“Análisis de la diversidad bacteriana y funcional de la tuba con un enfoque bioinformático”

T E S I S

QUE PARA OBTENER EL GRADO DE MAESTRO EN BIOCIENCIAS

PRESENTA:

I.B.Q. Fernando Astudillo Melgar

DIRECTOR: Dr. Gerardo Huerta Beristain

CODIRECTOR: Dr. José Utrilla Carreri

Chilpancingo de los Bravo, Gro., junio 2018



Este trabajo fue realizado en el Laboratorio de Investigación en Biotecnología de la Facultad de Ciencias Químico Biológicas de la Universidad Autónoma de Guerrero. Se contó con la colaboración del Laboratorio de Biología de Sistemas y Biología Sintética del Centro de Ciencias Genómicas de la Universidad Nacional Autónoma de México y con el Departamento de Microbiología Molecular del Instituto de Biotecnología de la Universidad Nacional Autónoma de México

Bajo la dirección de:

Dr. Gerardo Huerta Beristaín

La codirección de:

Dr. José Utrilla Carreri

La asesoría externa de:

Dr. Adrián Ochoa Leyva Biol. Filiberto Sánchez López

M.B. José Raunel Tinoco Valencia Dr. Carlos Ortuño Pineda

Al comité tutoral:

Dra. Ma. Elena Moreno Godínez Dra. Jeiry Toribio Jiménez

Dr. Miguel Ángel Mendoza Catalán

Durante los estudios en la Maestría en Biociencias, el C. Fernando Astudillo Melgar, recibió beca CONACYT con No. registro: 597135. Y apoyo del proyecto FOMIX-CONACyT-Edo. De Guerrero 249671.

También se agradece el apoyo proporcionado del proyecto DGAPA-PAPIIT IA201518


Agradecimientos.

Al Dr. Gerardo. No tengo palabras para agradecer todo lo que ha hecho por mi, es una excelente persona, confió en mi sin ni siquiera conocerme y eso es algo que no cualquiera hace. Hizo mi estadía en la maestría tan agradable, aprendí tanto de usted y no nada más como profesionista, sino también como persona. Gracias por las enseñanzas, las clases, los momentos de risa y el bully. Espero que siga teniendo mucho éxito Dr. Lo admiro mucho.

A mis asesores externos. Al Dr. José. Gracias por la oportunidad de trabajar en su laboratorio por brindarme las herramientas para el desarrollo del trabajo y por las observaciones y consejos durante el desarrollo del proyecto. Al Dr. Adrián. Por todo su apoyo, por el material y orientación brindada, por permitirme trabajar en su laboratorio y por su asesoría.

A mi sínodo. Al Dr Migue, Dra. Ma Elena y Dra. Jeiry. Por su apoyo, revisiones y consejos que tuvieron en bien darme durante el desarrollo del proyecto. Además gracias por facilitarme equipos y material cuando los necesité.

Biol. Filiberto y al M.B. Raunel. Por su apoyo y asesoría en la realización de experimentos.

A los profesores de seminario. Al Dr. Carlos. Muchas gracias por todo su apoyo, paciencia, consejos y tanto más, es una gran persona, me ayudo tanto en muchos aspectos, es un gran profesionista, lo admiro mucho. A la Dra. Paty. Gracias por todas las veces que me ayudo, siempre que necesite un favor usted tuvo en bien ayudarme, sus consejos, sus pláticas, la admiro mucho. Gracias a ambos son unos excelentes profesores de seminario y amigos.

A la Maestra Daysi. Gracias por brindarme la oportunidad de convivir con su familia, de abrirme las puertas de su casa, las pláticas, los momentos de risa, los consejos, los juegos y por muchas cosas más muchas gracias!. A Canek y Nikte son unos chicos muy geniales, gracias por todos los juegos y aventuras que tuvimos, estoy seguro que serán unos excelentes profesionistas y personas, ya que tienen muy buenos ejemplos en casa, los aprecio mucho.

A mis amigos de generación. No pongo compañeros porque para mí, todos son mis amigos, gracias por todo! Fue una aventura muy bonita, llena de momentos buenos, algunos malos, mucho aprendizaje, convivios, risas, estrés, bueno mucho de todo. Ana, Carlos, Dave, Mary, Augusto, Karen y Yonas, gracias por la oportunidad de haberlos conocido, de brindarme su amistad, espero siempre sigamos en contacto.

A todos los GHB team (Banduqui). Itzel, Muchas gracias por todo, eres una gran amiga y te quiero mucho, compartimos tanto estrés y alegrías, aventuras en donde estuvimos solos trabajando mientras toda la escuela en fiesta, bueno tanto, eres una gran amiga. Sofi, Romy, Angy, Ana, Rous, Alhee, son unas personas muy lindas, gracias por todo su apoyo y cariño, los momentos de locura y risa que me brindaron, la confianza, y la enseñanza de que con esfuerzo se pueden lograr muchas cosas, las quiero mucho. Marino, Leyva gracias por la ayuda y los momentos de risa. Ada, gracias por todas las buenas pláticas, los favores, las aventuras, el bully, pero sobre todo por tu amistad. Itzel, Aydee y Brian, mis “niños” gracias por confiar y seguir este proyecto tan bonito, gracias por permitirme conocerlos y en especial por ser muy buenos amigos. A todos los miembros “nuevos” del equipo, Gema, Ale, Liss, Alberto, Venecia, Genesis y Perla, por la convivencia compartida.

A mi amiga Amparo, por todos los buenos ratos vividos, los consejos, las pláticas, los momentos de estrés, bueno por muchas cosas, eres una gran amiga te quiero mucho.

Al Dr. Javier. Por su apoyo, las pláticas, sus consejos y por el equipo que me facilito durante toda la maestría.

A todos los profesores que conforman esta gran maestría que es Biociencias. Por todo el apoyo y enseñanzas obtenidas durante el transcurso del posgrado. A la Biol. Lupita por el apoyo en el área administrativa.

Atte:

Fernando Astudillo Melgar


Dedicatoria.

