process pH on protein structure and functionality. Lebensmittel-Wissenschaft und -Technologie- Food Science References and Technology, 57(2), 461–469. https://doi.org/10.1016/j.lwt.2014.02.035 Kapilan, R. (2015). Solid state fermentation for microbial products : A review. Archives of Adane, T. D., Yoseph, A. A., & Kusse, G. G. (2019). Determination of chlorogenic acid Applied Science Research, 7(8), 21–25. Retrieved from http://scholarsresearchlibrary. content in beans and leaves of Coffea arabica using UV/Vis spectrometer. African com/archive.html. Journal of Pure and Applied Chemistry, 13(5), 58–63. https://doi.org/10.5897/ Kaprasob, R., Kerdchoechuen, O., Laohakunjit, N., Sarkar, D., & Shetty, K. (2017). ajpac2018.0780 Fermentation-based biotransformation of bioactive phenolics and volatile Adebo, O. A., Njobeh, P. B., Adebiyi, J. A., & Kayitesi, E. (2018). Co-influence of compounds from cashew apple juice by select lactic acid bacteria. Process fermentation time and temperature on physicochemical properties, bioactive Biochemistry, 59, 141–149. https://doi.org/10.1016/j.procbio.2017.05.019 12 N.D.K. Akpabli-Tsigbe et al. L W T 1 49 (2021) 111809 Lee, A. Y., Kim, H. S., Jo, J. E., Kang, B. K., Moon, B. C., Chun, J. M., et al. (2012). and production of an extract rich in chlorogenic acids. Food and Bioproducts Optimization of extraction condition for major iridoid components in fruit of corni Processing, 115, 175–184. https://doi.org/10.1016/j.fbp.2019.04.001 (Cornus officinalis) by UPLC-PDA using response surface methodology. Food Science Sitanggang, A. B., Sinaga, W. S. L., Wie, F., Fernando, F., & Krusong, W. (2020). and Biotechnology, 21(4), 1023–1029. https://doi.org/10.1007/s10068-012-0133-y Enhanced antioxidant activity of okara through solid state fermentation of GRAS Li, S., Jin, Z., Hu, D., Yang, W., Yan, Y., Nie, X., et al. (2020). Effect of solid-state fungi. Food Science and Technology, 40(1), 178–186. https://doi.org/10.1590/ fermentation with Lactobacillus casei on the nutritional value, isoflavones, phenolic fst.37218 acids and antioxidant activity of whole soybean flour. Lebensmittel-Wissenschaft und Soccol, C. R., Scopel, E., Alberto, L., Letti, J., Karp, S. G., Woiciechowski, A. L., et al. -Technologie- Food Science and Technology, 125, 109264. https://doi.org/10.1016/j. (2017). Recent developments and innovations in solid state fermentation. https://doi. lwt.2020.109264 org/10.1016/j.biori.2017.01.002. Biotechnology Research and Innovation, xxx, Li, X., Wei, Y., Xu, J., Xu, N., & He, Y. (2018). Quantitative visualization of lignocellulose xxx–xxx. components in transverse sections of moso bamboo based on ftir macro- and micro- Su, L.-W., Cheng, Y.-H., Hsiao, F. S.-H., Han, J.-C., & Yu, Y.-H. (2018). Optimization of spectroscopy coupled with chemometrics. Biotechnology for Biofuels, 11(1), 1–16. mixed solid-state fermentation of soybean meal by Lactobacillus species and https://doi.org/10.1186/s13068-018-1251-4 Clostridium butyricum. Polish Journal of Microbiology, 67(3), 297–305. Loow, Y.-L., & Wu, T. Y. (2018). Transformation of oil palm fronds into pentose sugars Tang, W. lin, Zhang, M., & Fang, Z. (2015). Optimization of ultrasound-assisted- using copper (II) sulfate pentahydrate with the assistance of chemical additive. extraction of porcine placenta water-soluble proteins and evaluation of the Journal of Environmental Management, 216, 192–203. https://doi.org/10.1016/j. antioxidant activity. Journal of Food Science & Technology, 52(7), 4042–4053. jenvman.2017.04.084 https://doi.org/10.1007/s13197-014-1444-1 Loow, Y.-L., Wu, T. Y., Lim, Y. S., Tan, K. A., Siow, L. F., Md Jahim, J., et al. (2017). Taylor, J. R., & Duodu, K. G. (2015). Effects of processing sorghum and millets on their Improvement of xylose recovery from the stalks of oil palm fronds using inorganic phenolic phytochemicals and the implications of this to the health-enhancing salt and oxidative agent. Energy Conversion and Management, 138, 248–260. https:// properties of sorghum and millet food and beverage products. Journal of the Science doi.org/10.1016/j.enconman.2016.12.015 of Food and Agriculture, 95(2), 225–237. https://doi.org/10.1002/jsfa.6713 Magro, A. E. A., Silva, L. C., Rasera, G. B., & de