foods

Article
Microbial Dynamics between Yeasts and Acetic Acid
Bacteria in Kombucha: Impacts on the Chemical
Composition of the Beverage
Thierry Tran 1, * , Cosette Grandvalet 1 , François Verdier 2 , Antoine Martin 2 , Hervé Alexandre 1
and Raphaëlle Tourdot-Maréchal 1
1 UMR Procédés Alimentaires et Microbiologiques, Université de Bourgogne Franche-Comté/AgroSup Dijon,
Equipe Vin Alimentation Micro-organismes Stress (VAlMiS) Institut Universitaire de la Vigne et du Vin Jules
Guyot, 2 rue Claude Ladrey, BP 27877, 21000/26, bd Docteur Petitjean, BP 87999, 21079 Dijon, France;
[email protected] (C.G.); [email protected] (H.A.); [email protected] (R.T.-M.)
2 Biomère, 14 rue Audubon, 75120 Paris, France; [email protected] (F.V.); [email protected] (A.M.)
* Correspondence: [email protected]




Received: 24 June 2020; Accepted: 20 July 2020; Published: 21 July 2020


Abstract: Kombucha is a traditional low-alcoholic beverage made from sugared tea and transformed by
a complex microbial consortium including yeasts and acetic acid bacteria (AAB). To study the microbial
interactions and their impact on the chemical composition of the beverage, an experimental design with
nine couples associating one yeast strain and one AAB strain isolated from original black tea kombucha
was set up. Three yeast strains belonging to the genera Brettanomyces, Hanseniaspora, and Saccharomyces
and three strains of Acetobacter and Komagataeibacter species were chosen. Monocultures in sugared
tea were analyzed to determine their individual microbial behaviors. Then, cultivation of the original
kombucha consortium and cocultures in sugared tea were compared to determine the interactive
microbial effects during successive phases in open and closed incubation conditions. The results
highlight the main impact of yeast metabolism on the product’s chemical composition and the
secondary impact of bacterial species on the composition in organic acids. The uncovered microbial
interactions can be explained by different strategies for the utilization of sucrose. Yeasts and AAB
unable to perform efficient sucrose hydrolysis rely on yeasts with high invertase activity to access
released monosaccharides. Moreover, the presence of AAB rerouted the metabolism of Saccharomyces
cerevisiae towards higher invertase and fermentative activities.

Keywords: kombucha; yeasts; acetic acid bacteria; interactions; symbiosis; sucrose hydrolysis; pellicle

1. Introduction
Fermented food is often designated as a convenient model for the study of microbial communities
and interactions [1–3]. Kombucha is an increasingly exploited example of a fermented beverage
obtained by microbial communities. Kombucha, also named “kombucha tea”, results from the metabolic
interplay of a microbial consortium including acetic acid bacteria (AAB), yeasts, and often (but not
always) lactic acid bacteria in sugared tea liquor [4–6]. Infusion provides nitrogenous substances
extracted from tea necessary for the growth of microorganisms. Sucrose is converted into glucose and
fructose by periplasmic yeast invertase and ethanol is produced as a result of alcoholic fermentation.
AAB oxidize glucose into gluconic acid and ethanol into acetic acid through oxidative metabolism.
This metabolic scheme raises the question of a possible trophic dependency of AAB towards yeasts [7].
Glucose and fructose are also used for bacterial cellulose production, leading to the formation of the
pellicle, also known as “mother”, “tea fungus”, or Symbiotic Culture of Yeast and Bacteria (“SCOBY”),
since it can be used as inoculum with or instead of liquid culture [4,5,8]. As a result, the beverage has

Foods 2020, 9, 963; doi:10.3390/foods9070963 www.mdpi.com/journal/foods

Foods 2020, 9, 963 2 of 25

the profile of a soda with a sweet/sour balance and can also be carbonated naturally if the vessel is
left closed after a first phase of biological acidification at ambient temperature [9]. There is no single
“culture” or microbial consortium for the production of kombucha, but a multitude of matrix-adapted
consortia whose origins are unknown. This matrix also offers two distinct environments: A liquid
phase, where microorganisms are in a planktonic state, and a pellicle, where they are entrapped.
Biofilms are known to host numerous interaction mechanisms among microbial communities and the
cellulosic pellicle produced by the AAB of kombucha is no exception [10–12]. Beyond this trophic
interaction, little is known about other types of interactions that might occur during kombucha
fermentation at genus or species levels. The formation of the pellicle, as well as the drop of pH, can
work as protection against contamination by exogenous spoilage agents such as molds [13]. Inside the
consortium, inter-kingdom and intra-kingdom interaction mechanisms remain poorly documented.
Some studies have provided clues on the existence of interactions through the lens of microbial
dynamics. The results have shown that some bacterial species are dominant, but the dynamics are
dependent on the type of tea used (black or green) [14]. It has also been shown that the population of
some yeast species decreases during kombucha fermentation, while others remain stable, as was the case
for Torulaspora delbrueckii or Schizosaccharomyces pombe in the presence of Zygosaccharomyces bailii. In this
case, the impacted yeast species were dependent on the consortium and this phenomenon did not occur
in pellicles, with all populations being maintained [15]. Microbial interaction data based on indigenous
complex consortia are difficult to summarize because of the complexity of the original system. Focus
has been placed on microbial dynamics in terms of populations, without investigating the impact on the
chemical composition of the beverage. To the best of our knowledge, only one study has investigated
the interactions between individual isolates from kombucha by cultivating AAB (Acetobacter sp.)
in sugared black tea with the addition of autoclaved sugared black tea fermented by a single strain
of yeast (Saccharomyces cerevisiae, Brettanomyces bruxellensis, and Zygosaccharomyces bailii) [16]. This
study concluded on the positive impact of the addition of fermented medium on AAB growth and
also suggested that acetic acid produced by AAB could stimulate the production of ethanol by yeast.
A recent study compared an original kombucha consortium with an unknown microbial composition,
and a synthetic consortium including purchased isolates of species commonly found in kombucha [17].
Two AAB species, consisting of Acetobacter pasteurianus and Komagataeibacter xylinus, were used with
the yeast species Z. bailii. The study confirmed the technical suitability of using a synthetic consortium
of yeasts and AAB for the fermentation of kombucha beverages and highlighted the metabolic interplay
between yeasts and AAB.
The present study goes further in the study of the yeast–bacterium interaction that occurs during
kombucha fermentation in terms of both microbiological levels and the composition of the main
metabolites produced. Yeasts and bacterial strains were isolated from an original black tea kombucha
culture, identified, and selected to develop an experimental design with three yeast strains and three
AAB strains. Each yeast and AAB was characterized individually, co-inoculated (one yeast strain x
one AAB strain, with “x” indicating a coculture) in sugared black tea, and compared to the original
consortium. This methodology was inspired by the integrated design for the study of microbial
communities described by Lawson et al. (2020), consisting of joint studies of the original consortium,
with simplified and deconstructed consortia reassembled using isolates [18]. Moreover, on top of the
aerobic acidification phase, the anaerobic carbonation phase in a closed bottle was investigated for the
first time.

2. Materials and Methods

A diagram summing up the experiment is available as Supplementary Data (Figure S1).

2.1. Isolation and Identification of Microbial Species

Yeast and bacterial strains were isolated from liquid and pellicle samples of black tea kombucha
from the company Biomère (Paris, France). WallersteinLab (WL) agar medium from Thermo Fisher
Foods 2020, 9, 963 3 of 25

Scientific (Waltham, MA, USA) was used to isolate yeasts [19]. Bromocresol Green allowed a macroscopic
discrimination of colonies based on their aspect and color for the set of yeasts isolated in the present
study [20,21]. De Man Rogosa and Sharpe (MRS) (pH 6.2) from Condalab (Madrid, Spain), LAC
(pH 5.1), and Mannitol agar media [22] were used for the isolation of bacteria that may have different
nutritional requirements in both aerobic and anaerobic conditions of incubation [23]. All reagents used
in agar media were purchased from Merck (Darmstadt, Germany), if not otherwise specified.
Five yeast colonies per colony morphotype were separately picked-up for inoculation in Yeast
Peptone Dextrose (YPD) liquid medium (48 h at 28 ◦ C). Twenty-five bacterial colonies per agar medium
were separately picked-up for inoculation in LAC medium (48 h at 28 ◦ C in the aerobic condition of
incubation), since all species identified on Mannitol agar medium were also identified on LAC agar
medium. After incubation, culture media were eliminated by centrifugation (14.500× g, 3 min at 15 ◦ C).
A half volume of appropriate liquid medium with a half volume of 40% (v/v) glycerol was then added
to the samples to make a stock kept at −20 ◦ C.
DNA extraction from each isolated strain was performed using an Instagen Matrix kit (Bio-Rad,
Hercules, CA USA). For yeasts, the 26S rDNA region ribosomal non transcribed spacer 2 (NTS 2) was
amplified using the following primers: NL1 (50 -GCATATCAATAAGCGGAGGAAAAG-30 ) and NL4
(50 -GTCCGTGTTTCAAGACGG-30 ) [24]. Two microliters of extracted DNA were added to 25 µL PCR
mix (1.8 mM MgCl2 , 0.25 mM dNTPs, 1.25 µM of each primer, and 0.025 U Taq polymerase) (Promega
Corp., Madison, WI, USA). A Biorad (Hercules, CA, USA) thermocycler was used as described
previously [25].
For bacteria, 16S ribosomal DNA was amplified using the following primers: E517F
(50 -GCCAGCAGCCGCGGTAA-30 ) and E106R (50 -CTCACGRCACGAGCTGACG-30 ). Amplification
was performed in the same conditions as for the yeasts, except for the annealing temperature, which
was changed to 58 ◦ C [26]. Both strands Sanger sequencing was performed on amplified DNA
by Genewiz® (Leipzig, Allemagne) using NL1 and NL4 primers for yeasts and E515F and E106R
primers for bacteria. The sequences obtained were then analyzed using the software Geneious R7
(version 7.1.5) and the Basic Local Alignment Search Tool (BLAST) available on NCBI’s website
(https://blast.ncbi.nlm.nih.gov/Blast.cgi), thus returning genus and species names.

2.2. Growth Conditions

In order to generate pure precultures, yeasts were restreaked on WL agar and bacteria were
restreaked on MRS agar from stocks kept at −20 ◦ C and incubated at 28 ◦ C. MRS agar was chosen
because it is a commonly used growth medium for exigent bacteria, and the growth of each isolated
AAB species on this medium was tested prior to the experiment.
Cultivation of the original kombucha consortium and cocultures was initiated in sugared tea.
To produce 1 L of sugared black tea, tap water was filtered through a Brita (Taunusstein, Allemagne)
charcoal cartridge and sterile filtered by a Steritop® cartridge by Merck (Darmstadt, Allemagne). Then,
200 mL of sterile tap water was boiled and directly taken off the heating source. One gram of black
tea (Pu’er Grade 1 TN4107 from “Les jardins de Gaïa”) (Wittisheim, France) was immediately added
and left for infusion for one hour at room temperature. Tea was then removed using a sieve and
transferred to a sterile vessel. A total of 800 mL of sterile tap water at 26 ◦ C was added. Then, 60 g
sucrose was added and dissolved completely. The mix was kept at 26 ◦ C in a closed Schott® flask for
one hour, prior to inoculation. Yeast and AAB strains were cultivated in YPD and MRS liquid medium,
respectively. The tubes were not fully closed, in order to allow gaseous exchange during the incubation
at 28 ◦ C in static conditions for 3 days. Before inoculation in sugared tea, cells from pure cultures were
washed with sugared black tea and then centrifugated (3.000× g; 10 min at 4 ◦ C). The populations
of pure cultures were determined using a BD Accuri C6 (Franklin Lakes, NJ, USA) flow cytometer
coupled with 0.1 µg mL−1 propidium iodide marking for cell mortality [27].
The choice of the target inoculation rate for individual species was based on the population
levels determined for original kombucha at the inoculation time. Consequently, sugared black tea
Foods 2020, 9, 963 4 of 25

was then inoculated using precultures with an initial population of yeast or AAB of 1.105 cells mL−1 .
For cocultures, the same process as that employed for pure cultures was used, with initial populations
of 1.105 cells mL−1 for both yeast and bacterium strains.
The production of kombucha was carried out using a kombucha mother culture maintained in
the lab from the original black tea kombucha sample from Biomère (Paris, France) by adding a half
volume of sugared black tea each month. Kombucha mother culture was added to sugared black tea at
the rate of 12% (v/v) and kept at 26 ◦ C for 14 days in a 500 mL Schott® bottle, in order to produce a
fresh kombucha inoculum. 12% of this inoculum was added to fresh sugared black tea to produce the
traditional kombucha used in this study.
Monocultures, cocultures, and original kombucha incubations were performed in 125 mL
Boston flasks with a Specific Interfacial Surface (SIS) of 0.01 cm−1 in a final volume of 123 mL [28].
For monocultures in an open incubation condition (IC), bottlenecks were loosely covered with
tin foil to allow gas exchange, whereas for closed IC, flasks were closed using a cap to reflect
bottling. Monocultures were characterized in both open and closed IC at 26 ◦ C after 7 and 14 days for
microbiological analysis and after 14 days for chemical analyses. Cocultures and kombucha incubations
were carried out in two phases. The first phase (P1) of fermentation was performed for 14 days at
26 ◦ C in the same open IC as monocultures. Then, half of the volume was withdrawn for analysis
(“P1” samples). The bottles were then sealed, in order to trigger natural carbonation. This phase (P2)
was performed at 26 ◦ C for 10 days. At the end, the samples were analyzed (“P2” samples). Open
and closed IC do not strictly reflect aerobic or anaerobic systems, but rather the production process of
kombucha making in industrial conditions. Dissolved oxygen concentrations were not measured in
samples. A description of the research samples and abbreviations is detailed as Supplementary Data
(Table S2).
The populations of yeasts and/or bacteria during fermentation were determined by plating
successive decimal dilutions of samples on WL agar for yeasts and MRS agar for bacteria, with technical
triplicates for each biological triplicate.