El presente trabajo va dedicado principalmente a dos grandes personas que se han sacrificado y luchado por brindarme una buena educación y valores, y a los que les debo todo. A mi padre, al cual le agradezco todos sus consejos, amor, cuidado, apoyo, orientación, los cuales me han servido a lo largo de mi vida. A mi madre, que siempre ha estado cuidándome con sus consejos apoyo y amor incondicional. Son unos excelentes seres humanos, siempre han sido y serán mi motor para salir adelante, un ejemplo a seguir como profesionistas y como personas; los admiro y amo con todo el corazón.

A mis hermanos, mis sobrinos, mi cuñada. Gracias por todo el apoyo brindado, los consejos, los ánimos, los encargos que hacía, las vueltas que a veces los hice pasar, los abrazos que me llenan de energía para seguir adelante, los momentos que se preocupaban junto conmigo y por muchas cosas más gracias. Con este escrito quiero compartirles uno de mis sueños que se realiza porque es un gran logro que también les pertenece ya que me han brindado todo cuanto he necesitado, en especial el cariño y apoyo durante el transcurso de mi vida.

A mis tíos, mi abuelita y mi prima. Que durante mis estancias en Morelos siempre me estuvieron cuidando y al pendiente de mi. Sus apapachos, comidas, paseos, ánimos, consejos, en fin… toda la energía que a uno le hace falta cuando está lejos del hogar base, y digo base porque sin duda ustedes me hicieron sentir en un hogar. Gracias.

Fernando Astudillo Melgar.

ANÁLISIS DE LA DIVERSIDAD BACTERIANA Y FUNCIONAL DE LA TUBA CON UN ENFOQUE BIOINFORMÁTICO.

Table of contents

Abstract.................................................................................................................................. 2

Introduction............................................................................................................................ 2

Material and Methods............................................................................................................ 3

Results.................................................................................................................................... 5

Discussion............................................................................................................................... 7

References............................................................................................................................ 11

Figures.................................................................................................................................. 14

Tables.................................................................................................................................... 20

Supplementary material....................................................................................................... 21

  • 1 Bacterial diversity, population dynamics and functional analysis

  • 2 of commercial and laboratory fermented palm wine (Tuba).

  • 3 Fernando Astudillo-Melgar1,3, Adrián Ochoa-Leyva2, José Utrilla3* Gerardo Huerta-

  • 4 Beristain1*

5

  • 6 1.- Laboratorio de Investigación en Biotecnología, Universidad Autónoma de Guerrero,

  • 7 Chilpancingo, México.

  • 8 2.- Departamento de Microbiología molecular, Instituto de Biotecnología- Universidad

  • 9 Nacional Autónoma de México, Cuernavaca, México.

  • 10 3.- Programa de Biología de Sistemas y Biología Sintética, Centro de Ciencias Genómicas -

  • 11 Universidad Nacional Autónoma de México, Cuernavaca, México.

  • 12 *Co-corresponding authors

  • 13 e-mail: utrilla@ccg.unam.mx, hbgerardo@gmail.com

14

15

16


17 Abstract.





38 Introduction.

  • 39 A wide variety of fermented food products such as yogurt, alcoholic beverages, bread and

  • 40 sauces are produced worldwide. During the production process of these fermented foods

  • 41 different microorganisms contribute to the organoleptic and biochemical characteristics

  • 42 (Tang et al. 2017). Recent studies in fermented food have shown that microbial ecology

  • 43 aspects such as diversity, their spatial distribution and ecological interaction, have a strong

  • 44 influence on metabolic production and chemical composition (Escalante et al. 2015).

  • 45 Bacterial consortia interactions in fermented foods promote process of polymer degradation

  • 46 and production of metabolites of interest such as alcohol, aromatics, acetate, lactate among

  • 47 others that contribute to functional and organoleptic properties (Tamang et al. 2016).

  • 48 Palm wine is a traditional beverage made using the sap collected from palm trees. It is

  • 49 consumed in different parts of the world, in Africa it is known as "legmi", in South India as

  • 50 "kallu", while in Borneo it has the names of "bahar" and "goribon" (Velázquez-Monreal et

  • 51 al., 2011). The differences among these beverages are the production process, the coconut

  • 52 tree species and the plant part where the sap is collected (Santiago-Urbina & Ruíz-Terán

  • 53 2014). In Mexico, several traditional fermented beverages are produced such as pulque

  • 54 (Escalante et al. 2016), pozol (Díaz-Ruíz et al. 2003) and Tuba (De la Fuente-Salcido et al.

  • 55 2015). Tuba was brought to Mexico by Philippine influence during the Spanish colonial

  • 56 period. This beverage is produced in the southern pacific coast of Mexico (Guerrero,

  • 57 Colima, Michoacan states). It is obtained from the sap of the inflorescences of Cocos

  • 58 nucifera L and it is consumed as a traditional beverage empirically used as an aid in

  • 59 gastrointestinal problems and as a rehydration drink (Velázquez-Monreal et al. 2011; de la

  • 60 Fuente-Salcido et al. 2015).

  • 61 The importance of bacteria in fermented foods has promoted the application of different

  • 62 strategies to analyze the bacterial diversity and role during elaboration of fermented

  • 63 products. The use of massive sequencing technologies together with recent bioinformatics

  • 64 methods, such as QIIME for diversity analysis (Navas-Molina et al. 2015; Caporaso et al.

  • 65 2011) and PICRUST for functional inference (Langille et al. 2013), have increased the

  • 66 taxonomic and functional information of uncultured bacterial communities in different

  • 67 ecosystems (Filippis et al. 2017). However, those methods have been used mainly in

  • 68 projects such as Human Microbiome and Earth Microbiome (Creer et al. 2016).

  • 69 Nevertheless, in the food area, the applications of them are limited. Some studies in

  • 70 traditional Asian liquors and sauces have established a correlation between microbial

  • 71 diversity and organoleptic properties, increasing the information about bacterial

  • 72 communities in Asian products such as Yucha (Tang et al. 2017; Zhang et al. 2016) and

  • 73 some Mexican traditional beverage as Pulque (Escalante et al. 2008).

  • 74 Here, we study the fermentation profile, population dynamics and bacterial diversity of

  • 75 Tuba produced in the Guerrero coast of Mexico. We sampled sap that was fermented under

  • 76 controlled conditions and sampled commercial Tuba. Using 16S amplicon sequencing and

  • 77 metabolic characteristics, we were able to analyze the diversity and infer functionality of

  • 78 bacterial communities present in all tuba samples. This work provides a basis for the further

  • 79 functional characterization of Tuba in its production process, probiotic potential and other

  • 80 functions as antibiotic and antioxidant biosynthesis.

