The Effect of Dekkera bruxellensis Concentration and Inoculation Time on Biochemical Changes and Cellulose Biosynthesis by Komagataeibacter intermedius

Bacterial Cellulose (BC) is a biopolymer with numerous applications. The growth of BC-producing bacteria, Komagataeibacter intermedius, could be stimulated by Dekkera bruxellensis, however, the effect on BC yield needs further investigation. This study investigates BC production and biochemical changes in the K. intermedius-D. bruxellensis co-culture system. D. bruxellensis was introduced at various concentrations (103 and 106 CFU/mL) and inoculation times (days 0 and 3). BC yield was ~24% lower when D. bruxellensis was added at 103 CFU/mL compared to K. intermedius alone (0.63 ± 0.11 g/L). The lowest BC yield was observed when 103 CFU/mL yeast was added on day 0, which could be compromised by higher gluconic acid production (10.08 g/L). In contrast, BC yields increased by ~88% when 106 CFU/mL D. bruxellensis was added, regardless of inoculation time. High BC yield might correlate with faster sugar consumption or increased ethanol production when 106 CFU/mL D. bruxellensis was added on day 0. These results suggest that cell concentration and inoculation time have crucial impacts on species interactions in the co-culture system and product yield.


Introduction
Bacterial cellulose (BC) refers to the biomaterial of glucose monomers produced by bacteria [1]. Cellulose is mainly obtained from plants. However, BC was reported to possess attractive properties, including higher purity, crystallinity, tensile strength, waterholding capacity, thermal stability and malleability [2,3]. Hence, BC is used for different applications, such as wound healing, drug delivery, as potential electric capacitors, and as a food additive [1,2].
The bacteria genus Komagataeibacter is well known for its ability to produce BC [4]. This genus is also abundant during the production of kombucha, a fermented sweetened tea, where they also form BC [5]. However, kombucha is not produced by the bacteria alone but with the co-culture of various yeasts species [5]. It is believed that the yeasts break down sucrose to glucose and fructose, while the bacteria utilize it to form BC and carboxylic acids [4,6]. In addition, yeasts can also produce ethanol, which was reported to increase BC production and reduce the presence of non-BC producing bacteria [7,8]. Lastly, the bacteria can convert the ethanol to acetic acid, which has also been found to increase BC yield [9]. Due to the ability to increase the amount of reducing sugars and advantageous metabolites, it is, therefore, likely that the presence of yeasts can potentially promote BC production.
Komagataeibacter intermedius is one of the BC-producing bacteria found in kombucha and was isolated in a previous study [10]. K. intermedius was reported to produce more BC than Komagataeibacter xylinus, the model BC producer [11][12][13]. K. intermedius also produced a high BC yield in sugarcane molasses, a cheap alternative to current commercial Table 1. BC production experimental set up with K. intermedius mono-(B -control ) and co-cultures with D. bruxellensis at 30 • C. D. bruxellensis was added simultaneously at concentrations of 10 6 CFU/mL (BY H0 ) and 10 3 CFU/mL (BY L0 ) and sequentially at 10 6 CFU/mL (BY H3 ) and 10 3 CFU/mL (BY L3 ).

K. intermedius and D. bruxellensis Cell Enumeration
K. intermedius and D. bruxellensis enumeration was carried out by collecting 0.1 mL of the cultivation broth on each sampling day. The samples were then serially diluted up to 10 −8 using 0.85% NaCl solution and were plated on HS agar supplemented with 1% v/v acetic acid and PDA with 2% w/v NaCl for K. intermedius and D. bruxellensis, respectively. A previous study demonstrated that the addition of 2% w/v NaCl or 1% v/v acetic acid was able to inhibit the growth of either K. intermedius or D. bruxellensis, respectively, which allowed the observation of each microbe from the co-culture [19]. The colonies were enumerated after 3 days of incubation at 30 • C.
Specific growth rate µ (h −1 ) were calculated by Equation (1) [20]: where X t and X 0 are the microbial population (CFU mL −1 ) at t and initial time, respectively; t and t 0 are the t and initial time when the sample is measured, respectively; µ is specific growth rate (1 h −1 ).

Biochemical Analysis
Changes in pH during the cultivation were monitored using a pH meter (ST300, OHAUS, Parsippany, NJ, USA). Prior to biochemical analysis, samples were centrifuged at 1000× g for 10 min at room temperature. Then, the supernatant was collected into a new tube, and the process was repeated until no pellet was visually observed. Sugars (glucose and fructose) were measured using the K-SUFRG kit, ethanol was measured using K-ETOH kit, and acids (gluconic and glucuronic acids) were measured using K-GATE and K-URONIC kits (Megazyme, International Ireland Ltd., Bray, Ireland), according to the manufacturer's instructions.
Free amino nitrogen (FAN) concentration was measured using the ninhydrin analysis method [21]. The ninhydrin color reagent consisted of 0.3 g fructose, 6 g KH 2 PO 4 , 10 g Na 2 HPO 4 and 0.5 g ninhydrin dissolved in 100 mL of distilled water. A solution mixture was prepared by dissolving 2 g of potassium iodide into 600 mL of distilled water and 400 mL of 96% ethanol. The samples were first diluted 50 times using distilled water. Then, 2 mL of the diluted sample was mixed with 1 mL of the ninhydrin color reagent. The mixture was boiled for 16 minutes and left to cool down to room temperature using an ice bath for 20 min. Afterwards, 5 mL of the previously prepared solution mixture was added to each sample, and the absorbance was measured at 570 nm.