2.3. Chemical Analyses

For chemical analyses, samples were kept frozen at −20 ◦ C and centrifuged prior to chemical
analyses (3.000× g; 15 min, 10 ◦ C). pH values were measured with a Mettler Toledo Five Easy pH meter
coupled with an LE498 probe. The total acidity was determined by titration with 0.1 N NaOH and
0.2% phenolphthalein as a color indicator [29].
Acetic, lactic, malic, and succinic acid concentrations were determined by HPLC with a VWR
Hitachi (Tokyo, Japan) control unit, L-2350 oven, L-2200 autosampler injection device, and L-2130
pump. The column used was a Raptor ARC-18 2.7 µm 150 × 2.1 mm from RESTEK (Lisses, France) and
the detector was a UV Diode Array Detector VWR Hitachi L-2455 (Tokyo, Japan) at 210 nm. The mobile
phase was 97% 20 mM KH2 PO4 (pH 2.4) and 3% methanol. A flow gradient was applied at 35 ◦ C, with
the following sequence for a total duration time of 10 min: From 0.1 mL min−1 to 1.0 mL min−1 for
5 min, 1.0 mL min−1 for 3 min, and 0.1 mL min−1 for 2 min.
Sucrose, glucose, and fructose concentration determinations by HPLC involved the same
equipment, except for the column. A HyperREZ XP Carbohydrate Ca++ 8% column from
Thermofisher (Waltham, Etats-Unis) was used with a Spectrasystem RI-150 refractometer from
de JMBS (Mandelieu-Napoule, France). The flow rate was 0.6 mL min−1 with ultrapure water at 80 ◦ C.
Gluconic acid and ethanol concentrations were determined using enzymatic kits from Biosentec
(Auzeville-Tolosane, France).
The free amino nitrogen (FAN) concentration was determined according to the protocol from
MEBAK® (2013) [30], with the following adaptation. In total, 0.4 mL 20-fold diluted sample was added
to 0.2 mL reactive mix (0.71 M Na2 HPO4 , 0.44 M KH2 PO4 , 28 mM ninhydrin, and 17 mM fructose).
The mix was heated for 16 min at 100 ◦ C and left to cool down for 20 min at room temperature. Then,
0.8 mL dilution solution (40% (v/v) ethanol, 12 mM potassium iodide) was added before the absorbance
Foods 2020, 9, 963 5 of 25

reading at 570 nm using a UV-1800 spectrophotometer from Shimadzu (Kyoto, Japan). Distilled water
was used as the blank. A calibration curve was made using a glycine solution.
Analytical chemistry results were expressed as the balance, according to the following formula (1):

∆Concentration = ∆Endpoint concentration − ∆Initial concentration. (1)

Initial corresponds to day 0 and the “endpoint” corresponds to day 14 for monocultures and “P1”
or “P2” for cocultures.

2.4. Statistical Analyses

All samples were made in triplicate. Values were treated with ANOVA and in the case of significant
differences (p < 0.05), the Newman–Keuls pair test was applied. Principal Component Analysis (PCA)
coupled with Hierarchic Ascending Classification (HCA) was also performed for P1 and P2 values
separately. All statistical analyses were performed with R software (version 3.5.2.).

3. Results and Discussion

3.1. Isolation and Identification of Yeast and Bacterial Strains

Based on the colony morphology on WL agar, yeast species in the original black tea kombucha
sample were isolated and then identified by DNA sequencing of the 26S ribosomal region [24]. In the
liquid phase, Brettanomyces (Dekkera) bruxellensis (small white colonies), Hanseniaspora valbyensis (dark
green colonies), and Saccharomyces cerevisiae (large white colonies) were identified. In the biofilm,
the same species were isolated, in addition to Hanseniaspora opuntiae (bright green colonies), Pichia aff.
fermentans (white star-shaped colonies), and Galactomyces geotrichum (white filaments). Yeast colony
morphologies were all distinct for each species. On agar plates, all AAB colonies had a sticky and
translucid aspect. DNA sequencing of the 16S ribosomal region [26] allowed the identification of the
species Acetobacter indonesiensis, Acetobacter papayae, and Komagataeibacter saccharivorans, which were all
isolated in both liquid and biofilm phases. All species names were associated with E-values equal to 0.
No lactic acid bacteria were identified, despite the use of LAC medium in both aerobic and anaerobic
conditions of incubation. Except for Galactomyces geotrichum, the genera of isolated yeasts and bacteria
are commonly found in kombucha [4]. For this study, species found in the liquid phase were chosen
because they were present in both liquid and biofilm. Each of the selected strains was characterized in
sugared black tea.

3.2. Characterization of Pure Cultures in Sugared Black Tea

Figure 1 presents the population levels of each selected strain inoculated in sugared black tea
after inoculation on day 0 and after incubation for 7 days and 14 days at 26 ◦ C in open and closed
incubation conditions (IC).
Foods 2020, 9, 963 6 of 25
Foods 2020, 9, x FOR PEER REVIEW 6 of 25

Figure 1.
Figure Microbialpopulations
1.Microbial populationsduring
duringcultivation
cultivation in
insugared
sugared black
black tea
tea of
of aamonoculture
monoculture of
of yeast
yeast and
and
bacterial strains isolated from black tea kombucha determined by plate counting (CFU mL −1 ). Error
bacterial strains isolated from black tea kombucha determined by plate counting (CFU mL ). Error −1

bars correspond
bars correspond to to the
the confidence
confidence interval
interval with α=
with α 0.05 (n
= 0.05 = 3).
(n = 3). Cultures
Cultures were
were conducted
conducted inin open
open
vessel (left) or closed (right) conditions of incubation for 14
vessel (left) or closed (right) conditions of incubation for 14 days.days.

Yeast population levels increased from 1.0.1055 ± 5.1044 CFU mL−1 to values between 1.0.106 CFU
Yeast population levels increased from 1.0.10 ± 5.10 CFU mL −1 to values between 1.0.106 CFU
mL−1 and 6.0.106 CFU mL−1 , regardless of the open or closed IC after 7 days. These levels of population
mL−1 and 6.0.106 CFU mL−1, regardless of the open or closed IC after 7 days. These levels of population
were maintained 7 days later.
were maintained 7 days later.
Except for K. saccharivorans, AAB’s population levels increased from 1.0.105 5± 5.1044 CFU mL−1 to
Except for K. saccharivorans,
6
AAB’s population
−1 7
levels
−1
increased from 1.0.10 ± 5.10 CFU mL−1 to
values between 1.0.10 CFU mL and 1.0.10 CFU mL , regardless of the open or closed IC after
values between 1.0.106 CFU mL−1 and 1.0.107 CFU mL−1, regardless of the open or closed IC after 7
7 days. It can be supposed that, on top of assimilable nutrients, the medium initially possessed a
days. It can be supposed that, on top of assimilable nutrients, the medium initially possessed a
sufficient dissolved oxygen level to allow the growth of AAB in closed IC. These levels of population
sufficient dissolved oxygen level to allow the growth of AAB in closed IC. These levels of population
were maintained 7 days later. K. saccharivorans possessed a lower population at day 0 and the
were maintained 7 days later. K. saccharivorans possessed a lower population at day 0 and the
population only increased by less than 1 log. Discrepancies between the target inoculation rate and
population only increased by less than 1 log. Discrepancies between the target inoculation rate and
plate counting results at day 0 could be due to the viable but not culturable (VBNC) state, as reported
plate counting results at day 0 could be due to the viable but not culturable (VBNC) state, as reported
for Komagataeibacter xylinus [31,32].
for Komagataeibacter xylinus [31,32].
The average initial composition of sugared black tea was 66.6 ± 3.3 g L−1 total sugars (66.4 g L−1
The average−1initial composition of sugared black tea was 66.6 ± 3.3 g L−1 total sugars (66.4 g L−1
−1
sucrose, 0.2 g L fructose, and no glucose detected). The FAN concentration was 63 ± 4−1 µg L ,
sucrose, 0.2 g L fructose, and no glucose detected). The FAN concentration was 63−1± 4 µ g L , which
−1
which is very low compared to grape must, for example (between 50 and 150 mg L ) [33]. The initial
is very low compared to grape must, for example (between 50 and 150 mg L−1) [33]. The initial average
average pH value was 6.64 ± 0.47 units and average total acidity was lower than 1 meq L−1 . No organic
pH value was 6.64 ± 0.47 units and average total acidity was lower than 1 meq L . No organic acids
−1
acids nor ethanol were detected. The chemical composition variation of sugared black tea after the
nor ethanol were detected. The chemical composition variation of sugared black tea after the
incubation of each culture (day 14) is detailed in Table 1.
incubation of each culture (day 14) is detailed in Table 1.
For yeasts, the decrease of total sugars indicated a maximal sugar variation of −10.7 g L−1. No
groups could be formed among the different values, despite significant differences (p < 0.05). Sucrose
underwent variations from null to total disappearance (−68.3 g L−1), with significant differences
between the values (p < 0.05). The monosaccharide content varied significantly from 0.0 to 28.5 g L−1
for glucose (p < 0.05) and from 0.0 to 31.8 g L−1 for fructose (p < 0.05). Their production succeeded
sucrose hydrolysis. A significant production of ethanol was also observed, with variations between
0.1 and 3.2 from 0.0 to 28.5 g L−1 (p < 0.05), which is representative of the fermentation activity. The
results show very different sugar consumption behaviors across the three yeasts species and highlight
the link between invertase activity, respiration, and fermentation. B. bruxellensis is characterized by
high invertase activity and fermentative metabolism in both open and closed IC. For this strain,
ethanol production was significantly higher in open IC than closed IC (p < 0.05). The inhibition of
alcoholic fermentation under anaerobic conditions is called the Custer effect and was reported for B.
bruxellensis [34,35]. H. valbyensis is characterized by poor sucrose hydrolysis and fermentative
capacities, particularly in open IC. The very low consumption of sucrose was reported in the context
of cider production [36]. The same applied to S. cerevisiae in open IC, but the presence of invertase
activity is more evident and points to the induction of SUC genes in the presence of a low glucose
concentration [37,38]. This mechanism allows S. cerevisiae to progressively hydrolyze sucrose and
consume monosaccharides, without inducing catabolic repression by glucose. However, oxygen
Foods 2020, 9, 963 7 of 25

Table 1. Change in the chemical composition of sugared black tea by pure cultures of yeast and bacterial strains isolated from black tea kombucha after 14 days in
open and closed conditions of incubation.