81

82 Material and Methods.

83 Sample collection.

  • 84 Sap samples were collected from three visibly healthy palm trees in a rural area in

  • 85 Acapulco, Guerrero, Mexico. Commercial Tuba samples were obtained from four different

  • 86 artesian producers in diamante zone from Acapulco, Guerrero Mexico (Figure 1). The

  • 87 climatological conditions of the samples collection site at the sampling day are described in

  • 88 table 1. Samples were transported in sanitized coolers to the laboratory for fermentation and

  • 89 analysis. The sap from palm trees was tagged with the following code a “P” followed by

  • 90 the number of the palm tree and “T” which means the fermentation time (i.e. P1T0).

  • 91 Commercial samples were tagged using the letter L followed by a consecutive number that

  • 92 symbolize the number of the establishment where each sample was obtained.

93

94 Fermentation in laboratory controlled conditions.

  • 95 Each sap sample (100 mL as working volume) was fermented in four 250 mL Erlenmeyer

  • 96 flasks corresponding to 0, 12, 24 and 35 hours of fermentation. They were incubated at

  • 97 30°C and 100 rpm of shaking speed in an orbital incubator. Samples were centrifuged

  • 98 (4000 rpm x 15 min) and the pellets were used for DNA extraction, while the supernatants

  • 99 were stored at -20°C for further analysis.

100 Metabolic composition characterization.

101 Sucrose, glucose, fructose, water-soluble proteins, acetic acid, ethanol and pH.

  • 102 Sugars, organic acids and ethanol from laboratory fermented and commercial samples were

  • 103 quantified using two HPLC methods following column manufacturer conditions. Glucose,

  • 104 fructose, sucrose and xylose were quantified using an Aminex HPX-87P (Biorad) column

  • 105 with an IR detector. Acetate and ethanol concentrations were measured using Aminex

  • 106 HPX-87H (Biorad) column and a UV 210 nm detector. Water-soluble proteins were

  • 107 measured by Bradford method modified by Fernández & Galván, 2015. The pH was

  • 108 measured using a potentiometer with 1 mL of the sample.

  • 109 16S amplicon library preparation and sequencing.

  • 110 The DNA extraction from all the samples was performed using the ZR Soil Microbe DNA

  • 111 MiniPrep™ kit (Zymo Research) according to the manufacturer protocol. The DNA was

  • 112 quantified using Qubit Fluorometric Quantitation (Thermo Fisher Scientific). 12.5 ng of

  • 113 total DNA was used for PCR of amplicons of the V3-V4 regions of the 16S rRNA

  • 114 ribosomal gene (Table 2) as described by the Illumina Protocol. All the PCR products were

  • 115 purified (AMPure XP beads - Illumina products) and quantified (Qubit). Finally, all the

  • 116 libraries were sequenced by Illumina MiSeq.

  • 117 Bioinformatics and Statistical analysis

  • 118 The sequences were analyzed using QIIME version -1.9.1software (Caporaso et al. 2011) in

  • 119 Python 2.7. The total sequences were clustered using UCLUST into OTUs tables

  • 120 (operational taxonomic units) using the Greengene database (GG 13_8_otus) as reference

  • 121 with a range of 97% of similarity and using the closed system with the command

  • 122 pick_closed_reference_otus.py. Taxonomy summaries including relative abundance data

  • 123 were generated using summarize_taxa.py, plot_taxa_summary.py and

  • 124 plot_taxa_through_plots.py commands. In all the cases, we used the data filtering option of

  • 125 0.01% in abundances because it is reported that filtering data base decreases the estimation

  • 126 error (Kuczynski et al. 2012; Navas-Molina et al. 2015).

  • 127 Alpha diversity was evaluated using the function of alpha_rarefaction.py from QIIME, that

  • 128 calculate alpha diversity on each sample in an OTUs table, using a variety of alpha

  • 129 diversity metrics as Shannon-Wiener index, Simpson index, Otus_observed and Chao1

  • 130 value. Each metrics result were analyzed by ANOVA applying the Tukey-Kramer test (0.95

  • 131 confidence interval) to estimate significance difference between the samples. Beta diversity

  • 132 was calculated by beta_diversity_through_plots.py, a workflow script for computing beta

  • 133 diversity distance matrices (UniFrac unweighted method) and generating Principal

  • 134 coordinates analysis (PCoA) plots from QIIME.

  • 135 The normalized OTUs table (0.01% abundance filter) was used to estimate functional

  • 136 features present in the samples, using PICRUSt version 1.1.0 (Langille et al. 2013) and the

  • 137 Greengenes databases 16S_13_5 and KO_13_5. The OTUs table was normalized to obtain

  • 138 the metagenomic functional predictions at different hierarchical KEGG levels using

  • 139 normalize_metagenomes.py, predict_metagenomes.py and categorize_by_function.py

  • 140 scripts of the same software.

  • 141 For the statistical studies of the functions, we used STAMP (Statistical analysis of

  • 142 taxonomic and functional profiles version 2.1.3), through ANOVA analysis applying the

  • 143 Tukey-Kramer test (0.95 confidence interval) to evaluated gene abundance of each

  • 144 function. R statistical program (version 3.3.3) was used to make plots using “ggplot2” and

  • 145 “dplyr” libraries.

  • 146

147 Results.

148 Sample Composition

  • 149 To determine the microenvironmental conditions that affect the microbial communities and

  • 150 metabolic characteristics of the Tuba samples, we evaluated the sugars (sucrose, glucose

  • 151 and fructose), water-soluble proteins, ethanol and acetic acid concentrations as well as the

  • 152 pH value (Supplementary Table 1S). Tuba P1 was the sample with the highest

  • 153 concentration in glucose and fructose with 61.4 and 47.3 g/L respectively at 12 hours, 4.7%

  • 154 (v/v) in ethanol and 6.0 g/L in acetate at 35 hours (Figure 2A). Tuba P2 was the sample

  • 155 with lowest concentration of monosaccharides at the beginning of the fermentation and

  • 156 high sucrose concentration (121.7 g/L), however, at the last fermentation time the ethanol

  • 157 and acetate concentrations were low with 3.5 g/L and 0.6% (v/v) respectively (Figure 2B).