BC Yield Measurement
After 14 days of incubation, BC formed on the surface of the medium were collected and rinsed in distilled water until no excess media remained on the pellicle. The BC samples were then treated using 1 M NaOH at 80 • C for at least an hour to remove the remaining microbes and leftover broth. The treated samples were then rinsed in distilled water to remove the NaOH before being oven dried at 60 • C overnight until a stable weight was achieved. The dried BC weight was measured using an analytical balance, and the BC yield was reported as gram dry weight per liter of culture volume (g/L).

Statistical Analysis
The data were analyzed using R studio program [22] for one-way analysis of variance (ANOVA). One-way ANOVA followed by Tukey HSD test were carried out to investigate if the presence of yeast in each sample had any effect on the final BC yield and glucuronic acid concentration. A two-way ANOVA was also done on BC yield to investigate the interaction effects between the yeast concentration and inoculation time. The effect was considered statistically significant if the p-value was less than or equal to the selected significance level (p-values < 0.05).

Effect of D. bruxellensis Concentration and Inoculation Time on Bacteria-Yeast Interactions
In kombucha, yeasts are hypothesized to support bacteria by breaking down sucrose or providing ethanol as an additional carbon source [7]. Bacteria-yeast co-cultures in kefir and sourdough bread were also known to support each other by producing other essential nutrients, such as vitamin B6 and amino acids [23,24]. However, other studies reported that depending on the initial cell ratio, a microbe can outcompete the other by depleting the available nutrients or producing inhibitory compounds [25,26]. Furthermore, adding yeast sequentially can also impact bacterial fermentation, which has been observed in sour beer and soy sauce [17,18]. Therefore, K. intermedius productivity may also be affected by the cell ratio and inoculation time of D. bruxellensis.
As shown in Figure 1, K. intermedius in all samples had a similar specific growth rate (~0.374 h −1 ), indicating that the D. bruxellensis may not be antagonistic. Furthermore, K. intermedius in samples with D. bruxellensis added sequentially (BY H3 and BY L3 ) were also able to maintain a high concentration up to day 5 of the experiment. However, similar to B -control , K. intermedius in BY H3 and BY L3 also had no observable growth by the end of cultivation. On the other hand, samples with simultaneous addition of D. bruxellensis had a high concentration of K. intermedius up to day 7 and had observed bacterial growth until the end of cultivation, with 6.14 ± 0.28 log CFU/mL in BY H0 and 3.01 ± 0.54 log CFU/mL in BY L0 . The simultaneous addition of D. bruxellensis may have provided the K. intermedius with more resources, such as amino acids and ethanol, which allowed K. intermedius to survive for an extended time [6,8]. with more resources, such as amino acids and ethanol, which allowed K. intermedius to survive for an extended time [6,8]. D. bruxellensis population in yeast mono (Y-control) and co-cultures was also monitored. The initial counts of D. bruxellensis were lower by 1 log CFU/mL than the adjusted goal, which could be attributed to the accuracy of OD to log CFU/mL conversion, resulting from variations in yeast's physiological state and ability to grow on the agar. Then, adding the inoculum into the fermentation media would further dilute the cell concentration.
D. bruxellensis in Y-control and BYH0 were observed to grow at a similar rate reaching a final concentration of ~8 log CFU/mL by day 14 (Figure 2). While D. bruxellensis in BYL0 experienced a significantly slower initial growth, it grew to the same concentration as Ycontrol and BYH0 by day 7. Samples with the sequential addition of D. bruxellensis grew even slower. BYH3 experienced an initial decrease before fluctuating at approximately 5 log CFU/mL until the end of cultivation, while no growth of D. bruxellensis could be observed throughout incubation in BYL3. This suggests that adding D. bruxellensis at the later stage (on day 3), especially at a lower concentration, could be unfavorable for yeast. By that time, K. intermedius almost enters the stationary phase ( Figure 1), leaving less resources available and more harmful inhibitory compounds, such as acetaldehyde [27,28], creating a hostile environment for D. bruxellensis. intermedius at 30 C. D. bruxellensis was added simultaneously at concentrations of 10 6 CFU/mL (BYH0) and 10 3 CFU/mL (BYL0) and sequentially at 10 6 CFU/mL (BYH3) and 10 3 CFU/mL (BYL3).