Microorganism B. bruxellensis H. valbyensis S. cerevisiae A. indonesiensis A. papayae K. saccharivorans

Open Closed Open Closed Open Closed Open Closed Open Closed Open Closed
Incubation Condition
Vessel Vessel Vessel Vessel Vessel Vessel Vessel Vessel Vessel Vessel Vessel Vessel
∆Total sugars (g L−1 ) * −10.7 a −10.6 a −2.2 a −9.4 a 0.0 a −7.9 a −3.6 a −0.9 a −0.7 a 0.0 a −6.2 a −10.0 a
∆Sucrose (g L−1 ) −68.3 d −68.3 d −2.2 a −15.4 ab −36.1 c −68.3 d −6.9 ab −0.9 a −1.0 a 0.0 a −12.3 ab −19.6 b
∆Glucose (g L−1 ) 27.0 a 26.9 a 0.0 d 0.9 d 18.9 b 28.5 a 2.1 d nd nd nd 4.2 d 9.0 c
∆Fructose (g L−1 ) 29.8 a 30.8 a 0.0 d 5.0 c 17.2 b 31.8 a 1.1 cd nd 0.2 d 0.4 d 1.9 cd 0.6 d
∆Ethanol (g L−1 ) 3.2 a 2.5 b 0.1 d 1.5 c 0.4 d 2.4 b nd <0.1 nd nd nd <0.1
∆Free amino nitrogen (µg L−1 ) 37 a 52 a 7b 31 ab 3b 54 a 3b 30 ab 8b 29 ab 8b 28 ab
∆pH −2.13 cd −1.86 cd −1.07 b −2.26 cd −0.38 a −1.67 c −2.64 d −2.25 cd −2.05 cd −2.04 cd −2.61 d −2.50 d
∆Total acidity (meq L−1 ) 8.0 abc 6.3 abc 2.0 c 6.7 abc 1.0 c 4.7 bc 11.0 abc 19.3 a 7.0 abc 5.7 bc 19.3 a 16.7 ab
∆Acetic acid (g L−1 ) 0.38 a 0.25 a nd <0.1 <0.1 nd 0.25 a 0.49 a 0.28 a 0.38 a 0.40 a 0.56 a
∆Gluconic acid (g L−1 ) <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 0.91 a 0.11 b 0.56 ab 0.18 b 1.94 a 0.12 b
∆Lactic acid (g L−1 ) nd nd 0.46 a 0.20 abc 0.32 ab nd nd nd nd nd <0.1 nd
∆Succinic acid (g L−1 ) 1.27 a 1.19 a 0.11 c 0.76 b 0.34 c 1.27 a <0.1 <0.1 nd nd 0.14 c 0.16 c
∆Malic acid (g L−1 ) 0.92 a 1.4 a nd 1.00 a nd 1.24 a nd 0.22 b <0.1 0.18 b nd <0.1
*: ∆Concentration = ∆Endpoint concentration − ∆Initial concentration; nd = not detected; common letters imply non-significant differences between average values (ANOVA test with
α = 0.05 and n = 3).
Foods 2020, 9, 963 8 of 25

For yeasts, the decrease of total sugars indicated a maximal sugar variation of −10.7 g L−1 .
No groups could be formed among the different values, despite significant differences (p < 0.05).
Sucrose underwent variations from null to total disappearance (−68.3 g L−1 ), with significant differences
between the values (p < 0.05). The monosaccharide content varied significantly from 0.0 to 28.5 g L−1
for glucose (p < 0.05) and from 0.0 to 31.8 g L−1 for fructose (p < 0.05). Their production succeeded
sucrose hydrolysis. A significant production of ethanol was also observed, with variations between 0.1
and 3.2 from 0.0 to 28.5 g L−1 (p < 0.05), which is representative of the fermentation activity. The results
show very different sugar consumption behaviors across the three yeasts species and highlight the
link between invertase activity, respiration, and fermentation. B. bruxellensis is characterized by
high invertase activity and fermentative metabolism in both open and closed IC. For this strain,
ethanol production was significantly higher in open IC than closed IC (p < 0.05). The inhibition of
alcoholic fermentation under anaerobic conditions is called the Custer effect and was reported for
B. bruxellensis [34,35]. H. valbyensis is characterized by poor sucrose hydrolysis and fermentative
capacities, particularly in open IC. The very low consumption of sucrose was reported in the context
of cider production [36]. The same applied to S. cerevisiae in open IC, but the presence of invertase
activity is more evident and points to the induction of SUC genes in the presence of a low glucose
concentration [37,38]. This mechanism allows S. cerevisiae to progressively hydrolyze sucrose and
consume monosaccharides, without inducing catabolic repression by glucose. However, oxygen
limitation in closed IC induced fermentation and increased sugar consumption (Pasteur effect), as
reported by the data summarized in Marques et al. (2016) [39]. According to this review, the increase
in sugar consumption may be a consequence of the lower energetic yield of fermentative metabolism
compared to respiration. Then, it can be supposed that the low sucrose consumption observed in
S. cerevisiae monoculture in open IC is associated with respiratory metabolism, which could also
be the case for H. valbyensis. In addition, slightly lower glucose concentrations could be observed
for modalities associated with fermentative metabolism. Studies have reported preferential glucose
consumption during the alcoholic fermentation of grape must [40,41]. The free amino nitrogen (FAN)
concentration increased in all modalities between 3 and 54 µg L−1 and it also seems to be enhanced by
fermentative metabolism induced by oxygen limitations, although a significant increase between open
and closed IC was only observed for S. cerevisiae (p < 0.05).
The increase in total acidity (between +1 and +8 meq L−1 ) led to a decrease of the pH value,
ranging from −1.7 to −2.7 units. This increase in total acidity was significantly more intense for B.
bruxellensis compared to S. cerevisiae and the increase for S. cerevisiae was also more intense than for
H. valbyensis in open IC (p < 0.05). B. bruxellensis is characterized by the production of acetic acid.
Indeed, this species is strongly associated with the production of this organic acid [34]. For the whole
set of yeasts, the accumulation of succinic and malic acid and the absence of lactic acid production
seem to be linked to fermentative metabolism in these conditions.
Despite the initial absence of monosaccharides added, AAB were able to consume between 0.7
and 10 g L−1 total sugars. The presence of monosaccharides after 14 days points to the hydrolysis of
sucrose in the external medium. This phenomenon has been reported in several studies [17,42–44].
The total acidity increased between +5.7 and +19.3 meq L−1 and induced a drop of pH between −2.05
and −2.64 units, with significant differences between the values (p < 0.05). The production of gluconic
acid from the oxidation of released glucose between +0.11 and +1.94 g L−1 was significantly lower
with closed IC for all AAB species except A. papayae (p < 0.05) and could be explained by the limitation
of oxygen [7]. The production of acetic acid between +0.25 and +0.56 g L−1 without the initial presence
of ethanol can be explained through the glycolysis and pyruvate metabolism of AAB [44–46] and has
been reported in the study of Wang et al. (2020) [17]. Succinic acid production occurred between
+0.14 and +0.16 g L−1 for K. saccharivorans monocultures and malic acid production only occurred
for Acetobacter sp. cultures in closed IC. It is worth noting that no consistent biofilm was produced
after 14 days in all conditions; only floating cellulose fragments were visible. Overall, AAB could
Foods 2020, 9, 963 9 of 25

hydrolyze sucrose and consume monosaccharides. Oxygen limitations in closed IC impacted oxidative
metabolism converting glucose into gluconic acid, and thus a limited acidification of the medium.
The interpretation of coculture results in comparison with original kombucha fermentation on
the basis of monoculture characterization will help clarify the impact of each metabolic profile on the
yeast–AAB interactions.

3.3. Comparison of Yeast-Acetic Acid Bacteria Cocultures with Original Kombucha Fermentation

3.3.1. Microbial Dynamics

The experimental plan for yeast-AAB cocultures and their abbreviations are presented in Table 2.
These pairings can be seen as minimal kombucha consortia that allow the metabolic interplay necessary
for the fermentation of kombucha. This process occurred at 26 ◦ C in two phases: An initial 14-day
phase in open IC and a second phase of 10 days in closed IC.

Table 2. “Yeast x Acetic acid bacteria” couples used for cocultures.

Acetobacter Acetobacter Komagataeibacter

Cocultures
indonesiensis papayae saccharivorans
Brettanomyces bruxellensis BB x AI BB x AP BB x KS
Hanseniaspora valbyensis HV x AI HV x AP HV x KS
Saccharomyces cerevisiae SC x AI SC x AP SC x KS

All yeast populations in coculture (Figure 2a,c,e) at day 14 remained significantly lower (1 to 0.5
log) than those in yeast monocultures at day 14 in open IC (Figure 1). This could mean that the presence
of AAB lowered the population of yeasts, regardless of the species. Possible explanations could be
nutritional competition [1] or inhibition through the production of acetic acid (see Table 3) [47,48].
AAB populations increased significantly above 1.105 CFU mL−1 at day 7. Only cocultures BB x AP,
BB x KS, and HV x AI (Figure 2b,d,f) reached bacterial populations beyond 1.106 CFU mL−1 on day
7. Between day 14 and day 24, populations remained stable in the range between 1.105 and 1.106
CFU mL−1 . The strongest variations could be seen with the modalities SC x AP and SC x KS, with
populations below 1.105 CFU mL−1 . Populations of Acetobacter sp. in coculture at day 14 were lower
than those in monocultures in open IC at day 14 by around 1 log. On the contrary, populations of K.
saccharivorans were higher in cocultures by around 1.5 log.
 Foods 2020, 9, x FOR PEER REVIEW 10 of 25

than2020,
Foods those in
monocultures in open IC at day 14 by around 1 log. On the contrary, populations
9, 963 of25K.
10 of
saccharivorans were higher in cocultures by around 1.5 log.

.
Figure 2. Microbial populations of yeast and bacterial cocultures in sugared black tea determined by
plate counting (CFU mL −1 ). Error bars correspond to the confidence interval with α = 0.05 (n = 3). (a) and
Figure 2. Microbial populations of yeast and bacterial cocultures in sugared black tea determined by
(b) Yeast
plate and bacterial
counting (CFU mLpopulations, respectively,
−1). Error bars in cocultures
correspond to the confidence Brettanomyces
involvinginterval with α bruxellensis (BB).
= 0.05 (n = 3). (a)
(c)
andand(b)
(d) Yeast
Yeastandandbacterial populations,
bacterial respectively,
populations, in cocultures
respectively, Hanseniaspora
involvinginvolving
in cocultures valbyensis
Brettanomyces
(HV). (e) and(BB).
bruxellensis (f) Yeast and (d)
(c) and bacterial
Yeast populations,
and bacterialrespectively,
populations,inrespectively,
cocultures involving Saccharomyces
in cocultures involving
cerevisiae (SC). AI = Acetobacter indonesiensis, AP = Acetobacter papayae, and KS = Komagataeibacter
Hanseniaspora valbyensis (HV). (e) and (f) Yeast and bacterial populations, respectively, in cocultures
saccharivorans.
involving Saccharomyces cerevisiae (SC). AI = Acetobacter indonesiensis, AP = Acetobacter papayae, and KS
= Komagataeibacter saccharivorans.

During original kombucha fermentation, total yeasts and total bacterial populations at
inoculation were 3.5.105 ± 4.6.104 CFU mL−1 and 8.5.104 ± 1.3.104 CFU mL−1, respectively (Figure 3).
Yeast and bacterial populations increased to 3.2.106 ± 1.8.106 CFU mL−1 and 1.8.106 ± 3.2.105 CFU mL−1,
respectively, and stayed stable until day 24. The total yeasts population thus remained slightly higher
than the total bacteria population throughout the fermentation. This was not observed for cocultures.
Foods 2020, 9, 963 11 of 25

Table 3. Difference in the organic acid content of samples between day 0 (after inoculation) and the
end point (14 days for pure cultures, 14 days in an open vessel for P1, and 10 days in a closed vessel
following P1 for P2).