  • 158 Tuba P3 showed the highest concentration of ethanol (5% v/v) at the end of the

  • 159 fermentation, nevertheless, the glucose and fructose concentration were 39.8 and 29.1 g/L

  • 160 respectively at 12 hours (Figure 2C). The pH values in Tuba P1, P2 and P3 decreased from

  • 161 3.7 to 2.8 during the fermentation process. The water-soluble protein concentration of the

  • 162 Tuba samples showed low values from 0.006 to 0.01 g/L. In the case of the commercial

  • 163 samples, all of them presented different composition values, nevertheless they had an

  • 164 average values of 40.5 g/L of sucrose, 40.0 g/L of glucose, 42.53 g/L of fructose, 1.6 g/L of

  • 165 acetic acid, 0.1% (v/v) of ethanol and a pH of 4 (Figure 2D).

166 Taxonomic classification.

  • 167 A total of 302,398 sequences were obtained from the Tuba amplicon libraries, with an

  • 168 average of 75,594 sequences per Tuba fermented under controlled conditions (distributed as

  • 169 follows, for the Tuba P1 74,860 reads were obtained; for the Tuba P2, 75,623; for the Tuba

  • 170 P3 76,298) and the four commercial samples had an average of 75,617 sequences. A total

  • 171 of 123 OTUs were detected in all Tuba samples. However, filtering data base with 0.01%

  • 172 relative abundance filter, the OTUs were reduced to 28 as the more abundance. The

  • 173 taxonomic identification was elaborated using the last filter mentioned, which demonstrates

  • 174 the 10 most representative genera of the 16 Tuba samples (Figure 3). The genera that

  • 175 predominate in all the samples were Fructobacillus, Leuconostoc, Gluconacetobacter,

  • 176 Sphingomonas , Vibrio and some genera of the Enterobacteriaceae family. Additionally,

  • 177 analyzing the Enterobacteriaceae populations with the lower abundance we found genera as

  • 178 Erwinia, Klebsiella, Serratia and Cronobacter (Supplementary Figure 1S). The population

  • 179 dynamic had a similar trend in Tuba fermented in controlled conditions but with different

  • 180 percentage in the abundances; we observed an increase of lactic acid bacteria (LABs) until

  • 181 24 h, acetic acid bacteria (AABs) and some proteobacteria as Sphingomonas through the

  • 182 fermentation time and a decrease of Vibrio genus (Figure 3).

183 Diversity analysis.

  • 184 Alpha diversity tests were performed using the OUTs table obtained with the 0.01 % filter

  • 185 and grouped according to the origin of the sample. We observed a similar behavior in all

  • 186 the four analysis, that means, no matter what base-priority was in the analysis as richness

  • 187 (observed_otus), dominance (Simpson), equity (Shannon index) or singletones (Chao1

  • 188 value) it did not affect diversity results (Supplementary Figure 2S). Tuba P1 was the most

  • 189 diverse with the highest values in the four diversity index, then Tuba P3 and commercial

  • 190 Tuba samples had similar index values, and finally Tuba P2 was the least diverse with the

  • 191 lowest values. After of ANOVA statistical analysis, we found that in Chao1 and

  • 192 Observed_otus tests Tuba P2 was the only showing significant difference. Nevertheless, in

  • 193 Shannon and Simpson index the four groups showed significant difference among each

  • 194 other (Table 3).

  • 195 Beta diversity with Unweighted UniFrac distance was determined using the 0.01% filter.

  • 196 We did not observe groupings by fermentation time (Figure 4A) however, a grouping was

  • 197 observed by origin of the samples (Figure 4B). In the graphic of origin of the sample we

  • 198 also observe a grouping by quadrant of the all the Tuba samples, however Tuba P2 showed

  • 199 the greatest dispersion in the data, which indicated a big difference between the

  • 200 fermentation times in Tuba P2. Similar effect is observed in Tuba P1 where two

  • 201 fermentation times (0 h and 35 h) show similar beta diversity values compared to

  • 202 commercial samples and Tuba P3. Otherwise, the samples, which were in the same

  • 203 quadrant as Tuba P3 and the commercial Tuba, were considered strongly related (Figure 4).

204 Functional inference

  • 205 After diversity distribution analysis we sought to understand the functionality of the

  • 206 bacterial community in Tuba fermentation, therefore we used PICRUSt algorithm to predict

  • 207 the metagenomic profiles of the samples. Initially, we obtained functional characteristics of

  • 208 the 3 KEGG levels (Level 1: general cellular functions, Level 2: Specific functions i.e.

  • 209 different metabolism, and Level 3: Specific pathway associate with specific function)

  • 210 (

    http://www
    . genome.ad.jp/kegg/). We limited our analysis to the level 3 and we discarded

  • 211 elementary cellular functions such as replication, translation, and functions associated with

  • 212 human diseases (cancer) or poorly characterized functions, to analyze specific genes related

  • 213 with functions of biotechnological relevance. Considering the 328 registered functions on

  • 214 KEGG, we found the 19 most abundant functions associated with carbohydrates metabolic

  • 215 process, vitamins, amino acid, antibiotics and antioxidant molecules biosynthesis (Figure

  • 216 5), this suggested that the production of those compounds may be taking place during Tuba

  • 217 fermentation. After an ANOVA test, we found functions without significant difference as

  • 218 the carotenoid biosynthesis (Figure 6A), this means that no matter what is the sample

  • 219 origin, this function may have present at the same gene abundance in the four groups.

  • 220 Otherwise, there were functions with significant difference, such as peptidases biosynthesis

  • 221 that had more gene abundance in Tuba P2 samples (Figure 6B). Each sample had more

  • 222 abundance in genes associated with a specific function, for example, antioxidant, antibiotic

  • 223 compounds, and folate biosynthesis in Tuba P3, lipopolysaccharide biosynthesis and

  • 224 Lysine genes in Tuba P1, finally the 4 commercial Tuba samples may have bacteria with

  • 225 genes associated mainly with folate biosynthesis and peptidases. Our study allowed to

  • 226 analyze if some of bacterial genera found in Tuba may had gene associated with enzymes

  • 227 of carotenoid biosynthesis, we observed that Sphingomonas and Gluconacetobacter had

  • 228 more abundance percentage in the enzyme 15-cis-phytoene synthase (Figure 7).

