Banana Peels as Potential Prebiotic and Functional Ingredient

This study aims to determine the prebiotic potential of the banana peel on the growth of probiotic Lactobacillus spp. in vitro and to utilize the peel as a functional ingredient in preparing biscuits. Peels of dessert banana ( pisang berangan ) and plantain ( pisang nangka ) were oven-dried and homogenized, and the total sugar content was determined. Subsequently, different cultivation media were made by substituting the carbon source with Banana Peel Powder (BPP), Plantain Peel Powder (PPP), glucose, and inulin. These media were later fermented with probiotic Lactobacillus spp., extracted from a probiotic drink. The growth performance was accessed following 24 hours of incubation. BPP and PPP were incorporated into the preparation of biscuits as functional ingredients. A portion of wheat flour was substituted with 10%, 20%, and 30% of BPP and PPP, respectively, during the preparation of the biscuits. These biscuits were then analyzed for proximate composition, physical properties, and estimated Glycaemic Index (eGI). The supplementation of BPP and PPP in the media improved the probiotic bacteria's growth rate and generation time as the media had a significantly higher amount of Lactobacillus spp. compared to others. Both BPP-and PPP-supplemented media had significantly low pH, indicating intense metabolic activity of the bacteria utilizing the peels. Results also showed significant differences in the total dietary fiber and protein content of BBP-and PPP-incorporated biscuits. The addition of BPP and PPP did not significantly affect the physical properties of the biscuit, and such incorporation resulted in lower eGI when compared to the control. BPP and PPP possess potential prebiotic properties and can be utilized as functional ingredients. Further study is warranted to explore other prebiotic properties of banana peels and to investigate consumers' acceptance of banana peel-incorporated foods.


INTRODUCTION
Banana is a popular fruit widely cultivated and consumed in many tropical nations, including Malaysia, Thailand, and Indonesia (Zaini et al. 2022).To fulfil the increasing local demand for bananas, banana trees are widely cultivated in Malaysia (Fizar et al. 2022).High production of bananas is associated with the generation of banana by-products.Banana peel is usually discarded, thus generating massive amounts of household and industrial food waste (Ahmed et al. 2006).
Banana peel is an underpredicted functional food source as it contains soluble fibre oligosaccharide (Liang et al. 2022).Fructooligosaccharide was found in banana peels and had prebiotic properties (Azam et al. 2020).Besides soluble fibre, banana peel also contains insoluble fibre.Since banana peels have a remarkable amount of fibre; thus, they may exhibit a low Glycaemic Index (GI) value when the wheat flour is substituted with a specific ratio of banana peel powder (Chakraborty et al. 2021).Indeed, utilizing this agricultural waste as a functional food or value-added food item is critical for long-term sustainability, in line with the Sustainable Development Goal (SDG) No.12, which aims to ensure sustainable consumption and production patterns.
There is limited study on the prebiotic potential of banana peels.A few published studies have shown contradicted findings of banana peels (Hernández-Alcántara et al. 2016).Other studies have reported banana peels' antioxidant and antimicrobial properties, but there is still limited study of banana peels on the activity of Lactobacillus spp.growth.Besides, Tan et al.
there is no study comparing the GI of banana peel-incorporated biscuits from several types of bananas.Only one study compared GI between wheat-based and unripe banana peel biscuits (Bakar et al. 2020).Therefore, this study aims to determine the prebiotic potential of the banana peel on the growth of probiotic Lactobacillus spp. in vitro and to utilize the peel as a functional ingredient in the preparation of biscuits.

Design, location, and time
This was an in vitro experiment and it was carried out at the Nutrition Laboratory, Department of Nutrition, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, from July 2022 until February 2023.

