Accepted Manuscript

Sourdough microbial community dynamics: an analysis during French organic breadmaking processes
Emilie Lhomme , Charlotte Urien , Judith Legrand , Xavier Dousset , Bernard Onno ,
Delphine Sicard
PII:

S0740-0020(14)00292-5

DOI:

10.1016/j.fm.2014.11.014

Reference:

YFMIC 2304

To appear in:

Food Microbiology

Received Date: 8 September 2014

Revised Date:

20 November 2014

Accepted Date: 28 November 2014

Please cite this article as: Lhomme, E., Urien, C., Legrand, J., Dousset, X., Onno, B., Sicard, D.,
Sourdough microbial community dynamics: an analysis during French organic bread-making processes,
Food Microbiology (2015), doi: 10.1016/j.fm.2014.11.014.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Sourdough microbial community dynamics: an analysis during French organic bread-

making processes

Emilie Lhomme1 #, Charlotte Urien2 #, Judith Legrand3, Xavier Dousset4, Bernard Onno1* and

Delphine Sicard3, 5

6
#

These authors contributed equally to this work.

LUNAM Universit Oniris, Laboratoire de Microbiologie Alimentaire et Industrielle, rue de

SC

M
AN
U

8
9

RI
PT

10

la Graudire, Nantes Cedex 3, France.

11

INRA, UMR 0320/UMR 8120 Gntique Vgtale, F-91190 Gif-sur-Yvette, France.

12

Univ Paris-Sud, UMR 0320/UMR 8120 Gntique Vgtale, F-91190 Gif-sur-Yvette,

13

France.

14

15

Nantes Cedex 3, France.

16

TE
D

LUNAM Universit ONIRIS, UMR INRA 1014 Secalim, rue de la Graudire, BP 88225

INRA, UMR 1083 Sciences pour lnologie, 34060 Montpellier Cedex 2, France.

EP

17

* Corresponding author: Bernard Onno

19

E-mail: bernard.onno@oniris-nantes.fr

20

Tel: +33 (0)251785535

21

Fax: + 33 (0)251785520

22
23

Running title: Microbial community dynamics in French sourdoughs

AC
C

18

ACCEPTED MANUSCRIPT

Abstract

25

Natural sourdoughs are commonly used in bread-making processes, especially for organic

26

bread. Despite its role in bread flavor and dough rise, the stability of the sourdough microbial

27

community during and between bread-making processes is debated. We investigated the

28

dynamics of lactic acid bacteria (LAB) and yeast communities in traditional organic

29

sourdoughs of five French bakeries during the bread-making process and several months apart

30

using classical and molecular microbiology techniques. Sourdoughs were sampled at four

31

steps of the bread-making process with repetition. The analysis of microbial density over 68

32

sourdough /dough samples revealed that both LAB and yeast counts changed along the bread-

33

making process and between bread-making runs. The species composition was less variable.

34

A total of six LAB and nine yeast species were identified from 520 and 1,675 isolates,

35

respectively. The dominant LAB species was Lactobacillus sanfranciscensis, found for all

36

bakeries and each bread-making run. The dominant yeast species changed only once between

37

bread-making processes but differed between bakeries. They mostly belonged to the

38

Kazachstania clade. Overall, this study highlights the change of population density within the

39

bread making process and between bread-making runs and the relative stability of the

40

sourdough species community during bread-making process.

SC

M
AN
U

TE
D

EP

42

AC
C

41

RI
PT

24

43

Keywords: lactic acid bacteria, yeast, Lactobacillus sanfranciscensis, Kazachstania sp.,

44

Saccharomyces sp.

ACCEPTED MANUSCRIPT

1. Introduction

46

Sourdough is used as an alternative to bakers yeast to leave bread dough, although some

47

bakeries use both (De Vuyst and Neysens, 2005). In France, the regulations define

48

sourdough as dough made from wheat or rye () with added water and salt (optional),

49

containing a naturally acidifying microbiota made up primarily of lactic bacteria and

50

yeasts (Article 4 of Decree no. 93-1074 of September 13th 1993). In addition, the decree

51

specifies that sourdough bread has a potential maximum pH of 4.3 and an acetic acid

52

content of at least 900 ppm. Many studies have now demonstrated the nutritional,

53

sensory, texture and shelf-life advantages of using sourdough composed of lactic acid

54

bacteria (LAB) and yeasts (Galle and Arendt, 2013; Gnzle, 2014; Poutanen et al., 2009).