  • 229

230 Discussion.

  • 231 In the present study, we carried out for the first time the identification of bacterial diversity,

  • 232 the fermentation dynamics in terms of bacterial populations and metabolic changes during

  • 233 Tuba fermentations comparing between laboratory controlled conditions and commercial

  • 234 samples. This comparison was realized by a combination of metabolic analysis and 16S

  • 235 amplicon sequencing during the Tuba fermentation, as well as to infer functions of

  • 236 biotechnological interest that the Tuba may present during the fermentative process.

  • 237 We found that the average of total sugar concentration in sap of the palm trees in the Tuba

  • 238 samples was 130 g/L. Where it contained 77.06% of sucrose, 12.81% of glucose and

  • 239 10.15% of fructose, without presence of xylose. In a study with sap of Phoenix dactylifera

  • 240 was reported that it contained 95.27 % of sucrose, 2.51% glucose and 1.61% of fructose

  • 241 with a neutral pH of 7-7.4 (Santiago-Urbina & Ruíz-Terán 2014). These results suggested

  • 242 that the sap composition is dependent of the palm type. The concentration of sucrose from

  • 243 the sap samples at the start of the laboratory controlled fermentation was high, from 85 g/L

  • 244 to 121 g/L (Supplementary Table 1S). Then, after 12 hours of fermentation for Tuba P1 and

  • 245 P3, and after 24 hours for Tuba P2 the concentration of sucrose was reduced, increasing the

  • 246 concentration of glucose and fructose, presumably by invertase-mediated hydrolysis. The

  • 247 different behavior of Tuba P2 sample may be related with low sucrose hydrolysis, delaying

  • 248 the fermentation process. This may be caused by a lower yeast abundance than other

  • 249 samples, the yeast abundance was not measured in this study, but we did not identify any

  • 250 ethanologenic bacteria, then we may attribute all the ethanol production to yeasts present in

  • 251 the samples. Also, the high concentration of sucrose may also cause retro-inhibition of the

  • 252 invertase enzyme, and the pH values may reduce its catalytic activity (Hsieh et al. 2006;

  • 253 Goosen et al. 2007). Hence, we can only propose that the hydrolysis of sucrose is

  • 254 associated with the presence of Fructobacillus and Leuconostoc genera, because those

  • 255 microorganism present the genes that codes for the invertase β -fructofuranosidase

  • 256 (Supplementary Figure 3S), a further study is needed to show yeast abundance in Tuba

  • 257 samples and its implication in sucrose hydrolysis.

  • 258 In the laboratory controlled fermentations we observed near complete sucrose hydrolysis

  • 259 and lower ethanol production compared to other fermented beverages such as pulque.

  • 260 Pulque shown an absence of sucrose hydrolysis, high ethanol concentration and a high

  • 261 abundance of ethanolic bacteria such as Zymomonas mobilis (Escalante et al. 2008), that

  • 262 genus was not found in our work. Thus, these results suggest that the composition of the

  • 263 bacterial community in Tuba play an important role in the hydrolysis of sucrose at the start

  • 264 of the fermentation. These characteristics are related with the bacterial diversity, because

  • 265 several bacterial genera present in the sap has different metabolism and regulation types

  • 266 that in consequence may inhibit or delay the fermentative process (Tamang et al., 2016).

  • 267 We found the 10 more abundant bacterial genera that belong to three main groups, lactic

  • 268 acid bacteria (LABs), acetic acid bacteria (AABs) and proteobacteria (Figure 3). It has been

  • 269 reported that LABs are the main antibiotic and folate producers in fermented products (De

  • 270 la Fuente-Salcido et al. 2015; Rossi et al. 2011) both functions have an important impact on

  • 271 human health. Moreover, some LABs reported in here such as Fructobacillus and

  • 272 Leuconostoc genera are similar phylogenetic and metabolically, nevertheless,

  • 273 Fructobacillus is unable to produce ethanol, redirecting the carbon flow to the production

  • 274 of lactate, (Endo et al. 2015). Other genera found was Lactococcus that produce more

  • 275 lactate than ethanol (Makarova et al. 2006).

  • 276 The acetate production is related with the abundance of AABs such as Gluconacetobacter

  • 277 and Acetobacter genera that was found in all samples. Interestingly, sample P1 showed the

  • 278 higher abundances of Acetobacter, which contributed with the acetate and ethanol

  • 279 production in comparison with the Tuba P2 and P3. Nevertheless, we observed a smaller

  • 280 abundance of the AABs in the commercial samples; contributing with a lower acetate and

  • 281 ethanol concentrations with respect to the laboratory fermented samples. This result is in

  • 282 agreement with other studies, where the authors propose that the growth of the AABs of the

  • 283 Gluconacetobacter and Acetobacter genera is dependent on the presence of acetate and

  • 284 ethanol in the environment (Lisdiyanti et al. 2003). Other researches have stablished these

  • 285 genera as the main acetate producers in products from fruit fermentation (Dellaglio et al.

286 2017).

  • 287 Both Vibrio genera and Enterobacteriaceae family (both proteobacterias) were detected in

  • 288 all the analyzed Tuba samples. Vibrio have been reported as a "natural" pollutant of

  • 289 fermented products (Lee et al. 2015). The abundances of these bacterial groups was

  • 290 reduced through the Tuba fermentation process, this abundance in commercial Tuba

  • 291 samples was similar with the abundance to the initial fermentation points. Finally, we

  • 292 observed a relation between the increase of the abundance of LABs and the decrease of

  • 293 Enterobacteriaceae family. It has been shown that the secretion of peptidases by LAB and

  • 294 AAB limits the cell growth of pathogen such as Vibrio (De la Fuente-Salcido et al. 2015;

  • 295 Lee et al. 2015). Hence, in this case the limitation of the growth of some proteobacteria in

  • 296 the Tuba, may be caused by compounds produced by the bacterial community (such as the

  • 297 peptidases).