Procedure
Preparation of banana peel powder.The banana peel powder was prepared based on the method of Zahid et al. (2021) with some modifications.Dessert banana (pisang berangan) and plantain (pisang nangka) peels were washed with tap water and cut into small uniform pieces.They were dried in an oven at 50°C for 24 hours and ground into powder using a grinder.The dried Banana Peel Powder (BPP) and Plantain Peel Powder (PPP) were sieved to maintain a uniform particle size, packed in airtight bags, and stored in the chiller at 4°C until further use.
Determination of carbohydrate content in banana peel and banana peel-incorporated biscuits.The anthrone method by Ludwig and Goldberg (1956) was used to determine carbohydrate content.Samples were heated in a boiling water bath for 8 minutes after adding the anthrone reagent, and the absorbance was measured using a spectrophotometer at 620 nm.The carbohydrate content was extrapolated from a standard curve of glucose with varying concentrations (0.2-1 g/L).
Propagation of Lactobacillus spp.. Briefly, 0.01 mL of probiotic drink was aseptically spread onto MRS agar and incubated at 37°C aerobically for 24-48 hours.Then, one colony of Lactobacillus spp. was inoculated into MRS broth and incubated at 37°C aerobically for 48 hours.At 3-hour intervals, 1 mL was taken, where the Optical Density (OD) was measured, and log colony forming unit per milliliter (log CFU/mL) was calculated.A graph of log colony forming units (log CFU/mL) against optical density (OD600) of Lactobacillus spp. was plotted and these variables were positively correlated.Then the graph was used to extrapolate the amount of Lactobacillus spp.
Assessment of bacterial growth performance.The preparation of the batch culture medium and the fermentation of Lactobacillus spp. was based on the method in Azam et al. (2020) with some modifications.The amount of BPP and PPP with 2.39% w/v of carbohydrate content was extracted following the procedure by Pereira et al. (2018) to obtain the oligosaccharide.Peptone (1g/100 mL), yeast extract (0.5 g/100 mL), Tween 80 (0.1 mL/100 mL), and L-cysteine Prebiotic and functional properties of banana peels hydrochloride (0.05 g/100 mL) were mixed in five vessels, respectively, with the extracted BPP and PPP solutions, 2.39% w/v of inulin, 2.39% w/v of glucose, and distilled water (control).All media were sterilized using an autoclave.The final volume in each vessel was 100 mL.One (1) colony of Lactobacillus spp. was inoculated separately into five different vessels containing batch culture media and incubated at 37°C for 24h.Samples from each vessel were taken (10 mL) at 15, 18, 21, and 24 hours of fermentation at 37°C for enumeration of Lactobacillus spp.and pH measurement.Preparation and proximate composition of BPP-and PPP-incorporated biscuits.One hundred (100) g of wheat flour, 56 g of sugar, 23.5 g of shortening, 1.1 g NaHCO3, 0.89 g NaCl, and 12 mL of milk were mixed to form biscuit dough (Arun et al. 2015).The Wheat Flour (WF) was substituted with 10% (BPP10 or PPP10), 20% (BPP20, or PPP20), and 30% (BPP30 or PPP30) of BPP and PPP, respectively.The doughs were then molded into walnut-sized balls approximately 4 cm in diameter, placed on a greased tray, and baked in the oven for 15 minutes at 180°C or until golden brown.The biscuits were cooled at room temperature and stored in a polyethylene bag before analysis.Evaluation of moisture (gravimetric method with oven drying -AOAC 2000), ash (carbonization and incineration in the furnace -AOAC 2000), fat (Soxhlet method -AOAC 1996), protein (Kjeldahl -AOAC 2000), carbohydrate (Anthrone method), and total dietary fiber (enzymatic-gravimetric method -AOAC 991.43) of the cookies were conducted according to AOAC methods (Helrich 1990).

Determination of mean growth rate constant (k) and mean generation time, (g) of
Determination of physical properties of BPP-and PPP-incorporated biscuits.The diameter (D) and thickness (T) of the biscuits were measured using a ruler (Abu Bakar et al. 2018).The diameter of the biscuits was determined by arranging six biscuits edge to edge, and the overall diameter of the biscuits was measured.Then, the average diameter of the biscuits was recorded after six times of biscuit rearrangement.The procedure was repeated by stacking six biscuits on top of one another to determine the thickness of the biscuits.The averages of the diameter and thickness were reported in centimeters (cm) and used to calculate the spread ratio.
Estimation of glycaemic index.To predict the glycaemic response and estimate the Glycaemic iIndex (eGI) of the BPP-and PPP-incorporated biscuits, an in vitro starch hydrolysis was conducted according to Gibson et al. (2011).Briefly, samples were added into potassium phosphate buffer (pH 6.9) and kept at 37°C.Hydrochloric acid was used to adjust the pH of the samples to pH 2.5.Then, pepsin was added and placed in a water bath for 1 hour at 37°C.Potassium hydroxide was used to increase the samples' pH to 6.8.The mixtures were transferred into a dialysis tube after adding 2 mL α-amylase.Subsequently, the dialysis tube was placed in a buffer solution while the flask was set in a shaking water bath at 37°C.Forty (40) mL of buffer solution was extracted every 30-minute interval within 3 hours to determine the carbohydrate hydrolysis rate from the dialysis tube.3,5-Dinitrosalicylic Acid (DNSA) reagent was used to estimate the amount of reduced sugar.The concentration-over-time curve (area under the curve) was determined, and the Hydrolysis Index (HI) was calculated by comparing the Area Under the Curve (AUC) of the samples and the AUC of maltose, which acts as the standard of reference food.Then, HI was calculated as HI=(Area under the curve, sample / Area under the curve, reference) x 100% and it was later used to determine eGI.