55

Among LAB, the genera Pediococcus, Leuconostoc, Weissella, and especially

56

Lactobacillus, have frequently been identified (De Vuyst et al., 2009; Hammes et al.,

57

2005; Huys et al., 2013). Among yeasts, the species Saccharomyces cerevisiae and

58

Candida humilis are those most commonly found (Iacumin et al., 2009; Minervini et al.,

59

2012a; Vernocchi et al., 2004). Some stable and specific associations between LAB and

60

yeasts, due to particular nutritional, trophic, and metabolic interactions, have been

61

described (Corsetti and Settanni, 2007; De Vuyst and Neysens, 2005; De Vuyst and

62

Vancanneyt, 2007; Gnzle et al., 2007; Gobbetti et al., 2005).

63

Recently, a number of studies have considered the influence of different factors on the

65
66

SC

M
AN
U

TE
D

EP

AC
C

64

RI
PT

45

microbial species composition of sourdough. For instance, the microbial community has
been found to change according to the wheat species (Minervini et al., 2012a), the cereals
used (Vogelmann et al., 2009), the temperature and the back-slopping time (Vrancken et

67

al., 2011a), the location of the propagation (artisan bakery or laboratory) (Minervini et al.,

68

2012b), and technological factors (Vogelmann and Hertel, 2011).

ACCEPTED MANUSCRIPT

Despite the well-known microbial composition of sourdoughs in Italy (Lattanzi et al.,

70

2013; Minervini et al., 2012a, Minervini et al., 2012b), Belgium (Scheirlinck et al., 2007,

71

2008), and Germany (Meroth et al., 2003a; Meroth et al., 2003b), few studies have

72

investigated both LAB and yeast sourdough diversity in France (Huys et al., 2013 and

73

references therein). In fact, previous works on French sourdough have considered only

74

LAB diversity (Ferchichi et al., 2007; Robert et al., 2009; Valcheva et al., 2005; Vera et

75

al., 2011).

76

Furthermore, to our knowledge, no previous study has investigated the dynamics of the

77

microbial community during the bread-making process except for sweet baked products

78

(Venturi et al., 2012). In addition, only few data are available on the long-term stability of

79

sourdough ecosystems in bakeries (Gnzle and Vogel, 2003; Rosenquist and Hansen,

80

2000; Scheirlinck et al., 2008). The comparison of the microbiota in one conventional and

81

one organic sourdough suggests that the genetic diversity is less stable in the organic

82

sourdough compared to the conventional one (Rosenquist and Hansen 2000). However,

83

the stability of the microbiota was never investigated further in organic sourdoughs.

84

The aims of this study were (i) to analyze the stability of LAB and yeast counts in French

85

organic sourdoughs along the bread-making process and between bread-making runs, (ii)

86

to study the relationships between microbial counts and dough biochemical properties,

87

and (iii) to characterize the LAB and yeast species composition throughout the bread-

89

SC

M
AN
U

TE
D

EP

AC
C

88

RI
PT

69

making processes and their long-term stability.

ACCEPTED MANUSCRIPT

2. Materials and methods

91

2.1 Sourdough and bread collection

92

A total of five bakeries located in distinct regions of France were selected (Supplementary

93

material S1), based on their choice to promote biodiversity conservation and to limit their

94

ecological footprint through the use of natural sourdough and organic flour. A semi-structured

95

interview to describe their bread-making practices was carried out among the five bakeries.

96

The ingredients and technological parameters involved are summarized in Supplementary

97

material S1. In all the bakeries, the chief sourdough was obtained by taking a piece of final

98

leavened dough or dough after kneading at each bread-making run. Bakers practices differed

99

from one bakery to another (Supplementary material S1). Although temperatures were

100

relatively the same in the bakeries, the number of back-sloppings before the final leavened

101

dough step varied between bakeries, from one (in Bakeries 4 and 5) to three back-sloppings

102

(in Bakery 2). Some other specificities might be noticed: a firm sourdough recipe in Bakery 5,

103

a variation of the back-slopping period (short for Bakeries 1 and 2 (about 6h)) and the

104

percentage of sourdough added in final dough (higher for Bakeries 1 and 4 (>30%)).

105

Samples were taken at four steps during the bread-making process: chief sourdough (CS),

106

final leavened dough (LD), dough after kneading (AK) and bread before baking (BB)

107

(Supplementary material S2). Two independent samples were taken at each bread-making

108

process step except for Bakery 1 and for the first bread-making run sampling for Bakeries 2

109

and 3. For each bakery, the sampling was repeated during two different bread-making runs,

110

seven to ten months apart (Supplementary material S1). All samples were kept at 4 C for

111

three to seven hours before analysis. Then, LAB and yeast were enumerated for all collected

112

samples. Finally, LAB and yeast species were identified for all samples of the first collection

113

and for the final leavened dough samples of the second analyzed bread-making run.