Effect of D. bruxellensis Concentration and Inoculation Time on pH Changes
Komagataeibacter are well known for producing gluconic, glucuronic, and acetic acids [5]. D. bruxellensis is also known to produce acetic acid [29]. The accumulation of acids D. bruxellensis population in yeast mono (Y -control ) and co-cultures was also monitored. The initial counts of D. bruxellensis were lower by 1 log CFU/mL than the adjusted goal, which could be attributed to the accuracy of OD to log CFU/mL conversion, resulting from variations in yeast's physiological state and ability to grow on the agar. Then, adding the inoculum into the fermentation media would further dilute the cell concentration.
D. bruxellensis in Y -control and BY H0 were observed to grow at a similar rate reaching a final concentration of~8 log CFU/mL by day 14 (Figure 2). While D. bruxellensis in BY L0 experienced a significantly slower initial growth, it grew to the same concentration as Y -control and BY H0 by day 7. Samples with the sequential addition of D. bruxellensis grew even slower. BY H3 experienced an initial decrease before fluctuating at approximately 5 log CFU/mL until the end of cultivation, while no growth of D. bruxellensis could be observed throughout incubation in BY L3 . This suggests that adding D. bruxellensis at the later stage (on day 3), especially at a lower concentration, could be unfavorable for yeast. By that time, K. intermedius almost enters the stationary phase ( Figure 1), leaving less resources available and more harmful inhibitory compounds, such as acetaldehyde [27,28], creating a hostile environment for D. bruxellensis. with more resources, such as amino acids and ethanol, which allowed K. intermedius to survive for an extended time [6,8]. D. bruxellensis population in yeast mono (Y-control) and co-cultures was also monitored. The initial counts of D. bruxellensis were lower by 1 log CFU/mL than the adjusted goal, which could be attributed to the accuracy of OD to log CFU/mL conversion, resulting from variations in yeast's physiological state and ability to grow on the agar. Then, adding the inoculum into the fermentation media would further dilute the cell concentration.
D. bruxellensis in Y-control and BYH0 were observed to grow at a similar rate reaching a final concentration of ~8 log CFU/mL by day 14 (Figure 2). While D. bruxellensis in BYL0 experienced a significantly slower initial growth, it grew to the same concentration as Ycontrol and BYH0 by day 7. Samples with the sequential addition of D. bruxellensis grew even slower. BYH3 experienced an initial decrease before fluctuating at approximately 5 log CFU/mL until the end of cultivation, while no growth of D. bruxellensis could be observed throughout incubation in BYL3. This suggests that adding D. bruxellensis at the later stage (on day 3), especially at a lower concentration, could be unfavorable for yeast. By that time, K. intermedius almost enters the stationary phase ( Figure 1), leaving less resources available and more harmful inhibitory compounds, such as acetaldehyde [27,28], creating a hostile environment for D. bruxellensis. intermedius at 30 C. D. bruxellensis was added simultaneously at concentrations of 10 6 CFU/mL (BYH0) and 10 3 CFU/mL (BYL0) and sequentially at 10 6 CFU/mL (BYH3) and 10 3 CFU/mL (BYL3).

Effect of D. bruxellensis Concentration and Inoculation Time on pH Changes
Komagataeibacter are well known for producing gluconic, glucuronic, and acetic acids [5]. D. bruxellensis is also known to produce acetic acid [29]. The accumulation of acids

Effect of D. bruxellensis Concentration and Inoculation Time on pH Changes
Komagataeibacter are well known for producing gluconic, glucuronic, and acetic acids [5]. D. bruxellensis is also known to produce acetic acid [29]. The accumulation of acids during cultivation may impact the pH, affecting their growth and eventually the BC production. BC production can be inhibited when the pH is outside of the bacteria's optimum range [30]. Therefore, the pH of the media was monitored throughout the incubation. It was found that the pH of all samples increased to a final pH of at least 8, which was consistent with the previous work [10]. Such an increasing pH trend was rarely observed in acetic acid bacteria fermentation, which is more likely to cause pH reduction due to organic acid production [5,31]. Previous work reported that the pH increase was associated with the presence of acetate buffer in the medium [10,32]. Furthermore, ammonia production by K. intermedius might have also contributed to the pH increase, which is further discussed in Section 3.6. The ability of K. intermedius to produce BC at a pH of 9 has been previously reported [10,11]. Although the optimal pH for BC production is typically slightly acidic, different strains have also been found to prefer more alkaline conditions [30,33].
Even though the pH in all samples increased, the rate of the increase seemed to be dependent on the concentration and inoculation time of D. bruxellensis ( Figure 3). In B -control , the pH started to increase from day 3 to a final pH of 9.31 ± 0.02. The pH in samples with simultaneous addition of D. bruxellensis also experienced an increase from day 3, but at a slower rate. BY H0 obtained the lowest final pH among all samples (7.96 ± 0.31), followed by BY L0 (8.75 ± 0.1). In samples with the sequential addition of D. bruxellensis, the pH only increased after day 5 but resulted in a similar final pH as B -control .
during cultivation may impact the pH, affecting their growth and eventually the BC production. BC production can be inhibited when the pH is outside of the bacteria's optimum range [30]. Therefore, the pH of the media was monitored throughout the incubation. It was found that the pH of all samples increased to a final pH of at least 8, which was consistent with the previous work [10]. Such an increasing pH trend was rarely observed in acetic acid bacteria fermentation, which is more likely to cause pH reduction due to organic acid production [5,31]. Previous work reported that the pH increase was associated with the presence of acetate buffer in the medium [10,32]. Furthermore, ammonia production by K. intermedius might have also contributed to the pH increase, which is further discussed in Section 3.6. The ability of K. intermedius to produce BC at a pH of 9 has been previously reported [10,11]. Although the optimal pH for BC production is typically slightly acidic, different strains have also been found to prefer more alkaline conditions [30,33].
Even though the pH in all samples increased, the rate of the increase seemed to be dependent on the concentration and inoculation time of D. bruxellensis (Figure 3). In Bcontrol, the pH started to increase from day 3 to a final pH of 9.31 ± 0.02. The pH in samples with simultaneous addition of D. bruxellensis also experienced an increase from day 3, but at a slower rate. BYH0 obtained the lowest final pH among all samples (7.96 ± 0.31), followed by BYL0 (8.75 ± 0.1). In samples with the sequential addition of D. bruxellensis, the pH only increased after day 5 but resulted in a similar final pH as B-control.