Castro, R. J. S. (2019). Solid-state Torres-Mancera, M. T., Baqueiro-Peña, I., Figueroa-Montero, A., Rodríguez-Serrano, G., fermentation as an efficient strategy for the biotransformation of lentils: Enhancing González-Zamora, E., Favela-Torres, E., et al. (2013). Biotransformation and their antioxidant and antidiabetic potentials. Bioresources and Bioprocessing, 6(1). improved enzymatic extraction of chlorogenic acid from coffee pulp by filamentous https://doi.org/10.1186/s40643-019-0273-5 fungi. Biotechnology Progress, 29(2), 337–345. https://doi.org/10.1002/btpr.1696 Mintah, B. K., He, R., Agyekum, A. A., Dabbour, M., Golly, M. K., & Ma, H. (2020). Edible Verduzco-Oliva, R., & Gutierrez-Uribe, J. A. (2020). Beyond enzyme production: Solid insect protein for food applications: Extraction, composition, and functional state fermentation (SSF) as an alternative approach to produce antioxidant properties. Journal of Food Process Engineering, 43(4), 1–12. https://doi.org/ polysaccharides. Sustainability, 12(2). https://doi.org/10.3390/su12020495 10.1111/jfpe.13362 Wang, H., Zhang, Q., Mujumdar, A. S., Fang, X.-M., Wang, J., Pei, Y.-P., et al. (2020). Musa, A., Gasmalla, M. A. A., Ma, H., Sarpong, F., Wali, A., Awad, F. N., et al. (2019). High-humidity hot air impingement blanching (HHAIB) efficiently inactivates Effect of a multi-frequency counter-current S-type ultrasound pretreatment on the enzymes, enhances extraction of phytochemicals and mitigates brown actions of chili defatted corn germ protein: Enzymatic hydrolysis, ACE inhibitory activity and pepper. Food Control, 111(September 2019), 107050. https://doi.org/10.1016/j. structural characterization. Food and Function, 10(9), 6020–6029. https://doi.org/ foodcont.2019.107050 10.1039/c9fo01531k Wianowska, D., & Gil, M. (2019). Recent advances in extraction and analysis procedures Nabavi, S. F., Tejada, S., Setzer, W. N., Gortzi, O., Sureda, A., Braidy, N., et al. (2017). of natural chlorogenic acids. Phytochemistry Reviews, 18(1), 273–302. https://doi. Chlorogenic acid and mental diseases: From chemistry to medicine. Current org/10.1007/s11101-018-9592-y Neuropharmacology, 15(4), 471–479. https://doi.org/10.2174/ Yang, J., Wu, X., Chen, H., Sun-waterhouse, D., Zhong, H., & Cui, C. (2019). A value- 1570159x14666160325120625 added approach to improve the nutritional quality of soybean meal byproduct: Naveed, M., Hejazi, V., Abbas, M., Kamboh, A. A., Khan, G. J., Shumzaid, M., et al. Enhancing its antioxidant activity through fermentation by Bacillus (2018). Chlorogenic acid (cga): A pharmacological review and call for further amyloliquefaciens SWJS22. Food Chemistry, 272, 396–403. https://doi.org/10.1016/ research. Biomedicine & Pharmacotherapy, 97, 67–74. https://doi.org/10.1016/j. j.foodchem.2018.08.037 biopha.2017.10.064 Yazid, N. A., Barrena, R., Komilis, D., & Sánchez, A. (2017). Solid-state fermentation as a Pratt, D. E., & Birac, P. M. (1979). Source of antioxidant activity of soybeans and soy novel paradigm for organic waste valorization: A review. Sustainability, 9(2), 1–28. products. Journal of Food Science, 44, 1720–1722. https://doi.org/10.1111/j.1365- https://doi.org/10.3390/su9020224 2621.1979.tb09125.x Zhang, S. T., Shi, Y., Zhang, S. L., Shang, W., Gao, X. Q., & Wang, H. K. (2014). Whole Ruenroengklin, N., Zhong, J., Duan, X., Yang, B., Li, J., & Jiang, Y. (2008). Effects of soybean as probiotic lactic acid bacteria carrier food in solid-state fermentation. various temperatures and pH values on the extraction yield of phenolics from litchi Food Control, 41(1), 1–6. https://doi.org/10.1016/j.foodcont.2013.12.026 fruit pericarp tissue and the antioxidant activity of the extracted anthocyanins. Zhou, T., Huo, R., Kwok, L. Y., Li, C., Ma, Y., Mi, Z., et al. (2019). Effects of applying International Journal of Molecular Sciences, 9(7), 1333–1341. https://doi.org/ Lactobacillus helveticus H9 as adjunct starter culture in yogurt fermentation and 10.3390/ijms9071333 storage. Journal of Dairy Science, 102(1), 223–235. https://doi.org/10.3168/ Santos da Silveira, J., Durand, N., Lacour, S., Belleville, M. P., Perez, A., Loiseau, G., et al. jds.2018-14602 (2019). Solid-state fermentation as a sustainable method for coffee pulp treatment 13