Data analysis
Statistical analysis was conducted by using IBM SPSS Statistics 26.Results were expressed as means±standard deviation.One-Way ANOVA analysis was used to compare the data, and Post Hoc Comparison (Tukey HSD) was used to determine significant differences between the variables.P-values of <0.05 were regarded to be significant.
Tan et al.

Carbohydrate content in BPP and PPP
BPP had 34.0%carbohydrates, while PPP contained higher carbohydrates, 61.9%.According to Tsado et al. (2021), dessert banana peel and plantain peel contain around 63.8% and 74.1% of carbohydrates, respectively, compared to this study's result.The varied nutritional content is due to the genetic makeup, environmental factors, and addition of nutrients to the soil (Durgadevi et al. 2019).The ripening state can also affect the composition of the peels (Khawas & Deka 2016).Plantain peel contains more total starch and soluble sugar (glucose, fructose, and sucrose) than dessert banana peel at any maturation stage (Emaga et al. 2007).

Growth performance of Lactobacillus spp. in BPP-and PPP-supplemented media
The bacteria count significantly differed in media with different carbon sources (p<0.05).Specifically, the growth of Lactobacillus spp.media supplemented with BPP and PPP was significantly higher than inulin, glucose, and without carbon source (Table 1).The finding of this study was in line with previous studies that showed the banana peel promotes the growth of probiotic bacteria (Syahpura et al. 2019;Zahid et al. 2021).Study showed that fructooligosaccharide is present in both dessert banana and plantain peel and possess prebiotic properties (Azam et al. 2020).Dessert banana peel contains more fructooligosaccharide than plantain peel (Syahpura et al. 2019).Thus, it can be postulated that the higher count of Lactobacillus spp. in a medium with BPP could be due to the presence of fructooligosaccharide.
The findings of Rossi et al. (2005) indicated that the potential of fructooligosaccharide in promoting the growth of Lactobacillus spp. was higher than inulin, supporting the findings of this study.The growth of Lactobacillus spp. is more rapid with the presence of fructooligosaccharide compared to inulin because short-chain fructooligosaccharide has a lower degree of polymerization (<10) compared to long-chain inulin (>20).Inulin is a fructan with a longer chain length and larger molecule, which requires more time for the bacteria to ferment (Parhi et al. 2021).Thus, bacteria take a shorter time to ferment fructooligosaccharide.
As shown in Table 1, there was a significant difference in the final pH of the media with a different carbon source (p<0.05).The pH of media supplemented with BPP and PPP was significantly lower than inulin, glucose, and no carbon source.The finding of this study was in line with previously reported study that showed the banana peel as a carbon source has a pH-lowering effect (Safdari et al. 2021).Lactobacillus spp. is lactic acid bacteria that can ferment carbohydrates to produce lactic acid (Wang et al. 2021).Thus, the activity of Lactobacillus spp.can be detected by measuring the pH of customized media.
The result showed that there was a significant difference between the mean growth rate constant (k) and mean generation time (g) of Lactobacillus spp. in media with a different carbon source (p<0.05)(Table 1).The mean growth rate The generation time (g) of Lactobacillus spp. in a medium without a carbon source is significantly the longest.From the finding, it is estimated that around 33.83 hours to complete binary fission and produce the next generation, without any supplementation.Hence, the environment without a carbon source is the least preferred by Lactobacillus spp., as the bacteria can hardly undergo cell division.In contrast, Lactobacillus spp.only require around 1.11 hours to complete cell division in media with dessert banana peel.It shows that the environment is the most preferred by the Lactobacillus spp., where it divides actively, as found in the present study.
The generation time differs widely among different bacteria.Even within the bacteria of the same species, the generation time might also be different due to the environmental factor (Mason 1935).The finding of this study is slightly different from the previous study, where the generation time of Lactobacillus spp. is between 0.62 hours and 2.25 hours (Mason 1935;Ahmed et al. 2006).This indicated that the Lactobacillus spp.used in this study duplicate slower compared to other species.