Effect of D. bruxellensis Concentration and Inoculation Time on Sugar Concentration during Co-Culture with K. intermedius
Bacteria typically produce BC and other acids from glucose; hence, the production of both relies on the amount of available sugars. Yeasts may provide more glucose by breaking down available sucrose, but they can also potentially outcompete bacteria and consume the reducing sugars instead [7]. Therefore, the amount of reducing sugars was measured throughout the incubation to detect any competition for sugar when D. bruxellensis was added at a lower concentration or added sequentially. Figure 4 shows that K. intermedius in B-control prefers glucose, with 90% consumed by day 7 of the incubation, while fructose concentration remains constant until day 14 ( Figure  4A,B). Similarly, BYH0 experienced the same trend for glucose consumption with approximately 90% of the sugar consumed by day 7. Meanwhile, the glucose consumption rate in BYL0 was slower, with the same amount of sugar consumed by day 14. It was also found that BYH0 and BYL0 utilized fructose as well. When D. bruxellensis was added sequentially (BYH3 and BYL3), an increase in glucose concentration was observed by day 5. By the end of the incubation, BYH3 used up roughly 82% of the glucose, while BYL3 consumed the least

Effect of D. bruxellensis Concentration and Inoculation Time on Sugar Concentration during Co-Culture with K. intermedius
Bacteria typically produce BC and other acids from glucose; hence, the production of both relies on the amount of available sugars. Yeasts may provide more glucose by breaking down available sucrose, but they can also potentially outcompete bacteria and consume the reducing sugars instead [7]. Therefore, the amount of reducing sugars was measured throughout the incubation to detect any competition for sugar when D. bruxellensis was added at a lower concentration or added sequentially. Figure 4 shows that K. intermedius in B -control prefers glucose, with 90% consumed by day 7 of the incubation, while fructose concentration remains constant until day 14 ( Figure 4A,B). Similarly, BY H0 experienced the same trend for glucose consumption with approximately 90% of the sugar consumed by day 7. Meanwhile, the glucose consumption rate in BY L0 was slower, with the same amount of sugar consumed by day 14. It was also found that BY H0 and BY L0 utilized fructose as well. When D. bruxellensis was added sequentially (BY H3 and BY L3 ), an increase in glucose concentration was observed by day 5. By the end of the incubation, BY H3 used up roughly 82% of the glucose, while BY L3 consumed the least amount (~70%). The fructose in both samples showed a slight increase by the end of incubation, which indicates that the D. bruxellensis may also only utilize fructose when glucose is unavailable.
Other studies have reported that K. intermedius could utilize fructose, although they still primarily prefer glucose [14,34]. It is also noted that different strains of Komagataeibacter may prefer different kinds of sugars [35]. In another study it was demonstrated that the co-culture consume the total sugars at a higher rate than bacterial monoculture [10]. Therefore, total sugars may be used more quickly especially when the yeast is added simultaneously [36].

Effect of D. bruxellensis Concentration and Inoculation Time on Gluconic and Glucuronic acid Production by K. intermedius
Gluconic and glucuronic acid are commonly found during kombucha fermentation and result from the oxidation of glucose's 1st and 6th carbon [5,37]. In kombucha, the yeast has been observed to stimulate acid production by acetic acid bacteria, either by providing more readily available carbon sources or by producing other beneficial nutrients [5,6]. However, increased acid production may impact BC production by using more glucose or acidifying the environment [30]. Hence, gluconic and glucuronic acid production were detected in K. intermedius monoculture and co-cultures.
As shown in Figure 5, the gluconic acid concentration in all samples peaked by day 5 and then declined toward the end of the incubation period. The production followed by consumption of gluconic acid has been noted in other studies in Acetobacter xylinum NUST4.2 and Komagataeibacter hansenii [31,38]. While B-control peaked by day 5 at 8.45 ± 0.45 g/L, both co-cultures with simultaneous addition of yeast peaked earlier at day 3. However, BYH0 peaked with 16% less gluconic acid, while BYL0 produced 20% more gluconic acid than B-control. Similar results were also found in another study where a lower yeast ratio in Starmerella davenportii (yeast)-Gluconacetobacter intermedius (bacteria) co-culture produced thrice the amount of gluconic acid compared to bacteria monoculture due to Other studies have reported that K. intermedius could utilize fructose, although they still primarily prefer glucose [14,34]. It is also noted that different strains of Komagataeibacter may prefer different kinds of sugars [35]. In another study it was demonstrated that the co-culture consume the total sugars at a higher rate than bacterial monoculture [10]. Therefore, total sugars may be used more quickly especially when the yeast is added simultaneously [36].