Fermentation ∆Acetic ∆Gluconic ∆Lactic ∆Succinic ∆Malic Acid

Coculture
Phase Acid (g L−1 ) Acid (g L−1 ) Acid (g L−1 ) Acid (g L−1 ) (g L−1 )
BB x AI 0.7 bcd 1.6 abc nd 0.3 abc nd
BB x AP 0.6 bcd 1.4 abcd nd 0.3 abc nd
BB x KS 2.7 bc 1.7 ab nd nd nd
P1
Open vessel HV x AI nd 0.1 ef 0.3 a 0.2 bc nd
(14 days)
HV x AP 0.2 cd 0.1 f 0.4 a 0.2 abc nd
HV x KS 0.4 c <0.05 0.2 bc 0.2 abc 0.3 a
SC x AI 1.0 bcd 1.2 abcde nd 0.3 abc nd
SC x AP 0.8 c 2.2 a nd abc nd
0.3
SC x KS 2.7 bcd 1.3 abcde nd 0.4 abc nd
Kombucha bcd ab 0.1 c abc nd
1.6 1.7 0.2
BB x AI 2.0 bcd 1.6 abc nd 0.6 abc nd
BB x AP 1.6 bcd 1.2 abcde nd 0.7 ab nd
P2 a a
BB x KS 12.7 2.0 nd nd nd
Closed
vessel HV x AI 0.7 cd 0.3 def 0.2 bc 0.7 ab 1.0 a
(10 days HV x AP 2.4 bcd 0.5 cdef 0.1 cd 0.7 ab nd
after P1)
HV x KS 2.9 bcd 0.4 def 0.3 ab 0.5 abc 0.2 a
SC x AI 2.2 bcd 2.3 a nd 0.8 a nd
a a
SC x AP 2.2 bcd 2.4 nd 0.8 nd
SC x KS 11.3 a 2.2 a nd nd nd
Kombucha 2.1 bcd 0.8 bcdef 0.1 cd 0.3 abc nd
ANOVA was performed with α = 0.05 and n = 3, and common letters imply non-significant differences
between means. BB = Brettanomyces bruxellensis, HV = Hanseniaspora valbyensis, HO = Hanseniaspora opuntiae,
SC = Saccharomyces cerevisiae, AI = Acetobacter indonesiensis, AP = Acetobacter papayae, KS = Komagataeibacter
saccharivorans, and “x” = a coculture.

During original kombucha fermentation, total yeasts and total bacterial populations at inoculation
were 3.5.105 ± 4.6.104 CFU mL−1 and 8.5.104 ± 1.3.104 CFU mL−1 , respectively (Figure 3). Yeast
and bacterial populations increased to 3.2.106 ± 1.8.106 CFU mL−1 and 1.8.106 ± 3.2.105 CFU mL−1 ,
respectively, and stayed stable until day 24. The total yeasts population thus remained slightly
higher than the total bacteria population throughout the fermentation. This was not observed for
cocultures. Both yeast and AAB populations were lower in cocultures than in original kombucha
fermentation (Figure 2). The general higher population in original kombucha could indicate different
microbial dynamics of the consortium compared to cocultures. The differences in the yeast/AAB
balance could be explained by the absence of consistent pellicle formation in cocultures, with a fragile
veil appearing instead. The AAB population present in the biofilm in original kombucha could then
be found in the liquid phase of cocultures instead [49,50]. The different yeast colony morphotypes
allowed a discrimination of populations by species. At the inoculation time, H. valbyensis was the
major yeast species (6.6.104 ± 5.1.103 CFU mL−1 ), followed by B. bruxellensis. (2.8.105 ± 4.3.104 CFU
mL−1 ). The S. cerevisiae population was clearly lower, with 1.3.103 ±1.7.103 CFU mL−1 (Figure 3). After
7 days, all populations increased beyond 5 log and B. bruxellensis became the major species, followed
by H. valbyensis and then S. cerevisiae. The same levels of population were found at day 14. It is worth
noting that these proportions differ in comparison to those of the inoculum (fermentation of 14 days)
instead. The AAB population present in the biofilm in original kombucha could then be found in the
liquid phase of cocultures instead [49,50]. The different yeast colony morphotypes allowed a
discrimination of populations by species. At the inoculation time, H. valbyensis was the major yeast
species (6.6.104 ± 5.1.103 CFU mL−1), followed by B. bruxellensis. (2.8.105 ± 4.3.104 CFU mL−1). The S.
cerevisiae population was clearly lower, with 1.3.103 ±1.7.103 CFU mL−1 (Figure 3). After 7 days, all
Foods 2020, 9, 963 12 of 25
populations increased beyond 5 log and B. bruxellensis became the major species, followed by H.
valbyensis and then S. cerevisiae. The same levels of population were found at day 14. It is worth noting
that these proportions
and consequently differ in
the sample comparison
at day to thoseThis
0 (inoculation). of the inoculum
could (fermentation
be explained of 14 days)
by differences and
in oxygen
consequently
access due to the thegreater
sampleSISatofday
the 0preculture
(inoculation). This could
compared be explained
to the SIS used during bythe
differences
experiment.in Change
oxygen
access due to the greater SIS of the preculture compared to the SIS used during
occurred at day 18 and day 24 after closing the bottle, with a parallel decrease of the H. valbyensis the experiment.
Change occurred
population (−2.7.10at5 CFU
day 18
mLand day
−1 ) at day2424after
and closing the of
an increase bottle, with a(+1.8.10
S. cerevisiae parallel5 CFU
decrease
mL−1of). the
TheH. B.
valbyensis
bruxellensispopulation
population(−2.7.10
remained
5 CFU mLbetween
stable
−1) at dayday
24 and anday
7 and increase of S. cerevisiae
24, which highlights(+1.8.10
5 CFU mLrole
a predominant
−1).

The
duringB. the
bruxellensis population
whole process. remained dynamics
The population stable between
inducedday 7 andclosing
by vessel day 24, which
were highlights in
not observable a
predominant
cocultures (Figure role 2).
during the whole process. The population dynamics induced by vessel closing
were not observable in cocultures (Figure 2).

Figure 3. Microbial populations during cultivation in sugared black tea of the black tea kombucha
Figure 3. Microbial populations during cultivation in sugared black tea of the black tea kombucha
consortium determined by plate counting (CFU mL−1 ). Error bars correspond to the confidence interval
consortium determined by plate counting (CFU mL−1). Error bars correspond to the confidence
with α = 0.05 (n = 3).
interval with α = 0.05 (n = 3).
3.3.2. Utilization of Carbohydrates
3.3.2. Utilization of Carbohydrates
No groups could be determined among the total sugar variation values, ranging between −5 and
No
−27 g L−1 groups could
, despite the be determined
presence among the
of significant total sugar
differences (p =variation values,
0.008) (Figure ranging
4a). between
However, −5 and
the average
−27 g L−1, despite the presence of significant differences (p = 0.008) (Figure 4a). However, the average
values of total sugar variations are all lower between the aerobic acidification phase (14 days after
values of total
inoculation, P1)sugar variations
and anaerobic are all lower
acidification between
phase (10 daystheinaerobic
closed IC,acidification phasethat
P2). This means (14 most
daysof after
the
inoculation, P1) and anaerobic acidification phase (10 days in closed IC, P2). This
sugars were consumed during P1. The sucrose hydrolysis of cocultures, reflected by the disappearance means that most of
the sugars were consumed during P1. The sucrose hydrolysis of cocultures,
of sucrose, was similar to the capacities observed by the corresponding yeast monoculture, except reflected by the
disappearance
for S. cerevisiae.ofThis sucrose,
specieswas similarcloser
behaved to the capacities observed
to monocultures by theIC,
in the closed corresponding
with the completeyeast
monoculture, except for S. cerevisiae. This species behaved closer to monocultures
degradation of sucrose during P1 (Figure 4a and Table 1). All cocultures except those including in the closed IC,
with the complete degradation of sucrose during P1 (Figure 4a and Table
H. valbyensis achieved significantly stronger sucrose hydrolysis at P1 compared to kombucha (p <1). All cocultures except
those including
0.05), but this gap H.was
valbyensis
filled atachieved
P2 (Figure significantly stronger asucrose
3a). This highlights hydrolysis
strong impact at P1 on
of yeasts compared
the releaseto
kombucha (p < 0.05), but this gap was filled at P2 (Figure 3a). This highlights a
of monosaccharides that are used by AAB. The results displayed in Figure 4-b show a significantly strong impact of yeasts
on the release
greater releaseofofmonosaccharides
monosaccharidesthat of B.are used by AAB.
bruxellensis and S. The results compared
cerevisiae displayed toin H.
Figure 4-b show
valbyensis, thusa
confirming the release of glucose and fructose for cocultures including yeasts with strong known
invertase activity (p < 0.05 for glucose and p < 0.05 for fructose). It is noteworthy that in all cases
at P1, the fructose content was higher than glucose at P1 and P2, which was characteristic of yeast
metabolism in monocultures (Table 1), but preference for glucose could also occur for AAB.
 of yeast metabolism in monocultures (Table 1), but preference for glucose could also occur for AAB.
Original kombucha underwent an incomplete hydrolysis of sucrose at P1 (−52.0 g L−1 sucrose),
but the process was completed at P2 (−63.4 g L−1 sucrose), although the total sugar consumption did
not evolve significantly (−15.9 and −18.2 g L−1 total sugars at P1 and P2, respectively; the initial amount
of total sugar was 72.5 g L−1 with 71 g L−1 sucrose and 1.5 g L−1 fructose, and glucose was not detected).
Foods 2020, 9, 963 13 of 25
It can be supposed that the conditions in original kombucha led to a different rate of substrate
consumption.

Figure 4. Difference in total sugars and sucrose (a) and in glucose and fructose (b) (g L−1 ) of samples
Figure 4. Difference in total sugars and sucrose (a) and in glucose and fructose (b) (g L−1) of samples
between day 0 (after inoculation) and the end point (14 days for pure cultures, 14 days in an open
between day 0 (after inoculation) and the end point (14 days for pure cultures, 14 days in an open
vessel for P1, and 10 days in a closed vessel following P1 for P2). ANOVA was performed with
vessel for P1, and 10 days in a closed vessel following P1 for P2). ANOVA was performed with α =
α = 0.05 and n = 3. Common letters imply non-significant differences between means. Colors reflect the
0.05 and n = 3. Common letters imply non-significant differences between means. Colors reflect the
yeast species in the coculture. BB = Brettanomyces bruxellensis (purple), HV = Hanseniaspora valbyensis
yeast species
(green), in the coculture.
SC = Saccharomyces BB = Brettanomyces
cerevisiae bruxellensis indonesiensis,
(blue), AI = Acetobacter (purple), HV AP= Hanseniaspora
= Acetobactervalbyensis
papayae,
(green), SC = Saccharomyces cerevisiae (blue), AI = Acetobacter
KS = Komagataeibacter saccharivorans, and “x” = a coculture. indonesiensis, AP = Acetobacter papayae, KS
= Komagataeibacter saccharivorans, and “x” = a coculture.
Original kombucha underwent an incomplete hydrolysis of sucrose at P1 (−52.0 g L−1 sucrose),
but the process was completed at P2 (−63.4 g L−1 sucrose), although the total sugar consumption
did not evolve significantly (−15.9 and −18.2 g L−1 total sugars at P1 and P2, respectively; the initial
amount of total sugar was 72.5 g L−1 with 71 g L−1 sucrose and 1.5 g L−1 fructose, and glucose was
not detected). It can be supposed that the conditions in original kombucha led to a different rate of
substrate consumption.

3.3.3. Variations in Ethanol Content

The ethanol increase was lower at P1 (from +0.3 to +1.2 g L−1 ) than P2 because of the inhibition of
oxidative metabolism with oxygen deprivation (from +0.5 to +4.3 g L−1 ) (Figure 5). The production of
Foods 2020, 9, x FOR PEER REVIEW 13 of 25

3.3.3. Variations
Foods 2020, 9, 963 in Ethanol Content 14 of 25
The ethanol increase was lower at P1 (from +0.3 to +1.2 g L ) than P2 because of the inhibition
−1

of oxidative metabolism with oxygen deprivation (from +0.5 to +4.3 g L−1) (Figure 5). The production
ethanol by cocultures involving S. cerevisiae at P1 was higher than those occurring in monocultures
of ethanol by cocultures involving S. cerevisiae at P1 was higher than those occurring in monocultures
in open IC (+0.1 g L−1 ) (Table 3). This shows that this species in cocultures switched to fermentative
in open IC (+0.1 g L−1) (Table 3). This shows that this species in cocultures switched to fermentative
metabolism, even in open IC. This switch also involved an increase in invertase activity due to the lower
metabolism, even in open IC. This switch also involved an increase in invertase activity due to the
energetic yield of glucose consumed through fermentation compared to respiration (Figure 3). Between
lower energetic yield of glucose consumed through fermentation compared to respiration (Figure 3).
P1 and P2, the ethanol content increased significantly for BB x AI, BB x AP, HV x AP, and all cocultures
Between P1 and P2, the ethanol content increased significantly for BB x AI, BB x AP, HV x AP, and
involving S. cerevisiae (p < 0.05). The ethanol production of original kombucha was +1.0 g L−1 at P1,
all cocultures involving S. cerevisiae (p < 0.05). The ethanol production of original kombucha was +1.0
with no significant variation at P2. Most ethanol production was significantly higher in cocultures,
g L−1 at P1, with no significant variation at P2. Most ethanol production was significantly higher in
except BB x KS, HV x AI, and HV x KS. This could suggest that biological ethanol oxidation is influenced
cocultures, except BB x KS, HV x AI, and HV x KS. This could suggest that biological ethanol
by microbial dynamics. The presence of the pellicle in original kombucha could play a specific role in
oxidation is influenced by microbial dynamics. The presence of the pellicle in original kombucha
the enhancement of ethanol oxidation in particular.
could play a specific role in the enhancement of ethanol oxidation in particular.