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

90

114

ACCEPTED MANUSCRIPT

2.2 LAB and yeast enumeration and isolation

116

Sourdough samples (10 g) were homogenized by adding 90 ml of sterile tryptone salt (TS)

117

solution [0.85% (wt/vol) NaCl; 0.1% tryptone (wt/vol)] and mixing for 2 min using a

118

Stomacher (AES Laboratories, France). A 10-fold dilution series from 10-2 to 10-6 was made

119

and plated in triplicate using either a spiral plate method (Easyspiral, Intersciences, Saint

120

Nom, France) or a spreading plate method on MRS5 agar (Meroth et al., 2003b) and on Yeast

121

Extract (YE) media (yeast extract 1%, glucose 2%, agar 1.7%) to determine LAB and yeast

122

counts, respectively. LAB were counted after incubation for 48 h at 30 C under anaerobic

123

conditions (Anaerocult A, Merck, Darmstadt, Germany). Yeasts were enumerated after

124

incubation for 48 h to 72 h at room temperature under aerobic conditions.

125

For each sourdough sample, 30 LAB colonies and 40 yeast colonies were randomly picked

126

from the plates and isolated on MRS5 agar and YE agar, respectively. After catalase and Gram

127

tests, a total of 520 Gram-positive and catalase-negative LAB isolates were selected and

128

stored at -80 C in MRS5 medium containing glycerol (20%). A total of 1,675 yeast isolates

129

were stored at -80 C in YE medium containing glycerol (17.4%).

130

The strains used as references for LAB and yeast species identification are listed in

131

Supplementary material S3. The yeast type strains used in this study are preserved at CIRM-

132

Levures (http://www6.inra.fr/cirm/Levures).

SC

M
AN
U

TE
D

EP

AC
C

133

RI
PT

115

134

2.3 Determination of pH, titratable acidity and organic acids

135

Determination of pH, titratable acidity and organic acid content was carried out for each

136

sourdough and dough sample as well as for bread. Ten grams of sourdough, dough or bread

137

were mixed with 90 ml of TS solution in a Stomacher for 2 min. Then, 10 ml of the mixture

138

was homogenized and the pH and total titratable acidity (TTA) were measured with an

139

automatic titrator (pH-Matic 23, Crison). For D- and L-lactic and acetic acid measurements, 10

ACCEPTED MANUSCRIPT

ml of the homogenized mixture was centrifuged at 7, 000 g for 10 min at room temperature

141

and the supernatant was used for measurements using the EnzytecTM Kit as described in the

142

instruction manual, and by measuring absorbance at 340 nm (Genesys 10). Acid amounts

143

were expressed as g/kg of sourdough, dough or bread.

144

RI
PT

140

2.4 DNA extraction

146

The direct DNA extractions from sourdough or dough samples (30 g) were performed as

147

described previously (Jaffrs et al., 2009).

148

The chromosomal DNA of bacterial strains was extracted from the pellet of bacteria grown in

149

5 ml MRS5. Then, DNA purification was carried out using a Qiagen DNeasy Blood and

150

Tissue Kit, as described in the Qiagen instruction manual (Qiagen, SA, Courtaboeuf, France).

151

For DNA extraction from yeasts, strains were grown at 30 C in 1 ml of YE (1% YE, 2%

152

glucose) shaken at 200 rpm. After digestion of the pellet for one hour at 37 C using 10 units

153

of zymolyase (Euromedex, Souffelweyersheim, France) in 0.5 ml of sorbitol 1 M, Na2EDTA

154

0.1 M, pH 7.5 buffer, a classic method of DNA extraction was used in which the cell

155

membrane was broken with a detergent (SDS), deproteinization was carried out with

156

potassium acetate and finally DNA was purified on a Whatman unifilter (Whatman, Florham

157

Park, NJ). For 28 isolates coming from Bakery 2, DNA glass bead extraction was adapted

158

from Burke et al. (2000).

M
AN
U

TE
D

EP

AC
C

159

SC

145

160

2.5. LAB species identification

161

LAB isolates were identified by partial or whole 16S rDNA sequencing, PCR targeting some

162

specific genes, or by TTGE.