  • 298 The alpha diversity tests, showed that in Chao1 and Observed_otus the Tuba P2 had

  • 299 significant difference but with Shannon and Simpsons index all Tuba samples (P1, P2, P3

  • 300 and commercial) showed significant difference. That difference was due to the focus of

  • 301 each test, Observed_otus and Chao1 had low values for Tuba P2 that means low number of

  • 302 bacterial genus and high dominance. Although, Shannon and Simpson index analyzed the

  • 303 abundance and equity of the population, which means that the four Tuba groups have the

  • 304 same 10 genera but in different abundance (Figure 3). The low values of Tuba P2 in alpha

  • 305 tests may have related to high concentrations of sucrose and low acetate and ethanol. In a

  • 306 study of the bacterial diversity in pulque was established that the diversity is strongly

  • 307 correlated with ethanolic fermentation conditions and aguamiel and pulque composition

  • 308 (Escalante et al. 2008). The microbial beta diversity data showed no significant differences

  • 309 between the samples of each palm. Hence, the 10 most abundant genera of the 16 analyzed

  • 310 samples were associated with the origin of the samples. In some studies, the biotic and

  • 311 abiotic conditions (seasonality, plant physiology, age, soil conditions, and other abiotic

  • 312 variables such as water irrigation and other environmental factors) affected the bacterial

  • 313 diversity at different times of the fermentation (Staley et al. 2014; Fonseca-García et al.

  • 314 2016; Coleman-Derr et al. 2016). Hence, we propose that the sugar concentration and the

  • 315 pH of the Tuba, has an effect on the bacterial diversity of this beverage, contributing to

  • 316 define the metabolic composition and the dominant bacterial genera. Additionally, the sap

  • 317 samples were collected after it was harvested by the producer and we took all the

  • 318 precautions to conserve the initial bacterial community and took it to the laboratory for

  • 319 fermentation (all handling was done in aseptic conditions). Therefore the observed

  • 320 differences in bacterial diversity in the samples is a combination of the palm related abiotic

  • 321 variables and the harvesting procedure itself.

  • 322 Palm wine is consumed in many places in the world, the Tuba type that is the subject of this

  • 323 study has its own characteristics. It is produced near coconut palm production sites in the

  • 324 Mexican south pacific coast and is consumed as refreshing, hydration drink and empirically

  • 325 used as traditional aid for gastrointestinal discomfort, here we are showing its low alcohol

  • 326 content, however it can reach higher concentrations if fermented for longer time

  • 327 (Velázquez-Monreal et al. 2011). In this study, the functional analysis of the Tuba P1, P2,

  • 328 P3 and commercial samples using PICRUSt showed 18 functions of biotechnological

  • 329 interest (Figure 6), some of them showed significant differences as folate biosynthesis,

  • 330 antibiotic production and peptidases. Other functions were present on all Tuba samples

  • 331 without a significant difference among them such as terpene and carotenoid biosynthesis.

  • 332 These functions have been described in other fermented beverages such as pulque, where

  • 333 they proposed it as a functional product because it has mainly prebiotic and probiotic action

  • 334 with antimicrobial activity and production of nutrients (Escalante et al. 2016). In other

  • 335 study Cocos nucifera L. (Palmacea) water (CW), variety Chandrasankara, was tested for its

  • 336 ability to scavenge free radicals, and they found a good antioxidant activity percentage

  • 337 (Mantena et al. 2003). Beverages made from plants, seeds or fruits have high contents of

  • 338 phenolic compounds that have the capacity to stabilize reactive oxygen and nitrogen

  • 339 species (Richelle et al. 2001), especially red, pink and white color fermented beverages

  • 340 (Martins de Sá et al. 2014). In addition, as we found in this work (Figure 7), microbial

  • 341 communities may be able to produce antioxidant compounds, there are evidence that

  • 342 described LABs genera as antioxidant compound producers, mainly glutathione, folate and

  • 343 butyrate (Wang et al. 2017). Other studies reported bacteria that produce antioxidant

  • 344 compound but it was not been identified yet (Tabbene et al. 2010), or it is a pigment

  • 345 produced in specific conditions by the bacteria (Radhakrishnan et al. 2016).

  • 346 In this work we reported for the first time the bacterial diversity and potential functional

  • 347 analysis through the fermentation process of the Tuba. With the knowledge of microbiota

  • 348 diversity and metabolic functional inference, the Tuba production can be controlled

  • 349 adjusting the presence and abundance of beneficial genera that contributes with the

  • 350 functional characteristics of the Tuba. It also contributed to the stablishing of

  • 351 microbiological basis of its empirical uses. Additionally, the bacterial isolation from these

  • 352 samples may provide us with new species with probiotic potential.