Figure5.
Figure Difference in ethanol
5. Difference ethanol (g(gLL−1
−1) of samples
samplesbetween
betweenday day00(after
(afterinoculation)
inoculation)and
andthethe
end point
end (14
point
days
(14 for pure
days cultures,
for pure 14 days
cultures, in an in
14 days open
an vessel for P1, for
open vessel andP1,
10 and
days10in days
a closed
in avessel forvessel
closed P2 following
for P2
P1). ANOVA
following P1).was performed
ANOVA with α = 0.05
was performed and
with = 3. and
α =n 0.05 Common letters imply
n = 3. Common non-significant
letters differences
imply non-significant
between means. Colors reflect the yeast species in the coculture. BB = Brettanomyces
differences between means. Colors reflect the yeast species in the coculture. BB = Brettanomyces bruxellensis (purple),
HV = Hanseniaspora
bruxellensis (purple),valbyensis (green), SC =
HV = Hanseniaspora Saccharomyces
valbyensis cerevisiae
(green), (blue), AI = Acetobacter
SC = Saccharomyces cerevisiae indonesiensis,
(blue), AI =
AP = Acetobacter
Acetobacter papayae,AP
indonesiensis, KS == Acetobacter
Komagataeibacter saccharivorans,
papayae, and “x” = asaccharivorans,
KS = Komagataeibacter coculture. and “x” = a
coculture.
3.3.4. Variations in the Free Amino Nitrogen Content
An increase of the FAN concentration between +19 and +80 µg L−1 was also observed in cocultures
compared to sugared tea (initial average concentration of 63 µg L−1 ) (Figure 6). The FAN increase was
significantly lower for HV x KS compared to BB x KS and SC x KS at P1 (p < 0.05). The FAN increase in
original kombucha was 12 µg L−1 at P1 and was not significantly different compared to P2. The fact
that FAN increases were similar between P1 and P2 suggests that oxygen deprivation could occur at P1
in cocultures because of the oxidative metabolism of AAB, as this behavior contrasts with that of yeast
monocultures (Table 1).
Foods 2020, 9, x FOR PEER REVIEW 15 of 25
Foods 2020, 9, 963 15 of 25

Figure 6. Difference in free amino nitrogen (µg L−1 ) of samples between day 0 (after inoculation)
Figure 6. Difference in free amino nitrogen (µ g L−1) of samples between day 0 (after inoculation) and
and the end point (14 days for pure cultures, 14 days in an open vessel for P1, and 10 days in a
the end point (14 days for pure cultures, 14 days in an open vessel for P1, and 10 days in a closed
closed vessel for P2 following P1). ANOVA was performed with α = 0.05 and n = 3. Common letters
vessel for P2 following P1). ANOVA was performed with α = 0.05 and n = 3. Common letters imply
imply non-significant differences between means. Colors reflect the yeast species in the coculture.
non-significant differences between means. Colors reflect the yeast species in the coculture. BB =
BB = Brettanomyces bruxellensis (purple), HV = Hanseniaspora valbyensis (green), SC = Saccharomyces
Brettanomyces bruxellensis (purple), HV = Hanseniaspora valbyensis (green), SC = Saccharomyces cerevisiae
cerevisiae (blue), AI = Acetobacter indonesiensis, AP = Acetobacter papayae, KS = Komagataeibacter
(blue), AI = Acetobacter indonesiensis, AP = Acetobacter papayae, KS = Komagataeibacter saccharivorans, and
saccharivorans, and “x” = a coculture.
“x” = a coculture.
3.3.5. Acidification by the Production of Organic Acids
3.3.5. Acidification by the Production of Organic Acids
The pH of all cocultures dropped between −2 and −4 units (initial average of 6.64) and the
The pH increased
total acidity from +4
of all cocultures to +70 meq
dropped betweenL−1 −2 and value
(initial −4 units < 1(initial
meq Laverage of 6.64)
−1 ) (Figure and the
7). After total
14 days,
acidity increased
cocultures fromH.
involving +4 valbyensis
to +70 meqdisplayed
L (initialavalue
−1
drop<of 1 meq L ) (Figure
pH that −1 7). After 14weaker
was significantly days, cocultures
than the
others (p < 0.05) and their average total acidity values were the lowest (p < 0.05). Between P1 and
involving H. valbyensis displayed a drop of pH that was significantly weaker than the others (p < 0.05)
P2,
a significative drop of pH could be observed for HV x AP and HV x KS only (p < 0.05). Moreover,
and their average total acidity values were the lowest (p < 0.05). Between P1 and P2, a significative
drop of pH
the total couldincreased
acidity be observed for HV xfor
significantly APBB and HV HV
x KS, x KSx only
AP, HV (p <x0.05). Moreover,
KS, and the total
all cocultures acidity
involving
increased significantly for BB x KS, HV x AP, HV x KS, and all cocultures involving
S. cerevisiae, which suggests that the oxidative metabolism of AAB was maintained after the closing S. cerevisiae, which of
suggests
the vessel.that
Thisthe oxidative
could metabolism
have been allowedof byAAB was maintained
the presence of residual after the closing
oxygen after theofclosing
the vessel. This
of bottles.
couldThehave been
drop of allowed by the kombucha
pH of original presence ofremained
residual oxygen
the same after
at P1 theand
closing of bottles.
P2, with −1.7 units (average
initial pH of 6.04). The increase of total acidity was +47 meq L at P1 and +63.7 −1.7
The drop of pH of original kombucha remained the same at
−1 P1 and P2, with mequnits
L−1 at (average
P2, but
initial pH of 6.04).
these average valuesThewereincrease of total acidity
not significantly was(average
different +47 meqinitial
L attotal
−1 P1 and
acidity+63.7
of 3meq L atL−1
± 1 meq −1 P2,
). but
It is
these average values were not significantly different (average initial total acidity
noticeable that despite having one of the highest values for the total acidity, the pH decrease of original of 3 ± 1 meq L −1). It

is noticeablewas
kombucha thatless
despite
intense having onecocultures
than all of the highest values fora the
and supposes total acidity,
stronger buffering thecapacity.
pH decrease Whileofa
original
consistentkombucha
cellulosicwas less was
biofilm intense than in
present alloriginal
cocultures and supposes
kombucha, a stronger
no consistent buffering
pellicle capacity.
was visible for
While a consistent cellulosic biofilm was present in original kombucha, no
cocultures. Instead, cellulosic fragments were floating at the surface. Therefore, in the absence of consistent pellicle was
visible for cocultures.
a consistent Instead, cellulosic
pellicle, cocultures were ablefragments
to producewere floating
organic acidsatinthe surface.
similar Therefore,
quantities in the
to original
absence
kombucha, of awhich
consistent pellicle,
implies cocultures
that the biofilm were
is notable to produce
necessary organicthe
to complete acids
P1 in similar quantities to
phase.
original kombucha, which implies that the biofilm is not necessary to complete the P1 phase.
Foods 2020, 9, 963 16 of 25
Foods 2020, 9, x FOR PEER REVIEW 16 of 25

7. Difference in the
Figure 7. the (a)
(a) pH
pH value
value (arbitrary
(arbitrary unit)
unit) and
and (b)
(b)total
totalacidity (meqLL−1
acidity(meq −1 samples
) of samples
between day 0 (after inoculation) and the end point (14 days for pure cultures, 14 days in an open open
vessel for P1,
P1, and
and 1010days
daysininaaclosed
closedvessel
vesselfollowing
following P1P1forfor
P2). ANOVA
P2). ANOVA waswas
performed
performedwithwith
α=
α = 0.05
0.05 n = n3.=Common
andand 3. Common letters imply
letters non-significant
imply non-significantdifferences
differencesbetween
betweenmeans.
means. Colors
Colors reflect
reflect the
yeast species
species inin the
the coculture.
coculture. BBBB == Brettanomyces bruxellensis
bruxellensis (purple), HV == Hanseniaspora valbyensis
(purple), HV
(green), SC ==Saccharomyces
Saccharomycescerevisiae
cerevisiae(blue), AI= =
(blue),AI Acetobacter
Acetobacter indonesiensis,
indonesiensis, APAP = Acetobacter
= Acetobacter papayae,
papayae, KS
KS = Komagataeibacter saccharivorans, and “x” = a
= Komagataeibacter saccharivorans, and “x” = a coculture. coculture.

The production of
The production of organic
organicacids
acidsinincocultures
cocultures and
and original
original kombucha
kombucha is detailed
is detailed in Table
in Table 3.
3. No
No target
target organic
organic acidacid could
could be be detected
detected in in sugaredblack
sugared blacktea
teaand
andamong
amongthem,
them, no
no citric
citric acid
acid was
was
detected in any sample.
detected in any sample.
The main organic acids produced during P1 were acetic and gluconic acids, with the concentration
increase ranging below +2.7 g L−1 and below +2.2 g L−1 , respectively. The lowest increases were
associated with cocultures involving H. valbyensis. These organic acids are mainly produced by AAB
as a result of oxidative metabolism from ethanol and glucose. At P1, the gluconic acid concentrations
of cocultures involving H. valbyensis were significantly lower than those of cocultures involving
B. bruxellensis (p < 0.05). The results suggest that the production of organic acids reflects the capacity
of the paired yeast to make these substrates available, as it has been established that H. valbyensis
Foods 2020, 9, 963 17 of 25

possesses poor fermentative and invertase activities. Between P1 and P2, no significant change in
the concentration of acetic and gluconic acids occurred for most cocultures, except BB x KS and SC x
KS, whose acetic acid concentrations increased from +7.3 to +12.7 g L−1 (p < 0.05). This indicates a
capacity of K. saccharivorans to maintain the production of acetic acid in conditions of oxygen limitations
compared to Acetobacter sp., possibly through a pathway other than oxidative metabolism [44,45].
Other organic acids, such as succinic acid, were detected in all cocultures except for BB x KS at P1 and
BB x KS and SC x KS at P2, which suggests a link with the presence of K. saccharivorans, especially in
oxygen limitation conditions. Malic acid was only detected in HV x KS at P1 and HV x AI and HV
x KS at P2. Lactic acid was only detected in all cocultures involving H. valbyensis, which seems to
characterize this species (Table 1). The links established between the production of malic, succinic,
and lactic acids and the type of metabolism do not appear clearly for cocultures, probably due to the
complexification of metabolic pathways in cocultures.
In original kombucha, an increase of acetic acid and gluconic acid production of +1.6 and +1.7 g L−1 ,
respectively (initial average concentrations of acetic acid of 0.4 g L−1 and gluconic acid of 0.3 g L−1 )
and low production of lactic and succinic acids inferior to +0.2 g L−1 at P1 (lactic and succinic acids not
detected at day 0) (Table 3) were observed. Qualitatively, the organic acid composition of kombucha
gathers all organic acids detected in cocultures, with the exception of malic acid. On the contrary, lactic
acid was detected and was characteristic of cocultures with H. valbyensis. Quantitatively, the original
kombucha composition in acetic and gluconic matched those of all cocultures involving fermentative
yeasts (excluding H. valbyensis) at P1, but at P2, this only applied to cocultures involving Acetobacter sp.,
since the presence of K. saccharivorans induced an intense production of acetic acid that was probably
due to the intensification of ethanol through alcoholic fermentation induced by oxygen deprivation.
Indeed, this phenomenon did not occur in cocultures including H. valbyensis.
The coculture results underline the impact of yeasts and AAB metabolism on the matrix and
variation occurring between yeast and AAB species. The use of statistical treatment can help in
visualizing the phenomenon occurring in such complex systems and determine the similarities with
original kombucha.