163

The 16S rDNA (about 1,500 bp) of the pure LAB isolates was amplified by PCR as described

164

previously (Jaffrs et al., 2009) from chromosomal DNA using primers fD1 and rD1

165

(Supplementary material S4), which anneal to positions 8 to 27 and to positions 1524 to 1541
7

ACCEPTED MANUSCRIPT

respectively, of the 16S rRNA gene, E. coli numbering (GenBank accession number J01695)

167

(Weisburg et al., 1991).

168

The sequencing primer SP1 (position 338 forward), SP2 (position 514 reverse), SP3 (position

169

810 forward), and SP5 (position 1089 reverse) targeted to conserved regions of the 16S rRNA

170

gene, E. coli (GenBank J01695) (Supplementary material S4). Partial nucleotide sequence

171

(about 700 bp) of the amplified 16S rDNA gene was determined using the sequencing primer

172

SP1. The whole 16S rDNA gene was sequenced (about 1,500 bp) using sequencing primers

173

SP. DNA sequences were determined using the Beckman Coulter Genomics service, (Takeley,

174

UK). The overlapping double strand sequences obtained with SP1 and SP2 or with SP3 and

175

SP5 were merged using the BioEdit program. Identification queries were fulfilled by a

176

BLAST search in the National Center for Biotechnology Information database (NCBI,

177

Bethesda, USA). A similarity of at least 97.6% to 16S rDNA gene sequences of type strains

178

was used as the criterion for identification (Stackebrandt and Ebers, 2006).

179

As the 16S rDNA sequence did not allow discriminating Lactobacillus plantarum,

180

Lactobacillus paraplantarum and Lactobacillus pentosus, a multiplex PCR targeting the recA

181

gene (Supplementary material S4), as described by Torriani et al. (2001), was performed. In

182

addition, a PCR targeting the katA gene (407 bp) was used (Supplementary material S4), as

183

described by Ammor et al. (2005), to confirm the 16S rDNA identification of Lactobacillus

184

sakei species which is heme-dependent catalase (Chaillou et al., 2005). In the same way, a

185

PCR targeting the 16S 23S rDNA intergenic spacer region was used (Supplementary

186

material S4), as described by Ferchichi et al. (2008), to confirm the 16S rDNA identification

187

of Lactobacillus hammesii species. All PCR amplifications were checked by electrophoresis

188

and visualized by DNA gel staining (SYBR Safe, Invitrogen, Villebon-sur-Yvette, France).

189

For PCR-TTGE, primers V3P2 and V3P3-GC-Clamp (Supplementary material S4) were used

190

to amplify the V3 region ( 200 bp) of 16S rDNA as previously described (Jaffrs et al.,

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

166

ACCEPTED MANUSCRIPT

2009). Amplicons were visualized on a 1.5% agarose gel by staining with SYBR Safe. The

192

PCR products were subjected to a TTGE analysis, as described previously (Jaffrs et al.,

193

2009). After the run, gels were stained for 30 min in 300 ml of SYBR Safe 1 X then rinsed in

194

Milli-Q water and visualized by UV illumination.

RI
PT

191

195

2.6 Yeast species identification

197

Yeast species were identified on the basis of three molecular markers: the AluI digestion

198

profile of the amplified rDNA Non-Transcribed Spacer, NTS2 (Nguyen and Gaillardin, 1997),

199

the sequence of the rDNA D1/D2 region (Kurtzman and Robnett, 2003), and the sequence of

200

the actin encoding gene, ACT1 (Daniel and Meyer, 2003; Kan, 1993). All the oligonucleotide

201

PCR primers used were obtained from Invitrogen (Invitrogen, Cergy Pontoise, France) and

202

are listed in Supplementary material S4.

203

Species of all yeast isolates were characterized on the basis of their pattern of AluI digestion

204

of the amplified NTS2 region. For each sample, PCR was carried out according to Nguyen et

205

al. (2009) using primers LR13 and SR21 and home-made Taq polymerase. Then, 5 l of PCR

206

product was digested using 2.5 units of AluI restriction enzyme (New England Biolabs,

207

Ipswich, England) for 1 h at 37 C using the manufacturers recommended procedure.

208

Digestion profiles were visualized on a 1.2% agarose gel. For each bakery and each bread-

209

making run, from one to three representatives of each NTS2-AluI digestion profile were

210

selected for further characterization of the species by D1/D2 and ACT1 sequencing leading to

211

a total of 55 isolates.