Effect of D. bruxellensis Concentration and Inoculation Time on Gluconic and Glucuronic Acid Production by K. intermedius
Gluconic and glucuronic acid are commonly found during kombucha fermentation and result from the oxidation of glucose's 1st and 6th carbon [5,37]. In kombucha, the yeast has been observed to stimulate acid production by acetic acid bacteria, either by providing more readily available carbon sources or by producing other beneficial nutrients [5,6]. However, increased acid production may impact BC production by using more glucose or acidifying the environment [30]. Hence, gluconic and glucuronic acid production were detected in K. intermedius monoculture and co-cultures.
As shown in Figure 5, the gluconic acid concentration in all samples peaked by day 5 and then declined toward the end of the incubation period. The production followed by consumption of gluconic acid has been noted in other studies in Acetobacter xylinum NUST4.2 and Komagataeibacter hansenii [31,38]. While B -control peaked by day 5 at 8.45 ± 0.45 g/L, both co-cultures with simultaneous addition of yeast peaked earlier at day 3. However, BY H0 peaked with 16% less gluconic acid, while BY L0 produced 20% more gluconic acid than B -control . Similar results were also found in another study where a lower yeast ratio in Starmerella davenportii (yeast)-Gluconacetobacter intermedius (bacteria) co-culture produced thrice the amount of gluconic acid compared to bacteria monoculture due to more glucose produced from sucrose breakdown [6,16]. However, samples with D. bruxellensis added sequentially had a lower gluconic acid production rate and yield than B -control but more sugars remaining (Figure 4). The higher production rate in BY H0 and BY L0 could be due to D. bruxellensis being more competitive forcing K. intermedius to produce acids or perhaps the metabolites produced by D. bruxellensis could stimulate acid production, as its presence has been found to correlate with higher acid production in kombucha [6,39]. more glucose produced from sucrose breakdown [6,16]. However, samples with D. bruxellensis added sequentially had a lower gluconic acid production rate and yield than Bcontrol but more sugars remaining (Figure 4). The higher production rate in BYH0 and BYL0 could be due to D. bruxellensis being more competitive forcing K. intermedius to produce acids or perhaps the metabolites produced by D. bruxellensis could stimulate acid production, as its presence has been found to correlate with higher acid production in kombucha [6,39]. Previous studies on co-culture have shown that the presence of D. bruxellensis in general is able to promote glucuronic acid production by Gluconacetobacter intermedius [15]. The higher yield could either be due to D. bruxellensis providing more glucose from sucrose breakdown or by producing acetic acid, which can inhibit glycolysis in G. intermedius and therefore stimulate glucuronate synthesis [15]. Furthermore, the optimal ratio of D. bruxellensis to G. intermedius was 4:6, similar to the inoculum condition of BYL0 [15].  Previous studies on co-culture have shown that the presence of D. bruxellensis in general is able to promote glucuronic acid production by Gluconacetobacter intermedius [15]. The higher yield could either be due to D. bruxellensis providing more glucose from sucrose breakdown or by producing acetic acid, which can inhibit glycolysis in G. intermedius and therefore stimulate glucuronate synthesis [15]. Furthermore, the optimal ratio of D. bruxellensis to G. intermedius was 4:6, similar to the inoculum condition of BY L0 [15]. more glucose produced from sucrose breakdown [6,16]. However, samples with D. bruxellensis added sequentially had a lower gluconic acid production rate and yield than Bcontrol but more sugars remaining (Figure 4). The higher production rate in BYH0 and BYL0 could be due to D. bruxellensis being more competitive forcing K. intermedius to produce acids or perhaps the metabolites produced by D. bruxellensis could stimulate acid production, as its presence has been found to correlate with higher acid production in kombucha [6,39]. Previous studies on co-culture have shown that the presence of D. bruxellensis in general is able to promote glucuronic acid production by Gluconacetobacter intermedius [15]. The higher yield could either be due to D. bruxellensis providing more glucose from sucrose breakdown or by producing acetic acid, which can inhibit glycolysis in G. intermedius and therefore stimulate glucuronate synthesis [15]. Furthermore, the optimal ratio of D. bruxellensis to G. intermedius was 4:6, similar to the inoculum condition of BYL0 [15].

Effect of D. bruxellensis Concentration and Inoculation Time on Ethanol Production
Ethanol is known to improve BC synthesis as an additional energy source allowing glucose to be used mainly for BC production only [9]. Ethanol is also an alternative carbon source used by acetic acid bacteria [24]. In previous studies, D. bruxellensis isolated from kombucha has been shown to produce a high amount of ethanol [8]. Therefore, the ethanol production by D. bruxellensis during co-culture with K. intermedius was observed. The results confirm the previous studies as the ethanol production increased over time and peaked by day 14 (Figure 7). By day 14, Y -Control showed the highest final concentration with 6952 ± 923 mg/L (data not shown) followed by BY H0 with 1235 ± 256 mg/L, while BY L0 showed considerably lower concentration and produced only 105 ± 9 mg/L by the end of the incubation. Samples with D. bruxellensis added sequentially resulted in a lower final ethanol concentration, producing 19 ± 1.4 mg/L in BY H3 and a negligible amount in BY L3 , which corresponds to the yeast's undetectable growth (Figure 2). This finding indicates that inoculating D. bruxellensis simultaneously at 10 6 CFU/mL enables yeast to reach a high population more rapidly, which is crucial for increased ethanol production.