3.3.6. Principle Component Analysis as a Visual Tool for Understanding Complex

Microbial Interactions
Principle Component Analysis (PCA) was performed on cocultures and original kombucha for P1
and P2 separately, in order to characterize the two different phases in terms of the impact on the chemical
composition of the sugared tea matrix (microbial population data were excluded). Both PCA possessed
a satisfactory sum of proper values with Dim1 and Dim2 axes (77.40% for P1 and 67.91% for P2).
The parameter plot of P1 (Figure 8a) was primarily structured by the Dim1 axis positively
characterized by the glucose, fructose, acetic acid, gluconic acid, ethanol, FAN concentrations, and total
acidity (cos2 ≥ 0.7), and negatively by the pH, and sucrose and lactic acid concentrations (cos2 ≤ −0.7).
Dim1 translated the biological acidification process targeted during P1 in an open vessel. This
acidification (increase of total acidity and decrease of pH) was mainly caused by the production of
acetic and gluconic acids by AAB. Their production was dependent on the release of monosaccharides
(glucose and fructose) from sucrose that can efficiently be hydrolyzed by yeast invertase activity.
Lactic acid production was associated with low invertase and fermentative activities, as observed in
H. valbyensis or S. cerevisiae monocultures in open IC and in cocultures with H. valbyensis (Tables 1 and 3).
Dim 2 was positively characterized by total sugars concentrations and negatively characterized by malic
acid, and translated differences in the profiles of organic acids from yeasts and the sugar consumption
of cocultures. As a result, the sample plot (Figure 8b) coupled with hierarchical clustering discriminated
cocultures with different levels. The predominant difference separated cocultures including H. valbyensis
from other yeasts. This was caused by the low sucrose hydrolysis and fermentative capacity of
H. valbyensis that prevented the access of AAB to monosaccharides and ethanol and thus efficient
acidification of the medium. Secondarily, the clustering discriminated the cocultures, including
 negatively characterized by malic acid, and translated differences in the profiles of organic acids from
yeasts and the sugar consumption of cocultures. As a result, the sample plot (Figure 8b) coupled with
hierarchical clustering discriminated cocultures with different levels. The predominant difference
separated cocultures including H. valbyensis from other yeasts. This was caused by the low sucrose
hydrolysis
Foods and fermentative capacity of H. valbyensis that prevented the access of AAB
2020, 9, 963 to
18 of 25
monosaccharides and ethanol and thus efficient acidification of the medium. Secondarily, the
clustering discriminated the cocultures, including K. saccharivorans and Acetobacter sp., for all coupled
K. saccharivorans and Acetobacter sp., for all coupled yeasts except S. cerevisiae. Original kombucha
yeasts except S. cerevisiae. Original kombucha belonged to the cluster including BB x KS, highlighting
belonged to the cluster including BB x KS, highlighting the closeness in chemical compositions.
the closeness in chemical compositions.

Figure8.8.Principle
Figure Component
Principle Component Analysis of coculture
Analysis and original
of coculture and kombucha samples using
original kombucha parameters
samples using
of the chemical composition. (a) Parameter or vector plot for P1, (b) sample plot for P1, (c)
parameters of the chemical composition. (a) Parameter or vector plot for P1, (b) sample plot for P1,parameter or
vector plot for P2, and (d) sample plot for P2. Continuous circles gather samples of the same
(c) parameter or vector plot for P2, and (d) sample plot for P2. Continuous circles gather samples of primary
clusters
the sameand dashed
primary line circles
clusters gather line
and dashed samples
circlesofgather
the same sub-clusters
samples according
of the same to hierarchical
sub-clusters according
ascendant clustering analysis.
to hierarchical ascendant clustering analysis.

The parameter plot of P2 (Figure 8c) was mainly structured by Dim1, which was positively
The parameter plot of P2 (Figure 8c) was mainly structured by Dim1, which was positively
characterized by glucose, fructose, and gluconic acid concentrations (cos2 ≥ 0.7), and negatively
characterized by glucose, fructose, and gluconic acid concentrations (cos² ≥ 0.7), and negatively
characterized by the pH, and lactic acid and sucrose concentrations. Dim2 was positively characterized
characterized by the pH, and lactic acid and sucrose concentrations. Dim2 was positively
by the ethanol concentration and negatively characterized by the total acidity (cos2 ≤ −0.7). Specifically,
characterized by the ethanol concentration and negatively characterized by the total acidity (cos² ≤
Dim1 could be representative of the invertase activity and Dim2 of AAB activity (oxidation of ethanol
−0.7). Specifically, Dim1 could be representative of the invertase activity and Dim2 of AAB activity
and increase of the total acidity). It can be noticed that total sugar and ethanol are not anticorrelated,
since, during P1 and even P2, the balance between total sugars and ethanol is biased by its conversion
in acetic acid by AAB. The clustering (Figure 8d) again discriminated cocultures including H. valbyensis
from the others and cocultures including K. saccharivorans from those with Acetobacter sp. (except for
cocultures including H. valbyensis). Moreover, original kombucha again belonged to the cluster
including BB x KS and SC x KS.
Foods 2020, 9, 963 19 of 25

The comparison of parameter plots of P1 and P2 reflects the shift from the combined yeasts and
AAB metabolisms, with a positive correlation of invertase activity, fermentation, and acidification
during the acidification phase (all strongly contributing to Dim1) (Figure 8a), towards the prevalence
of yeast metabolism (strong contribution to Dim1), with a lower correlation with AAB oxidative
metabolism during the carbonation phase (ethanol and total acidity strongly contributing to Dim 2)
(Figure 8c).
These results show that yeast metabolism was key throughout the process of fermentation:
P1 required yeasts to hydrolyze sucrose and produce ethanol for their conversion into acetic and
gluconic acids by AAB and P2 relied on fermentation to complete natural carbonation. Nevertheless,
K. saccharivorans still strongly impacted the chemical composition during P2 because of its ability to
maintain oxidative metabolism, especially when ethanol production is abundant, which was not as
visible for Acetobacter sp. More generally, yeast metabolism is the main factor influencing the kombucha
composition and the influence of AAB is secondary and mostly impacts the organic acid profile.

3.4. Sucrose Utilization Strategies as a Basis of Microbial Interactions in Kombucha

The experimental results obtained in this study confirm the metabolic interplay between yeasts and
acetic acid bacteria that has been theorized for kombucha made from sugared black tea [17,49,51,52].
They highlight, for the first time, the key role of yeasts and their metabolisms in terms of invertase activity
and the fermentation capacity, which can differ, depending on yeast strains (Table 1). Bibliographic
resources and the NCBI database have reported the existence of genes encoding invertase in S. cerevisiae
(internal and extracellular enzyme) [53–56] and B. bruxellensis (internal and extracellular enzyme) [57–
60], but few data is available for H. valbyensis. However, genomic studies on the Hanseniaspora genus
reported the loss of the SUC gene of the branch to which H. valbyensis belongs, thus supporting
the low sucrose hydrolysis observed (Table 1) [61]. On top of glucose, fructose, and ethanol, yeasts
may also release nitrogenous substrates through the conversion of bound amino-acids, such as
proteins and peptides, into free amino acids, since the results suggest a correlation between FAN
and fermentation (Figure 8a). Yeast can release free amino acid through extracellular proteolytic
activity and/or autolysis [62–65], which could benefit AAB as well [66]. As a result, the metabolism of
yeasts directly impacts the flux of available substrates for AAB. Nevertheless, AAB studied in this
work were able to grow in sugared black tea without the presence of yeasts to hydrolyze sucrose,
similar to the AAB strains used in Wang et al. (2020) [17]. From the perspective of the present
results and the experimental conditions of the work cited before, the hypothesis of a snowball effect is
supported: Initial spontaneous hydrolysis of sucrose in acidic conditions would be amplified by a pH
drop resulting from organic acid production by AAB (Figure S3) [67]. It could even be hypothesized
that such a phenomenon also applied for cultures involving H. valbyensis (Table 1, Figures 4 and 7).
However, the existence of enzymatic invertase activity in AAB is not to be excluded and more research
needs to be carried out in terms of the metabolic pathways of AAB [68]. However, Balasubramaniam
and Kannangara (1982) reported low invertase activity of K. xylinus and determined that sucrose
consumption was achieved through the activity of sucrose phosphorylase [69]. In all cases, sucrose
hydrolysis performed by yeasts except H. valbyensis was more efficient than that induced by AAB alone
(Table 1). It should be underlined that besides oxidative metabolism, AAB can use monosaccharides
through glycolysis, tricarboxylic acids (TCA), and pentose phosphate pathways [46]. The presence
of acetic, succinic, and malic acids in AAB monocultures are markers of these metabolisms (Table 1),
in agreement with coculture results. However, the metabolic activity of AAB mostly depends on the
substrates released by yeast and the composition in AAB can impact the profile of produced organic
acids (Table 3 and Figure 8).
The microbial dynamics of original kombucha after bottling (P2) were investigated for the first time
and highlighted the microbial interaction resulting in the decrease in the population of H. valbyensis
triggered by oxygen limitations (Figure 3). As S. cerevisiae exhibited more metabolic activity in
monocultures and cocultures in oxygen-limited conditions, it can be hypothesized that a negative
Foods 2020, 9, 963 20 of 25

interaction with H. valbyensis could occur in original kombucha. Saccharomyces–non-Saccharomyces

interactions have been studied and are not fully understood, but possible mechanisms include the
production of toxic compounds [70–73], competition with other nutrients such as sugars or amino
acids [74], cell–cell contact mechanisms [75], and flocculation [76]. It was observed that S. cerevisiae’s
metabolism in cocultures switched to high invertase activity and fermentation during P1 (Figures 4
and 5). This difference of behavior with the monocultures (Table 1) could be explained by an evolutionary
mechanism in S. cerevisiae, often compared to “The Prisoner’s Dilemma”, which modulates its invertase
activity according to the presence of “cheaters” without such enzymatic activity [13,77,78]. Therefore,
the SUC gene coding for the invertase is kept at a minimum expression in monocultures, so that
S. cerevisiae can hydrolyze sucrose progressively. A shift to a higher expression level can occur in
the presence of other “cheater” species, leading S. cerevisiae to rapidly consume sugars to outgrow
other populations and inhibit them through the release of toxic metabolites, such as ethanol [39].
However, in the case of kombucha, this benefits AAB because of the increase of oxidizable substrates
and possibly through the increased availability of amino acids. Analytical data regarding oxygen
consumption during kombucha fermentation is needed to assess its role in microbial interactions,
namely taking into account the presence of the pellicle, which is thought to reduce the access to oxygen
in the liquid phase [50,79]. It is worth noting that the pellicle was not necessary for AAB in cocultures
to perform efficient organic acid production, thus raising questions about additional functions other
than facilitating oxygen access to AAB during kombucha fermentation [50].
Differences between cocultures and kombucha also appeared in the chemical composition of the
liquid phase. Some kinetics in original kombucha seemed to be intermediate compared to cocultures,
especially with sucrose hydrolysis at P1 (Figure 3), the ethanol concentration at P2 (Figure 5), and the
FAN concentration (Figure 6). This could be explained by the balance in the yeast population, namely
between B. bruxellensis and H. valbyensis, which could lead to a leveling of global yeast metabolism
during original kombucha fermentation (Figure 3). The role of H. valbyensis remains enigmatic
regarding the technological aspect of kombucha fermentation. As the present study did not explore
volatile compounds, it cannot be excluded that this yeast strain could contribute to the olfactive profile
of kombucha beverages. More subtle microbial interactions could also take place in terms of amino
acid consumption or chemical signaling between H. valbyensis and other yeasts and bacteria [74].
Overall, the results point to an eco-evolution in kombucha between species that possess and do not
possess SUC genes necessary for the use of sucrose as a substrate. H. valbyensis and AAB would
then be labeled as “cheaters” by taking benefits from the invertase activity of other yeasts, such as B.
bruxellensis and S. cerevisiae, without contributing themselves to the production of public goods (here,
monosaccharides) [80,81]. However, it has been reported in S. cerevisiae x Escherichia coli cocultures that,
by lowering the population of public good producers (S. cerevisiae) at high population levels, cheaters
(E. coli) could consequently stabilize the sucrose hydrolysis rate, thus maintaining both population
levels [78]. A similar phenomenon could also have taken place in the present study, since lower yeast
populations in cocultures with AAB compared to monocultures were also observed (Figures 1 and 2).
H. valbyensis and AAB could therefore contribute to the stability of the community, as it was reported
that Hanseniaspora sp. could achieve fast growth due to the evolutionary loss of cell-cycle components,
thus limiting nutrient access [82], and acidification of the medium by AAB induced the selection of
acidophilic species [83].