Proximate composition of BPP-and PPPincorporated biscuits
As shown in Table 2, the moisture content in BPP-incorporated biscuits was significantly lower as the level of BPP substitution increased (p<0.05).There was a significant difference and increasing trend in ash content between control and BPP-incorporated biscuits as the BPP level increased, while only PPP30 biscuits showed significance (p<0.05).Besides, the protein content and total dietary fiber (TDF) were significantly different between the control and BPP-and PPPincorporated biscuits (p<0.05).Lower moisture content in BPP-and PPP-incorporated biscuits, indicates longer shelf life (Amarasinghe et al. 2021).Mahloko et al. (2019) also supported that microbial growth, mold, and insect manifestation could be resisted when the moisture content is less than 14%.It was expected that BPP-and PPPincorporated biscuits to have low protein as their source from the wheat flour has been reduced.Nonetheless, these biscuits could be alternative sources of dietary fiber because they had higher TDF than wheat flour biscuits, as found in this study.

Diameter, thickness, and spread ratio of BPPand PPP-incorporated biscuits
There was no significant difference in the diameter, thickness, and spread ratio of the biscuits (p<0.05)(Table 3).The results were contradicted by other studies since the diameter of the biscuits tend to decrease as the level of BPP increases (Arun et al. 2015;Abu Bakar et al. 2018;Mahloko et al. 2019).On the other hand, Mahloko et al. (2019) demonstrated that biscuits containing both BPP and PPP also did not show any significance from the control (p<0.05).As a result, the insignificant physical properties in both BPP-and PPP-incorporated biscuits Tan et al.
demonstrated that the substitution of BPP and PPP did not cause major changes in the physical characteristics when compared to the standard one.

Hydrolysis Index (HI) and Estimated Glycaemic Index (eGI)
There was a significant difference in HI and eGI of control biscuits with both BPPand PPP-incorporated biscuits (p<0.05)(Table 4).On the other hand, there was no significant difference in HI and eGI between BPP-and PPPincorporated biscuits (p>0.05).The biscuits with the formulation of 100% wheat flour showed significantly greater values of HI and eGI than the other formulations.The incorporation of BPP and PPP in biscuits slows the starch hydrolysis rate due to the presence of fiber content.Indeed, BPP-and PPP-incorporated biscuits have much higher fiber content compared to wheat flour, thus resulting in lower HI and eGI.Furthermore, α-amylase has a lower capacity to digest the dietary fiber contained in the biscuits, thus, resulting in less digestion of starch (Chakraborty et al. 2021).High-fiber food allows the stomach to give signals to the liver to cease glucose production.Therefore, low hydrolysis rate and increment of glucose level can be seen over time.The lower hydrolysis rate determines a lower GI value (Bakar et al. 2020).Both BPP-and PPP-incorporated biscuits meet with the classification of medium GI food (56-69) compared to control biscuits that fell into the high GI food category (≥70) (Dona et al. 2010).

CONCLUSION
The cultivation of Lactobacillus spp. in customized batch culture media containing BPP and PPP as the sole carbon source resulted in intense bacterial metabolic activities as observed by high cell counts (log CFU/mL), decreased in pH values and acceptable results of mean growth rate constant (k) as well as mean generation time (g) besides obtaining satisfactory prebiotic properties.Besides, incorporating BPP and PPP in biscuits can increase the total dietary fiber content and decrease the protein content of biscuits.Furthermore, such incorporation slowed down starch digestion and possessed lower eGI.In contrast, the physical properties of both BPP-and PPP-incorporated biscuits were not significantly affected.
For the future study, the use of probiotic bacteria with different species and strains (Bifidobacterium spp.and Eubacterium spp.) and pathogenic bacteria (E.coli) are recommended to determine the prebiotic effect of the banana peel on different probiotics.Moreover, the banana peel fiber can be introduced for Lactobacillus  Prebiotic and functional properties of banana peels spp.fermentation since insoluble fiber might also promote the growth of probiotics.Since the finding presented here is based on in vitro study, further in vivo and human research can be pursued to validate the prebiotic properties of banana peel and its incorporated biscuits.In addition, sensory evaluation is recommended to assess consumer satisfaction and acceptance towards the appearance, odor, taste, color, hardness, and texture of the biscuit incorporated with banana peels.
Lactobacillus spp.. Formulas from Willey et al. (2018) were used to calculate the mean growth rate constant (k) and mean generation time (g) as follows:

Table 1 .
Final bacteria count (log CFU/mL), final pH, and growth performance of Lactobacillus spp.employed with different carbon sources Prebiotic and functional properties of banana peels constant (k) and mean generation time (g) of Lactobacillus spp. in a medium supplemented with BPP as a carbon source was significantly higher than in a PPP-supplemented medium.

Table 2 .
Proximate composition of BPP-and PPP-incorporated biscuits

Table 4 .
Hydrolysis index and estimated glycaemic index of BPP-and PPP-incorporated biscuits