Effect of D. bruxellensis Concentration and Inoculation Time on Ethanol Production
Ethanol is known to improve BC synthesis as an additional energy source allowing glucose to be used mainly for BC production only [9]. Ethanol is also an alternative carbon source used by acetic acid bacteria [24]. In previous studies, D. bruxellensis isolated from kombucha has been shown to produce a high amount of ethanol [8]. Therefore, the ethanol production by D. bruxellensis during co-culture with K. intermedius was observed. The results confirm the previous studies as the ethanol production increased over time and peaked by day 14 (Figure 7). By day 14, Y-Control showed the highest final concentration with 6,952 ± 923 mg/L (data not shown) followed by BYH0 with 1,235 ± 256 mg/L, while BYL0 showed considerably lower concentration and produced only 105 ± 9 mg/L by the end of the incubation. Samples with D. bruxellensis added sequentially resulted in a lower final ethanol concentration, producing 19 ± 1.4 mg/L in BYH3 and a negligible amount in BYL3, which corresponds to the yeast's undetectable growth (Figure 2). This finding indicates that inoculating D. bruxellensis simultaneously at 10 6 CFU/mL enables yeast to reach a high population more rapidly, which is crucial for increased ethanol production.

Effect of D. bruxellensis Concentration and Inoculation Time on Free Amino Nitrogen (FAN) Concentration
The absence of a nitrogen source is known to result in 30% less BC yield, while an overabundance of nitrogen may instead only promote Komagataeibacter growth [40]. A nitrogen source is also crucial for D. bruxellensis growth, and the yeast has also been known to produce FAN, which is theorized to stimulate bacterial activity [6]. Therefore, the FAN concentration was measured in all samples during cultivation to investigate how the microbes interact (Figure 8). It was found that FAN in K. intermedius monoculture (B-control) only fluctuated slightly, with less than 5% being used, resulting in a final concentration of 96.78 ± 6.1 mg/L. This indicates that K. intermedius only needs a certain amount of nitrogen. Several studies have also reported that nitrogen consumption may vary depending on the Komagataeibacter strain and available sources [40,41]. A consistent decline was observed when a low concentration of D. bruxellensis was added on day 0 (BYL0), with only 16% of the FAN remaining by the end of the incubation (15.93 ± 2.6 mg/L). This could be because the FAN has been used for the yeast in BYL0 to grow rapidly (Figure 2). When a low concentration of D. bruxellensis was added on day 3 (BYL3), the FAN concentration followed a similar trend as B-control, with a slightly higher final concentration (115.6 ± 4.3 mg/L). The higher final concentration in BYL3 may be due to autolysis of D. bruxellensis as it could not compete with K. intermedius [6,42].
In both BYH0 and BYH3, an initial increase in FAN can be observed, with both samples peaking above 140 mg/L on day 3 and day 5, respectively. However, the FAN concentration then declined to approximately 75.24 ± 1.84 mg/L in BYH3 by day 14, while in BYH0, the FAN concentration decreased rapidly up to day 7 and then increased again toward Figure 7. Ethanol production by D. bruxellensis during 14 days of co-culture with K. intermedius at 30 • C. D. bruxellensis was added simultaneously at concentrations of 10 6 CFU/mL (BY H0 ) and 10 3 CFU/mL (BY L0 ) and sequentially at 10 6 CFU/mL (BY H3 ) and 10 3 CFU/mL (BY L3 ).

Effect of D. bruxellensis Concentration and Inoculation Time on Free Amino Nitrogen (FAN) Concentration
The absence of a nitrogen source is known to result in 30% less BC yield, while an overabundance of nitrogen may instead only promote Komagataeibacter growth [40]. A nitrogen source is also crucial for D. bruxellensis growth, and the yeast has also been known to produce FAN, which is theorized to stimulate bacterial activity [6]. Therefore, the FAN concentration was measured in all samples during cultivation to investigate how the microbes interact ( Figure 8). It was found that FAN in K. intermedius monoculture (B -control ) only fluctuated slightly, with less than 5% being used, resulting in a final concentration of 96.78 ± 6.1 mg/L. This indicates that K. intermedius only needs a certain amount of nitrogen. Several studies have also reported that nitrogen consumption may vary depending on the Komagataeibacter strain and available sources [40,41]. A consistent decline was observed when a low concentration of D. bruxellensis was added on day 0 (BY L0 ), with only 16% of the FAN remaining by the end of the incubation (15.93 ± 2.6 mg/L). This could be because the FAN has been used for the yeast in BY L0 to grow rapidly (Figure 2). When a low concentration of D. bruxellensis was added on day 3 (BY L3 ), the FAN concentration followed a similar trend as B -control , with a slightly higher final concentration (115.6 ± 4.3 mg/L). The higher final concentration in BY L3 may be due to autolysis of D. bruxellensis as it could not compete with K. intermedius [6,42].
In both BY H0 and BY H3 , an initial increase in FAN can be observed, with both samples peaking above 140 mg/L on day 3 and day 5, respectively. However, the FAN concentration then declined to approximately 75.24 ± 1.84 mg/L in BY H3 by day 14, while in BY H0 , the FAN concentration decreased rapidly up to day 7 and then increased again toward the end of incubation from 36.61 ± 1.42 mg/L to 107.41 ± 11.64 mg/L. These results are similar to the findings of Tran et al. [6], where an equal amount of Komagataeibacter saccharivorans and D. bruxellensis bacteria-yeast co-culture was found to have a higher final FAN concentration compared to pure bacterial culture. However, it should be noted that the ninhydrin method can be used to detect small proteins and ammonia as well [43]. Yeasts in kombucha are known to produce amino acids from the available nitrogen in tea. D. bruxellensis is also particularly known for its ability to produce γ-Aminobutyric acid (GABA) [44]. Finally, acetic acid bacteria are known to produce ammonia to survive acidic conditions [45]. Therefore, the final FAN concentration in this study may indicate not only amino acid consumption but also ammonia production, which could explain the pH increase toward the end of incubation (Figure 3). the end of incubation from 36.61 ± 1.42 mg/L to 107.41 ± 11.64 mg/L. These results are similar to the findings of Tran et al. [6], where an equal amount of Komagataeibacter saccharivorans and D. bruxellensis bacteria-yeast co-culture was found to have a higher final FAN concentration compared to pure bacterial culture. However, it should be noted that the ninhydrin method can be used to detect small proteins and ammonia as well [43]. Yeasts in kombucha are known to produce amino acids from the available nitrogen in tea. D. bruxellensis is also particularly known for its ability to produce γ-Aminobutyric acid (GABA) [44]. Finally, acetic acid bacteria are known to produce ammonia to survive acidic conditions [45]. Therefore, the final FAN concentration in this study may indicate not only amino acid consumption but also ammonia production, which could explain the pH increase toward the end of incubation (Figure 3).