4. Conclusions
Microbial interactions occurring in kombucha have often been described as “symbiotic”, which
does not imply mutual benefits, as described by the word “mutualistic”, but rather that all species
manage a stable coexistence, possibly through interactions that are yet to be defined [1,13]. Although
the present study lacks an exhaustive insight into the volatile and non-volatile metabolites, it was
possible to characterize nutritional interactions. As AAB are not dependent on yeasts to access
substrates, it appears that yeast–AAB interactions display a non-strict parasitism relationship (as
Foods 2020, 9, 963 21 of 25

sucrose hydrolysis occurred in AAB monocultures and was sufficient to increase their population at the
same level as in original kombucha). No evidence of a beneficial nutritional interaction for yeast could
be determined. There is also evidence that AAB induce the production of substrates by S. cerevisiae by
indirectly rerouting its metabolism, thus enhancing the benefits of this parasitic relationship. Moreover,
sucrose appears to play a central role in kombucha symbiosis not only in yeast–bacteria interactions,
but also potentially between yeast species. While the AAB composition secondarily impacts the profile
in organic acids, the yeast composition in kombucha consortia is crucial for the global microbial
dynamics, the chemical composition of the beverage, and the sensory characteristics of the product [9].
This study only includes one kombucha consortium, but many different microbial compositions
exist. Nevertheless, many research works have focused on the identification of yeast and bacteria
genera and it could be interesting to determine if the yeast profiles include genera or species with
variable invertase activities and fermentative capacities. In silico studies relying on genome and
protein databases combined with the results of past identification works of kombucha consortia could
highlight the existence of patterns in yeast compositions in terms of links between species and activities.
Yeast–yeast interactions in the context of kombucha fermentation should also be addressed in future
research works, now that the importance of their role has been uncovered.

Supplementary Materials: The following are available online at http://www.mdpi.com/2304-8158/9/7/963/s1,

Figure S1: Sum-up diagram of the experiment, Table S2: Description of research samples with abbreviations,
Figure S3: Hypothetical “snowball effect” explaining sucrose hydrolysis by acetic acid bacteria.
Author Contributions: Conceptualization, validation, methodology, and visualization, T.T., C.G., and R.T.-M.;
resources, F.V., A.M., and H.A.; formal analysis, investigation, data curation, and writing—original draft, T.T.;
writing—review and editing, all authors; supervision, C.G., F.V., and R.T.-M.; project administration, A.M. and
R.T.-M.; funding acquisition, T.T., C.G., F.V., A.M., and R.T.-M. All authors have read and agreed to the published
version of the manuscript.
Funding: This work was supported by the Regional Council of Bourgogne-Franche Comté, the “Fond Européen
de DEveloppement Régional (FEDER) and funded by Biomère (Paris, France).
Acknowledgments: The authors thank Laurence Noret for her help with handling and maintaining the HPLC
system and Peio Elichiry-Ortiz for setting up and sharing the HPLC method for sugar analysis.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Ivey, M.; Massel, M.; Phister, T.G. Microbial Interactions in Food Fermentations. Annu. Rev. Food Sci. Technol.
2013, 4, 141–162. [CrossRef] [PubMed]
2. Sieuwerts, S.; de Bok, F.A.M.; Hugenholtz, J.; van Hylckama Vlieg, J.E.T. Unraveling Microbial Interactions in
Food Fermentations: From Classical to Genomics Approaches. Appl. Environ. Microbiol. 2008, 74, 4997–5007.
[CrossRef] [PubMed]
3. Wolfe, B.E.; Dutton, R.J. Fermented Foods as Experimentally Tractable Microbial Ecosystems. Cell 2015, 161,
49–55. [CrossRef]
4. Villarreal-Soto, S.A.; Beaufort, S.; Bouajila, J.; Souchard, J.-P.; Taillandier, P. Understanding Kombucha Tea
Fermentation: A Review. J. Food Sci. 2018, 83, 580–588. [CrossRef]
5. Jayabalan, R.; Malbaša, R.V.; Lončar, E.S.; Vitas, J.S.; Sathishkumar, M. A Review on Kombucha
Tea-Microbiology, Composition, Fermentation, Beneficial Effects, Toxicity, and Tea Fungus. Compr. Rev. Food
Sci. Food Saf. 2014, 13, 538–550. [CrossRef]
6. Dufresne, C.; Farnworth, E. Tea, Kombucha, and health: A review. Food Res. Int. 2000, 33, 409–421. [CrossRef]
7. Lynch, K.M.; Zannini, E.; Wilkinson, S.; Daenen, L.; Arendt, E.K. Physiology of Acetic Acid Bacteria and
Their Role in Vinegar and Fermented Beverages. Compr. Rev. Food Sci. Food Saf. 2019, 18, 587–623. [CrossRef]
8. La China, S.; Zanichelli, G.; De Vero, L.; Gullo, M. Oxidative fermentations and exopolysaccharides production
by acetic acid bacteria: A mini review. Biotechnol. Lett. 2018, 40, 1289–1302. [CrossRef]
9. Tran, T.; Grandvalet, C.; Verdier, F.; Martin, A.; Alexandre, H.; Tourdot-Maréchal, R. Microbiological and
technological parameters impacting the chemical composition and sensory quality of kombucha. Compr. Rev.
Food Sci. Food Saf. 2020, 19, 2050–2070. [CrossRef]
Foods 2020, 9, 963 22 of 25

10. Carlier, A.; Pessi, G.; Eberl, L. Microbial Biofilms and Quorum Sensing. In Principles of Plant-Microbe
Interactions; Lugtenberg, B., Ed.; Springer International Publishing: Cham, Switzerland, 2015; pp. 45–52.
ISBN 978-3-319-08574-6.
11. Flemming, H.-C.; Wingender, J.; Szewzyk, U.; Steinberg, P.; Rice, S.A.; Kjelleberg, S. Biofilms: An emergent
form of bacterial life. Nat. Rev. Microbiol. 2016, 14, 563–575. [CrossRef]
12. Gullo, M.; La China, S.; Falcone, P.M.; Giudici, P. Biotechnological production of cellulose by acetic acid
bacteria: Current state and perspectives. Appl. Microbiol. Biotechnol. 2018, 102, 6885–6898. [CrossRef]
13. May, A.; Narayanan, S.; Alcock, J.; Varsani, A.; Maley, C.; Aktipis, A. Kombucha: A novel model system for
cooperation and conflict in a complex multi-species microbial ecosystem. PeerJ 2019, 7, e7565. [CrossRef]
14. Coton, M.; Pawtowski, A.; Taminiau, B.; Burgaud, G.; Deniel, F.; Coulloumme-Labarthe, L.; Fall, A.; Daube, G.;
Coton, E. Unraveling microbial ecology of industrial-scale Kombucha fermentations by metabarcoding and
culture-based methods. FEMS Microbiol. Ecol. 2017, 93. [CrossRef]
15. Teoh, A.L.; Heard, G.; Cox, J. Yeast ecology of Kombucha fermentation. Int. J. Food Microbiol. 2004, 95,
119–126. [CrossRef]
16. Liu, C.-H.; Hsu, W.-H.; Lee, F.-L.; Liao, C.-C. The isolation and identification of microbes from a fermented
tea beverage, Haipao, and their interactions during Haipao fermentation. Food Microbiol. 1996, 13, 407–415.
[CrossRef]
17. Wang, S.; Zhang, L.; Qi, L.; Liang, H.; Lin, X.; Li, S.; Yu, C.; Ji, C. Effect of synthetic microbial community on
nutraceutical and sensory qualities of kombucha. Int. J. Food Sci. Technol. 2020. [CrossRef]
18. Lawson, C.E.; Harcombe, W.R.; Hatzenpichler, R.; Lindemann, S.R.; Löffler, F.E.; O’Malley, M.A.; García
Martín, H.; Pfleger, B.F.; Raskin, L.; Venturelli, O.S.; et al. Common principles and best practices for
engineering microbiomes. Nat. Rev. Microbiol. 2019, 17, 725–741. [CrossRef]
19. Green, S.R.; Gray, P.P. A differential procedure for bacteriological studies useful in the fermentation industry.
Arch. Biochem. Biophys. 1951, 32, 59–69. [CrossRef]
20. Hall, J.F. Detection of wild yeasts in the brewery. J. Inst. Brew. 1971, 77, 513–516. [CrossRef]
21. Pallmann, C.L.; Brown, J.A.; Olineka, T.L.; Cocolin, L.; Mills, D.A.; Bisson, L.F. Use of WL Medium to Profile
Native Flora Fermentations. Am. J. Enol. Vitic. 2001, 52, 198–203.
22. Simonin, S.; Alexandre, H.; Nikolantonaki, M.; Coelho, C.; Tourdot-Maréchal, R. Inoculation of Torulaspora
delbrueckii as a bio-protection agent in winemaking. Food Res. Int. 2018. [CrossRef] [PubMed]
23. Salma, M.; Rousseaux, S.; Sequeira-Le Grand, A.; Alexandre, H. Cytofluorometric detection of wine lactic
acid bacteria: Application of malolactic fermentation to the monitoring. J. Ind. Microbiol. Biotechnol. 2013, 40,
63–73. [CrossRef] [PubMed]
24. Maoura, N.; Mbaiguinam, M.; Nguyen, H.V.; Gaillardin, C.; Pourquie, J. Identification and typing of the
yeast strains isolated from bili bili, a traditional sorghum beer of Chad. Afr. J. Biotechnol. 2005, 4, 646–656.
[CrossRef]
25. Esteve-Zarzoso, B. Identification of yeasts by RFLP analysis of the 5.85 rRNA gene and the two ribosomal
internal transcribed spacers. Int. J. Syst. Bacteriol. 2019, 49, 329–337. [CrossRef] [PubMed]
26. Wang, Y.; Qian, P.-Y. Conservative Fragments in Bacterial 16S rRNA Genes and Primer Design for 16S
Ribosomal DNA Amplicons in Metagenomic Studies. PLoS ONE 2009, 4, e7401. [CrossRef]
27. Stiefel, P.; Schmidt-Emrich, S.; Maniura-Weber, K.; Ren, Q. Critical aspects of using bacterial cell viability
assays with the fluorophores SYTO9 and propidium iodide. BMC Microbiol. 2015, 15, 36. [CrossRef]
28. Cvetković, D.; Markov, S.; Djurić, M.; Savić, D.; Velićanski, A. Specific interfacial area as a key variable in
scaling-up Kombucha fermentation. J. Food Eng. 2008, 85, 387–392. [CrossRef]
29. OIV. Méthode OIV-MA-AS313-01: Acidité Totale; Recueil International des Méthodes d’Analyse-OIV: Paris,
France, 2009.
30. MEBAK 2.6.4.1.1 Ninhydrin Method (Photometric Method, EBC). In MEBAK Wort, Beer, Beer-Based Beverages:
Collection of Brewing Analysis Methods of the Mitteleuropäische Brautechnische Analysenkommission (MEBAK);
Freising-Weihenstephan: Munich, Germany, 2013.
31. Zou, X.; Zhang, S.; Chen, L.; Hu, J.; Hong, F.F. Determination of live and dead Komagataeibacter xylinus cells
and first attempt at precise control of inoculation in nanocellulose production. Microb. Biotechnol. 2020, 13,
458–469. [CrossRef]
32. Millet, V.; Lonvaud-Funel, A. The viable but non-culturable state of wine micro-organisms during storage.
Lett. Appl. Microbiol. 2000, 30, 136–141. [CrossRef]
Foods 2020, 9, 963 23 of 25