353 References.

  1. 354 Caporaso, J.G. et al., 2011. QIIME allows analysis of high-throughput community
  2. 355 sequencing data. Nature America, 7(5), pp.335–336.
  3. 356 Coleman-Derr, D. et al., 2016. Plant compartment and biogeography affect microbiome
  4. 357 composition in cultivated and native Agave species. New Phytologist , 209, pp.798–
  5. 359 811.
  6. 360 Creer, S. et al., 2016. The ecologist’s field guide to sequence-based identification of
  7. 361 biodiversity. Methods in Ecology and Evolution, 7, pp.1008–1018.
  8. 362 De la Fuente-Salcido, N.M. et al., 2015. Isolation and characterization of bacteriocinogenic
  9. 363 lactic bacteria from M-Tuba and Tepache, two traditional fermented beverages in
  10. 364 México. Food Science & Nutrition, 3(5), pp.434–442. Available at:
  11. 365
    http://doi.wiley.com/10.1002/fsn3.236
    .
  12. 366 Dellaglio, F. et al., 2017. Description of Gluconacetobacter swingsii sp. nov. and
  13. 367 Gluconacetobacter rhaeticus sp. nov., isolated from Italian apple fruit. International
  14. 368 Journal of Systematic and Evolutionary Microbiology , 55(2005), pp.2365–2370.
  15. 369 Díaz-Ruíz, G. et al., 2003. Microbial and Physiological Characterization of Weakly
  16. 370 Amylolytic but Fast-Growing Lactic Acid Bacteria : a Functional Role in Supporting
  17. 371 Microbial Diversity in Pozol , a Mexican Fermented Maize Beverage. Applied and
  18. 372 Enviromental Microbiology , 69(8), pp.4367–4374.
  19. 373 Endo, A. et al., 2015. Comparative genomics of Fructobacillus spp . and Leuconostoc spp .
  20. 374 reveals niche- specific evolution of Fructobacillus spp . BMC Genomics, 16(1117).
  21. 375 Available a t:
    http://dx.doi.org/10.1186/s12864-015-2339-x
    .
  22. 376 Escalante, A. et al., 2008. Analysis of bacterial community during the fermentation of
  23. 377 pulque, a traditional Mexican alcoholic beverage, using a polyphasic approach.
  24. 378 International Journal of Food Microbiology , 124(2), pp.126–134.
  25. 379 Escalante, A. et al., 2016. Pulque, a Traditional Mexican Alcoholic Fermented Beverage :
  26. 380 Historical , Microbiological , and Technical Aspects. frontiers in Microbiology, 7,
  27. 381 pp.1–18.
  28. 382 Escalante, A.E. et al., 2015. Ecological perspectives on synthetic biology : insights from
  29. 383 microbial population biology. frontiers in Microbiology, 6, pp.1–10.
  30. 384 Fernández & Galván, 2015. Métodos para la cuantificación de proteínas. Aviable at:
  31. 385
    http://sgpwe.izt.uam.mx/files/users/uami/acym/27_metodos_para_la_cuantificacion_d
  32. 386 e_proteinas.pdf.
  33. 387 Filippis, F. De et al., 2017. A comparison of bioinformatic approaches for 16S rRNA gene
  34. 388 pro fi ling of food bacterial microbiota. International Journal of Food Microbiology,
  35. 389 265, pp.9–17. Available a t:
    http://dx.doi.org/10.1016/j.ijfoodmicro.2017.10.028
    .
  36. 390 Fonseca-García, C. et al., 2016. The Cacti Microbiome : Interplay between Habitat-
  37. 391 Filtering and. frontiers in Microbiology, 7, pp.1–16.
  38. 392 Goosen, C. et al., 2007. Molecular and Biochemical Characterization of a Novel
  39. 393 Intracellular Invertase from Aspergillus niger with Transfructosylating Activity. , 6(4),
  40. 394 pp.674–681.
  41. 395 Hsieh, C.-W. et al., 2006. Molecular Cloning and Functional Identification of Invertase
  42. 396 Isozymes from Green Bamboo Bambusa oldhamii. Agricultural and Food Chemestry,
  43. 397 54, pp.3101–3107.
  44. 398 Kuczynski, J. et al., 2012. Using QIIME to analyze 16s rRNA gene sequences from
  45. 399 microbial communities. Current Protocols in Microbiology, pp.1–28.
  46. 400 Langille, M.G.I. et al., 2013. Predictive functional profiling of microbial communities
  47. 401 using 16S rRNA marker gene sequences. Nature Biotechnology, 31(9), pp.814–821.
  48. 402 Available a t:
    http://dx.doi.org/10.1038/nbt.2676
    .
  49. 403 Lee, S.H., Jung, J.Y. & Jeon, C.O., 2015. Bacterial community dynamics and metabolite
  50. 404 changes in myeolchi-aekjeot , a Korean traditional fermented fi sh sauce, during
  51. 405 fermentation. International Journal of Food Microbiology, 203, pp.15–22. Available
  52. 406 a t:
    http://dx.doi.org/10.1016/j.ijfoodmicro.2015.02.031
    .
  53. 407 Lisdiyanti, P. et al., 2003. Diversity of Acetic Acid Bacteria in Indonesia, Thailand, and the
  54. 408 Philippines. Microbiology and Culture Collections, 19(2), pp.91–99.
  55. 409 Makarova, K. et al., 2006. Comparative genomics of the lactic acid bacteria. Proceedings of
  56. 410 the National Academy of Sciences , 103(42), pp.15611–15616. Available at:
  57. 411
    www.pnas.org/cgi/doi/10.1073/pnas.0607117103
    .
  58. 412 Mantena, S.K. et al., 2003. In vitro evaluation of antioxidant properties of Cocos nucifera.
  59. 413 Nahrung /Food , 47(2), pp.126–131.
  60. 414 Martins de Sá, L.Z.C. et al., 2014. Antioxidant potential and vasodilatory activity of
  61. 415 fermented beverages of jabuticaba berry ( Myrciaria jaboticaba ). Journal of
  62. 416 Functional Foods , 8, pp.169–179. Available at:
  63. 417
    http://dx.doi.org/10.1016/j.jff.2014.03.009
    .
  64. 418 Navas-Molina, J.A. et al., 2015. Advancing our understanding of the human microbiome
  65. 419 using QIIME. Methods Enzymol, 531(2013), pp.371–444.
  66. 420 Radhakrishnan, M. et al., 2016. In vitro antioxidant activity and antimicrobial activity
  67. 421 against biofilm forming bacteria by the pigment from Desert soil Streptomyces sp
  68. 422 D25. , 6(6), pp.148–150.
  69. 423 Richelle, M., Tavazzi, I. & Offord, E., 2001. Comparison of the Antioxidant Activity of
  70. 424 Commonly Consumed Polyphenolic Beverages (Coffee, Cocoa, and Tea) Prepared per
  71. 425 Cup Serving. Journal of Agricultural and Food Chemistry, 49(7), pp.3438–3442.
  72. 426 Rossi, M., Amaretti, A. & Raimondi, S., 2011. Folate Production by Probiotic Bacteria.
  73. 427 nutrients , 3, pp.118–134.
  74. 428 Santiago-Urbina, J.A. & Ruíz-Terán, F., 2014. Microbiology and biochemistry of
  75. 429 traditional palm wine produced around the world. International Food Research
  76. 430 Journal , 21(4), pp.1261–1269.
  77. 431 Staley, C. et al., 2014. Core functional traits of bacterial communities in the Upper
  78. 432 Mississippi River show limited variation in response to land cover. frontiers in
  79. 433 Microbiology , 5(414), pp.1–11.
  80. 434 Tabbene, O. et al., 2010. A new antibacterial and antioxidant S07-2 compound produced by
  81. 435 Bacillus subtilis B38. FEMS Microbiology Letters, 303, pp.176–182.
  82. 436 Tamang, J.P., Watanabe, K. & Holzapfel, W.H., 2016. Review : Diversity of
  83. 437 Microorganisms in Global Fermented Foods and Beverages. frontiers in Microbiology,
  84. 438 7, pp.1–28.
  85. 439 Tang, J. et al., 2017. Analysis of the Bacterial Communities in Two Liquors of Soy Sauce
  86. 440 Aroma as Revealed by High-Throughput Sequencing of the 16S rRNA V4
  87. 441 Hypervariable Region. BioMed Research International, 2017, pp.1–9.
  88. 442 Velázquez-Monreal, J. et al., 2011. Tuba: una bebida fermentada tradicional de Colima.
  89. 443 Cienciacierta 25:11–14.
  90. 444 Wang, Y. et al., 2017. Antioxidant Properties of Probiotic Bacteria. nutrients, 9(521), pp.1–
  91. 445 15.
  92. 446 Zhang, J. et al., 2016. Metagenomic approach reveals microbial diversity and predictive
  93. 447 microbial metabolic pathways in Yucha , a traditional Li fermented food. Scientific
  94. 448 Reports , 6, pp.1–9.