Effect of D. bruxellensis Concentration and Inoculation Time on BC Production by K. intermedius
Varying yeast concentrations and inoculation time may affect how bacteria and yeast interact, eventually impacting BC production. This study found that adding D. bruxellensis at a high concentration (10 6 CFU/mL), regardless of inoculation time, could enhance BC yield by 74-102% (Figure 9). B-control could only produce 0.63 ± 0.11 g/L BC after 14 days, while the yield increased to 1.09 ± 0.02 g/L and 1.27 ± 0.33 g/L when a high concentration of D. bruxellensis was simultaneously (BYH0) and sequentially (BYH3)added, respectively. On the other hand, adding D. bruxellensis at a low concentration (10 3 CFU/mL), regardless of the inoculation time, decreased the BC yield less than that of the bacteria alone. The lowest BC yield was observed in the co-culture with a low concentration of D. bruxellensis added simultaneously (BYL0), which produced 0.44 ± 0.012 g/L, a 30% lower yield than Bcontrol, while sequential addition D. bruxellensis (BYL3) had 18% lower BC yield compared to B-control (0.52 ± 0.017 g/L). The BC yield is noticeably lower in this study compared to our previous study [10], potentially due to NaOH treatment for BC purification. Other studies have found that NaOH treatment of BC at high temperatures resulted in a lower BC mass due to interactions with the BC structure [46,47]. Additionally, the BC mass could further decrease due to NaOH treatment being more intensive to completely remove the dark colored sugarcane molasses-based media.
In this study, it was observed that a lower concentration of D. bruxellensis (BYL0) gave a lower BC yield and more gluconic acid (10.08 ± 0.12 g/L), while BYH3 produced less gluconic acid (4.64 ± 0.19 g/L) and gave a higher BC yield. According to a study by Gilbert et al. [48], adding a lower concentration of S. cerevisiae to Komagataeibacter rhaeticus could increase the BC yield due to less competition for glucose. Perhaps the sequential addition of D. bruxellensis (BYH3) was more similar to this condition since more glucose was available for the bacteria ( Figure 4A). In addition to being competitive for glucose, BYL0 also

Effect of D. bruxellensis Concentration and Inoculation Time on BC Production by K. intermedius
Varying yeast concentrations and inoculation time may affect how bacteria and yeast interact, eventually impacting BC production. This study found that adding D. bruxellensis at a high concentration (10 6 CFU/mL), regardless of inoculation time, could enhance BC yield by 74-102% (Figure 9). B -control could only produce 0.63 ± 0.11 g/L BC after 14 days, while the yield increased to 1.09 ± 0.02 g/L and 1.27 ± 0.33 g/L when a high concentration of D. bruxellensis was simultaneously (BY H0 ) and sequentially (BY H3 )added, respectively. On the other hand, adding D. bruxellensis at a low concentration (10 3 CFU/mL), regardless of the inoculation time, decreased the BC yield less than that of the bacteria alone. The lowest BC yield was observed in the co-culture with a low concentration of D. bruxellensis added simultaneously (BY L0 ), which produced 0.44 ± 0.012 g/L, a 30% lower yield than B -control, while sequential addition D. bruxellensis (BY L3 ) had 18% lower BC yield compared to B -control (0.52 ± 0.017 g/L). The BC yield is noticeably lower in this study compared to our previous study [10], potentially due to NaOH treatment for BC purification. Other studies have found that NaOH treatment of BC at high temperatures resulted in a lower BC mass due to interactions with the BC structure [46,47]. Additionally, the BC mass could further decrease due to NaOH treatment being more intensive to completely remove the dark colored sugarcane molasses-based media.
In this study, it was observed that a lower concentration of D. bruxellensis (BY L0 ) gave a lower BC yield and more gluconic acid (10.08 ± 0.12 g/L), while BY H3 produced less gluconic acid (4.64 ± 0.19 g/L) and gave a higher BC yield. According to a study by Gilbert et al. [48], adding a lower concentration of S. cerevisiae to Komagataeibacter rhaeticus could increase the BC yield due to less competition for glucose. Perhaps the sequential addition of D. bruxellensis (BY H3 ) was more similar to this condition since more glucose was available for the bacteria ( Figure 4A). In addition to being competitive for glucose, BY L0 also promoted gluconic acid production, which may have caused less BC to be produced [49]. On the other hand, the high BC yield observed in BY H0 could be due to the D. bruxellensis supporting the K. intermedius growth for a longer period (Figure 1), either by ethanol or FAN production (Figures 7 and 8) [6,48].
The significance of the D. bruxellensis concentration and inoculation time was calculated with the General Linear Model at a 95% confidence interval. The result shows that the D. bruxellensis concentration (p-value 0.01) and the time of inoculation (p-value 0.031) are significant on the final BC yield. However, the interrelation between D. bruxellensis ratio and the inoculation time (p-value 0.21) at a 5% confidence level was not significant on the final BC yield. Figure 9. Effect of D. bruxellensis concentration and inoculation time on BC final yield by produced by K. intermedius. D. bruxellensis was added simultaneously at concentrations of 10 6 CFU/mL (BYH0) and 10 3 CFU/mL (BYL0) and sequentially at 10 6 CFU/mL (BYH3) and 10 3 CFU/mL (BYL3). Means with different letters are significantly different (p < 0.05).