33. Lee, J.; Schreiner, R.P. Free amino acid profiles from ‘Pinot noir’ grapes are influenced by vine N-status and
sample preparation method. Food Chem. 2010, 119, 484–489. [CrossRef]
34. Aguilar Uscanga, M.G.; Délia, M.-L.; Strehaiano, P. Brettanomyces bruxellensis: Effect of oxygen on growth
and acetic acid production. Appl. Microbiol. Biotechnol. 2003, 61, 157–162. [CrossRef] [PubMed]
35. Custer, M. Onderzoekingen over Het Gistgeslacht Brettanomyces. Ph.D. Thesis, University of Delft, Delft, The
Netherlands, 1940.
36. Nogueira, A.; le Quéré, J.M.; Drilleau, J.F.; Wosiacki, G. Fermentation Behavior of Apiculate Yeast Hanseniaspora
valbyensis in Cider Processing Conditions; International Commission of Agricultural Engineering (CIGR),
Institut fur Landtechnik: Bonn, Germany, 2008; p. Ref.20.
37. Özcan, S.; Vallier, L.G.; Flick, J.S.; Carlson, M.; Johnson, M. Expression of the SUC2 Gene of Saccharomyces
cerevisiae is Induced by Low Levels of Glucose. Yeast 1997, 13, 127–137. [CrossRef]
38. Meijer, M.M.C.; Boonstra, J.; Verkleij, A.J.; Verrips, C.T. Glucose Repression in Saccharomyces cerevisiae Is
Related to the Glucose Concentration Rather Than the Glucose Flux. J. Biol. Chem. 1998, 273, 24102–24107.
[CrossRef] [PubMed]
39. Marques, W.L.; Raghavendran, V.; Stambuk, B.U.; Gombert, A.K. Sucrose and Saccharomyces cerevisiae: A
relationship most sweet. FEMS Yeast Res. 2016, 16, fov107. [CrossRef] [PubMed]
40. Berthels, N.; Corderootero, R.; Bauer, F.; Thevelein, J.; Pretorius, I. Discrepancy in glucose and fructose
utilisation during fermentation by wine yeast strains. FEMS Yeast Res. 2004, 4, 683–689. [CrossRef]
41. Berthels, N.J.; Cordero Otero, R.R.; Bauer, F.F.; Pretorius, I.S.; Thevelein, J.M. Correlation between
glucose/fructose discrepancy and hexokinase kinetic properties in different Saccharomyces cerevisiae wine
yeast strains. Appl. Microbiol. Biotechnol. 2008, 77, 1083–1091. [CrossRef]
42. Velasco-Bedrán, H.; López-Isunza, F. The unified metabolism of Gluconacetobacter entanii in continuous and
batch processes. Process Biochem. 2007, 42, 1180–1190. [CrossRef]
43. Sharma, C.; Bhardwaj, N.K. Biotransformation of fermented black tea into bacterial nanocellulose via
symbiotic interplay of microorganisms. Int. J. Biol. Macromol. 2019, 132, 166–177. [CrossRef]
44. De Ley, J. Comparative carbohydrate metabolism and a proposal for a phylogenetic relationship of the acetic
acid bacteria. Microbiology 1961, 24, 31–50. [CrossRef]
45. De Ley, J.; Schell, J. Lactate and Pyruvate Catabolism in Acetic Acid Bacteria. J. Gen. Microbiol. 1962, 29,
589–601. [CrossRef]
46. Prust, C.; Hoffmeister, M.; Liesegang, H.; Wiezer, A.; Fricke, W.F.; Ehrenreich, A.; Gottschalk, G.;
Deppenmeier, U. Complete genome sequence of the acetic acid bacterium Gluconobacter oxydans. Nat.
Biotechnol. 2005, 23, 195–200. [CrossRef] [PubMed]
47. Dong, Y.; Hu, J.; Fan, L.; Chen, Q. RNA-Seq-based transcriptomic and metabolomic analysis reveal stress
responses and programmed cell death induced by acetic acid in Saccharomyces cerevisiae. Sci. Rep. 2017, 7.
[CrossRef] [PubMed]
48. Drysdale, G.S.; Fleet, G.H. The effect of acetic acid bacteria upon the growth and metabolism of yeasts during
the fermentation of grape juice. J. Appl. Bacteriol. 1989, 67, 471–481. [CrossRef]
49. Chen, C.; Liu, B.Y. Changes in major components of tea fungus metabolites during prolonged fermentation.
J. Appl. Microbiol. 2000, 89, 834–839. [CrossRef]
50. Goh, W.N.; Rosma, A.; Kaur, B.; Fazilah, A.; Karim, A.A.; Bhat, R. Fermentation of black tea broth (Kombucha):
I. Effects of sucrose concentration and fermentation time on the yield of microbial cellulose. Int. Food Res. J.
2012, 19, 109–117.
51. Blanc, P.J. Characterization of the tea fungus metabolites. Biotechnol. Lett. 1996, 18, 139–142. [CrossRef]
52. Sievers, M.; Lanini, C.; Weber, A.; Schuler-Schmid, U.; Teuber, M. Microbiology and fermentation balance in
a kombucha beverage obtained from a tea fungus fermentation. Syst. Appl. Microbiol. 1995, 18, 590–594.
[CrossRef]
53. Andjelković, U.; Milutinović-Nikolić, A.; Jović-Jovičić, N.; Banković, P.; Bajt, T.; Mojović, Z.; Vujčić, Z.;
Jovanović, D. Efficient stabilization of Saccharomyces cerevisiae external invertase by immobilisation on
modified beidellite nanoclays. Food Chem. 2015, 168, 262–269. [CrossRef]
54. Beteta, P.; Gascon, S. Localization of invertase in yeast vacuoles. FEBS Lett. 1971, 13, 297–300. [CrossRef]
55. Margetić, A.; Vujčić, Z. Comparative study of stability of soluble and cell wall invertase from Saccharomyces
cerevisiae. Prep. Biochem. Biotechnol. 2017, 47, 305–311. [CrossRef]
Foods 2020, 9, 963 24 of 25

56. Neumann, N.P.; Lampen, J.O. Glycoprotein structure of yeast invertase. Biochemistry 1969, 8, 3552–3556.
[CrossRef] [PubMed]
57. Blondin, B.; Ratomahenina, R. Purification and Properties of the β-Glucosidase of a Yeast Capable of
Fermenting Cellobiose to Ethanol: Dekkera intermedia Van der Walt. Eur. J. Microbiol. Biotechnol. 1983, 17, 1–6.
[CrossRef]
58. Conterno, L.; Joseph, C.M.L.; Arvik, T.J.; Henick-Kling, T.; Bisson, L.F. Genetic and Physiological
Characterization of Brettanomyces bruxellensis Strains Isolated from Wines. Am. J. Enol. Vitic. 2006,
139–147.
59. Fia, G.; Giovani, G.; Rosi, I. Study of beta-glucosidase production by wine-related yeasts during alcoholic
fermentation. A new rapid fluorimetric method to determine enzymatic activity. J. Appl. Microbiol. 2005, 99,
509–517. [CrossRef] [PubMed]
60. Roach, M.J.; Borneman, A.R. New genome assemblies reveal patterns of domestication and adaptation across
Brettanomyces (Dekkera) species. BMC Genom. 2020, 21, 14. [CrossRef]
61. Steenwyk, J.L.; Opulente, D.A.; Kominek, J.; Shen, X.-X.; Zhou, X.; Labella, A.L.; Bradley, N.P.; Eichman, B.F.;
Čadež, N.; Libkind, D.; et al. Extensive loss of cell-cycle and DNA repair genes in an ancient lineage of
bipolar budding yeasts. PLoS Biol. 2019, 17, e3000255. [CrossRef]
62. Alexandre, H.; Heintz, D.; Chassagne, D.; Guilloux-Benatier, M.; Charpentier, C.; Feuillat, M. Protease A
activity and nitrogen fractions released during alcoholic fermentation and autolysis in enological conditions.
J. Ind. Microbiol. Biotechnol. 2001, 26, 235–240. [CrossRef]
63. Guilloux-Benatier, M.; Chassagne, D.; Alexandre, H.; Charpentier, C.; Feuillat, M. Influence of yeast autolysis
after alcoholic fermentation on the development of Brettanomyces/Dekkera in wine. OENO One 2001, 35, 157.
[CrossRef]
64. Charoenchai, C.; Fleet, G.H.; Henschke, P.A.; Todd, B.E.N. Screening of non-Saccharomyces wine yeasts for
the presence of extracellular hydrolytic enzymes. Aust. J. Grape Wine Res. 1997, 3, 2–8. [CrossRef]
65. Younes, B.; Cilindre, C.; Villaume, S.; Parmentier, M.; Jeandet, P.; Vasserot, Y. Evidence for an Extracellular
Acid Proteolytic Activity Secreted by Living Cells of Saccharomyces cerevisiae PlR1: Impact on Grape Proteins.
J. Agric. Food Chem. 2011, 59, 6239–6246. [CrossRef] [PubMed]
66. Guilloux-Benatier, M.; Chassagne, D. Comparison of Components Released by Fermented or Active Dried
Yeasts after Aging on Lees in a Model Wine. J. Agric. Food Chem. 2003, 51, 746–751. [CrossRef]
67. Torres, A.P.; Oliveira, F.A.R.; Silva, C.L.M.; Fortuna, S.P. The influence of pH on the kinetics of acid hydrolysis
of sucrose. J. Food Process Eng. 1994, 17, 191–208. [CrossRef]
68. Jakob, F.; Quintero, Y.; Musacchio, A.; Estrada-de los Santos, P.; Hernández, L.; Vogel, R.F. Acetic acid bacteria
encode two levansucrase types of different ecological relationship. Environ. Microbiol. 2019, 21, 4151–4165.
[CrossRef]
69. Balasubramaniam, K.; Kannangara, P.N. Sucrose Phosphorylase and Invertase Activities in Bacteria. J. Natl.
Sci. Found. Sri Lanka 1982, 10, 169–180. [CrossRef]
70. Albergaria, H.; Francisco, D.; Gori, K.; Arneborg, N.; Gírio, F. Saccharomyces cerevisiae CCMI 885 secretes
peptides that inhibit the growth of some non-Saccharomyces wine-related strains. Appl. Microbiol. Biotechnol.
2010, 86, 965–972. [CrossRef]
71. Branco, P.; Viana, T.; Albergaria, H.; Arneborg, N. Antimicrobial peptides (AMPs) produced by Saccharomyces
cerevisiae induce alterations in the intracellular pH, membrane permeability and culturability of Hanseniaspora
guilliermondii cells. Int. J. Food Microbiol. 2015, 205, 112–118. [CrossRef] [PubMed]
72. Pereznevado, F.; Albergaria, H.; Hogg, T.; Girio, F. Cellular death of two non-Saccharomyces wine-related
yeasts during mixed fermentations with Saccharomyces cerevisiae. Int. J. Food Microbiol. 2006, 108, 336–345.
[CrossRef] [PubMed]
73. Wang, C.; Mas, A.; Esteve-Zarzoso, B. The Interaction between Saccharomyces cerevisiae and Non-Saccharomyces
Yeast during Alcoholic Fermentation Is Species and Strain Specific. Front. Microbiol. 2016, 7, Article 502.
[CrossRef] [PubMed]
74. Gobert, A.; Tourdot-Maréchal, R.; Morge, C.; Sparrow, C.; Liu, Y.; Quintanilla-Casas, B.; Vichi, S.; Alexandre, H.
Non-Saccharomyces Yeasts Nitrogen Source Preferences: Impact on Sequential Fermentation and Wine
Volatile Compounds Profile. Front. Microbiol. 2017, 8. [CrossRef] [PubMed]
75. Renault, P.E.; Albertin, W.; Bely, M. An innovative tool reveals interaction mechanisms among yeast
populations under oenological conditions. Appl. Microbiol. Biotechnol. 2013, 97, 4105–4119. [CrossRef]
Foods 2020, 9, 963 25 of 25

76. Soares, E.V. Flocculation in Saccharomyces cerevisiae: A review: Yeast flocculation: A review. J. Appl. Microbiol.
2011, 110, 1–18. [CrossRef]
77. Greig, D.; Travisano, M. The Prisoner’s Dilemma and polymorphism in yeast SUC genes. Proc. R. Soc. Lond.
B Biol. Sci. 2004, 271. [CrossRef]
78. Celiker, H.; Gore, J. Competition between species can stabilize public-goods cooperation within a species.
Mol. Syst. Biol. 2012, 8, 621. [CrossRef] [PubMed]
79. Jayabalan, R.; Marimuthu, S.; Swaminathan, K. Changes in content of organic acids and tea polyphenols
during kombucha tea fermentation. Food Chem. 2007, 102, 392–398. [CrossRef]
80. Harrington, K.I.; Sanchez, A. Eco-evolutionary dynamics of complex social strategies in microbial
communities. Commun. Integr. Biol. 2014, 7, e28230. [CrossRef] [PubMed]
81. Sanchez, A.; Gore, J. Feedback between Population and Evolutionary Dynamics Determines the Fate of
Social Microbial Populations. PLoS Biol. 2013, 11, e1001547. [CrossRef] [PubMed]
82. Langenberg, A.-K.; Bink, F.J.; Wolff, L.; Walter, S.; von Wallbrunn, C.; Grossmann, M.; Heinisch, J.J.;
Schmitz, H.-P. Glycolytic Functions Are Conserved in the Genome of the Wine Yeast Hanseniaspora uvarum,
and Pyruvate Kinase Limits Its Capacity for Alcoholic Fermentation. Appl. Environ. Microbiol. 2017, 83.
[CrossRef]
83. Leistner, L. Basic aspects of food preservation by hurdle technology. Int. J. Food Microbiol. 2000, 55, 181–186.
[CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).