449

450

451 Figures.

452

453 Figure 1. Sample collection. A) Sampling sites. The yellow stars represent the location of

454 the four commercial establishments (commercial samples) and the blue star show the area

455 where sap samples for the laboratory controlled fermentation were obtained. B) Cocos

456 nucifera L (palm tree). Yellow square signaling sap collection zone. C) Sap collection zone.

457 Red arrow indicate the palm structure (inflorescence) where the sap is collected.

458

459 Figure 2. Metabolic composition of the laboratory fermented Tuba and commercial

460 Tuba. A) Tuba P1, B) Tuba P2, C) Tuba P3 and D) Commercial Tuba samples. Each

461 number correspond to one sample. Right axis represented pH value.

462

463

464 Figure 3. Taxonomic identification. The graph represented the top ten of genera using

465 0.01% abundance filter OTUs table.

466

467

468 Figure 4. Beta diversity. A) Associate with respect to the fermentation time. B) Associate

469 with respect to the origin of the sample. Analysis performed by the Unifrac unweighted

470 technique with 0.01% abundance filter and plotted with the Principal Coordinates Analysis

471 (PCoA). The color boxes show a grouping data.

472

473 Figure 5. Abundance of sequences associate with functions. An ANOVA was performed

474 with Tukey-Kramer (0.95), the percentage of genes associated with functions, discarding

475 elementary cellular functions. Asterisk show functions with significant difference (p<0.05).

476

477

478 Figure 6. Box plot of two functions of interest. With A) not significant difference and B)

479 significant difference. An ANOVA was performed with Tukey-Kramer (0.95) and plotted

480 with STAMP.

481

482

483 Figure 7. Main bacteria with 15-cis-phytoene synthase gene (K02291 KEGG code).

484 Analysis performed with the function “metagenome_contributions.py” obtained by

485 PICRUSt analysis and plotted with R studio.

486

487

488 Tables.

489 Table 1. Climatological conditions of the study sites.

PARAMETER PALM SAP COLLECT COMMERCIAL COLLECT
Date 14/07/2016 16/08/2016
Coordinates North 16 ° 46'54.53 '' West 99 ° 47'02.73 ''
Altitude 12
Temperature Max. 32°C y Min. 24°C Max. 30°C y Min. 24°C
Humidity 87% 89%
Pressure 0.996 atm 0.996 atm
Weather Light rain Light rain

490 Data provided by Comisión Nacional del Agua (CONAGUA).

491

492 Table 2. PCR primers targeting 16S rRNA V3-V4 region of bacteria

Amplicon size: 550 bp
Forward 5'- TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGN GGCWGCAG-3’
Reverse 5' GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHV GGGTATCTAATCC-3’.

493

494 Table 3. Alpha diversity. Asterisk show significant difference with p<0.05.

495

496

497 Supplementary material.

498

499

500 Figure 1S. Taxonomic identification. Most abundant OTU’s using the 0.01% abundance

501 filter OTUs table.

502

503

504 Figure 2S. Alpha diversity rarefaction plots with 0.01%. A) Observed_otus, B) Chao1,

505 C) Shannon and D) Simpson. Each population is represented for a specific color in all the

506 graphics.

507

508

509 Figure 3S. Abundance of invertase gene (K01193). Analysis performed with the function

510 “metagenome_contributions.py” obtained by PICRUSt analysis and plotted with R studio.

511

512 Table 1S. Chemical composition of the Tuba.

Tuba P1 Tuba P2 Tuba P3 Commercial L1 Commercial L2 Commercial L3 Commercial L4
Time (h) Proteins (g/L)
0 0.0068 0.0115 0.0097
12 0.0094 0.0038 0.0109 0.0087 0.0173 0.0166 0.0054
24 0.0057 0.0084 0.0046
35 0.0082 0.0147 0.0042
Time (h) Sucrose (g/L)
0 85.1478 121.7608 95.2033
12 0.3932 113.2938 19.8372 45.3103 13.4554 99.2911 4.3160
24 0.3748 102.1780 0.8837
35 0.2629 4.2062 0.7476
Time (h) Glucose (g/L)
0 21.1467 13.1798 14.6654
12 61.4149 3.4526 39.8145 59.2407 39.8149 35.0787 51.8465
24 60.1442 21.6494 7.2646
35 21.0363 59.0502 0.8876
Time (h) Fructose (g/L)
0 16.0169 11.0782 11.9225
12 47.3585 1.9186 29.1776 50.8742 45.5284 32.7653 40.9536
24 48.5620 13.0289 39.7342
35 44.8364 45.6336 28.1220
Time (h) Acetate (g/L)
0 0.6867 0.8935 0.6452
12 3.0672 1.4066 1.1047 1.5209 1.6449 1.5271 1.9845
24 6.9141 2.7640 2.8354
35 6.0741 3.5594 3.9577
Time (h) Ethanol (g/L)
0 0.4116 0.1669 0.2528
12 2.2396 0.0000 1.1756 1.5176 1.4198 0.7016 4.4201
24 4.4278 0.0000 41.3655
35 47.1141 6.8062 50.6929
Time (h) pH
0 3.73 3.65 3.64
4 3.34 3.38 3.3
5 3.29 3.28 3.26
6 3.28 3.28 3.25
10 3.15 3.21 3.12 4.0 4.0 4.0 4.0
11 3.12 3.19 3.11
12 3.1 3.19 3.11
23 2.97 3.13 2.97
24 2.93 3.12 2.98
35 2.94 2.81 2.94

513