Conclusions
This study aimed to investigate the effect of D. bruxellensis concentrations (10 3 and 10 6 CFU/mL) and inoculation time (days 0 and 3) to promote BC synthesis by K. intermedius. The results show that regardless of inoculation time, a lower concentration of D. bruxellensis (10 3 CFU/mL) reduces the BC yield by 18-30% (0.44-0.52 g/L) compared to that of K. intermedius in monoculture (0.63 ± 0.11 g/L). On the other hand, adding a higher concentration of D. bruxellensis (10 6 CFU/mL), either simultaneously or sequentially, could increase the BC yields by 74-102% (1.09-1.27 g/L). Adding a higher D. bruxellensis concentration at the start of incubation could improve the survival of both species and stimulate metabolic reactions favorable for BC production, such as increased sugar consumption and higher ethanol and FAN production. In contrast, lower D. bruxellensis concentration promotes gluconic and glucuronic acid production, resulting in lower BC yields. These results show that it is feasible to regulate the BC production of K. intermedius in co-culture by controlling the inoculum proportion and time of addition of D. bruxellensis. Prior to scaling up the co-culture, more studies can be done on the metabolomics of the co-cultures in response to the varying ratios and inoculation time of D. bruxellensis. After identifying the role of each metabolite produced by D. bruxellensis, the metabolites responsible for stimulating BC production can be further optimized, while metabolites that stimulate acid production can be limited.
Author Contributions: Conceptualization, P.V.P.D., K.K. and S.A.; methodology, P.V.P.D., K.K. and S.A.; validation, P.V.P.D., K.K. and S.A.; formal analysis, F.P.; investigation, F.P.; resources, S.A.; data curation, P.V.P.D., K.K. and S.A.; writing-original draft preparation, P.V.P.D.; writing-review and editing, K.K., S.A. and M.J.T. ; visualization, P.V.P.D. and S.A.; supervision, P.V.P.D., K.K. and S.A.; project administration, P.V.P.D., K.K. and S.A.; funding acquisition, P.V.P.D., K.K. and S.A. All authors have read and agreed to the published version of the manuscript. The significance of the D. bruxellensis concentration and inoculation time was calculated with the General Linear Model at a 95% confidence interval. The result shows that the D. bruxellensis concentration (p-value 0.01) and the time of inoculation (p-value 0.031) are significant on the final BC yield. However, the interrelation between D. bruxellensis ratio and the inoculation time (p-value 0.21) at a 5% confidence level was not significant on the final BC yield.

Conclusions
This study aimed to investigate the effect of D. bruxellensis concentrations (10 3 and 10 6 CFU/mL) and inoculation time (days 0 and 3) to promote BC synthesis by K. intermedius. The results show that regardless of inoculation time, a lower concentration of D. bruxellensis (10 3 CFU/mL) reduces the BC yield by 18-30% (0.44-0.52 g/L) compared to that of K. intermedius in monoculture (0.63 ± 0.11 g/L). On the other hand, adding a higher concentration of D. bruxellensis (10 6 CFU/mL), either simultaneously or sequentially, could increase the BC yields by 74-102% (1.09-1.27 g/L). Adding a higher D. bruxellensis concentration at the start of incubation could improve the survival of both species and stimulate metabolic reactions favorable for BC production, such as increased sugar consumption and higher ethanol and FAN production. In contrast, lower D. bruxellensis concentration promotes gluconic and glucuronic acid production, resulting in lower BC yields. These results show that it is feasible to regulate the BC production of K. intermedius in co-culture by controlling the inoculum proportion and time of addition of D. bruxellensis. Prior to scaling up the co-culture, more studies can be done on the metabolomics of the co-cultures in response to the varying ratios and inoculation time of D. bruxellensis. After identifying the role of each metabolite produced by D. bruxellensis, the metabolites responsible for stimulating BC production can be further optimized, while metabolites that stimulate acid production can be limited.