212

Species identification was extended by sequencing the D1/D2 rDNA region (Kurtzman and

213

Robnett, 2003). PCR amplifications were run in 50 l containing 2 mM of each dNTP, 0.5

214

M of each primer NL1 and NL4 (ODonnell, 1993), 2 mM of MgCl2, 1.25 units of GoTaq

215

DNA polymerase (Promega, Madison, USA), 1 X Taq PCR buffer [80 mM Tris-HCl [pH 8.8],

AC
C

EP

TE
D

M
AN
U

SC

196

ACCEPTED MANUSCRIPT

0.05% (v/v)] Tween-20, 335 mM [(NH4)2SO4] and 25 to 125 ng of template DNA from pure

217

cells. After an initial denaturation at 94 C for 4 min, 30 cycles of amplification were

218

conducted as follows: denaturation at 94 C for 30 sec, annealing at 54 C for 40 sec, and

219

extension at 72 C for 1 min and 30 sec. Amplicons were visualized on a 0.8% agarose gel.

220

Finally, 20 l of PCR products were sent for sequencing of both strands (Genoscreen, Lille,

221

France).

222

Species identification was completed for 55 strains by sequencing ACT1 gene (Daniel and

223

Meyer, 2003; Kan, 1993). PCR amplifications were run in 25 l containing 0.2 mM of each

224

dNTP, 0.2 M of each primer CA14 and CA5R (Kan, 1993), 1 unit of Takara Ex Taq DNA

225

polymerase (Takara Bio Inc., Shiga, Japan), 1 X Ex Taq Takara buffer and 25 ng of template

226

DNA from pure cells. After an initial denaturation at 95 C for 3 min, 30 cycles of

227

amplification were conducted as follows: denaturation at 95 C for 30 sec, annealing at 54 C

228

for 40 sec, and extension at 72 C for 1 min. The final extension was at 72 C for 7 min. PCR

229

products were Sanger sequenced on both strands (Genoscreen, Lille, France). Amplicons were

230

visualized on a 0.8% agarose gel.

231

DNA sequence alignments were carried out as described above. Identification queries were

232

fulfilled by a BLAST search in the NCBI (see above) and the YeastIP databases

233

(http://genome.jouy.inra.fr/yeastip/). A similarity of more than 99% to gene sequences of type

234

strains was used as the criterion for identification.

SC

M
AN
U

TE
D

EP

AC
C

235

RI
PT

216

236

2.7 Microbial count analysis

237

We tested whether LAB and yeast counts varied according to the following co-factors:

238

bakeries, bread-making steps of the process, bread-making runs, and within the dough.

239

Because the samples from three bakeries were plated with the spiral plate method while those

240

from the other two bakeries were plated with the spreading plate method, the bakery effect

10

ACCEPTED MANUSCRIPT

and the enumeration method (plating or spiral) effect were partly mixed up. Hence, to avoid

242

an overestimation of the bakery effect, the plating method was accounted for in the model.

243

In order to test the effect of various co-factors on the variations in LAB and yeast counts, a

244

linear mixed effect model was built as described below:

RI
PT

241

where is the log-transformed LAB (yeast, respectively) count n (n = 1; 3) of the

246

dough sample m (m = 1; 2) carried out at the process step j (j = 1; 4) during the bread-making

247

run l (l = 1; 2) in the bakery k (k = 1; 5) measured by the colony plating method i (i = 1; 2).

248

The colony plating method and the process step were modeled as fixed effects (i and j,

249

respectively). The bakery, the bread-making run and the dough replicate were modeled as

250

Gaussian random effects (Bk, Dl (k) and Rm(kl), respectively) with variance denoted by B, D

251

and R, respectively. The residuals ijklmn were assumed independent and identically

252

distributed according to a normal distribution (, ).

253

Fixed and random effects were tested using likelihood-ratio tests with a 5% significance level.

254

For random effect tests, p-values were estimated by simulating the likelihood-ratio under the

255

null hypothesis. To compare the means of counts at the different process steps, Tukey tests

256

were performed. REML algorithms were used for estimates while ML algorithms were used

257

for tests. Statistical analyses were performed with the nlme and multcomp R packages

258

(http://www.r-project.org/).

M
AN
U

TE
D

EP

AC
C

259

SC

245

260

2.8 Species diversity analysis

261

The species diversity was estimated using the observed richness and three diversity indexes

262

(Marcon and Morneau, 2010), which measure the local diversity in a delimited system: the

263

Shannon index (Shannon and Weaver, 1949), which takes into account the number of species

264

and their frequency, the Simpson index (Simpson, 1949), which characterizes the probability

11

ACCEPTED MANUSCRIPT

265

of two randomly chosen individuals belonging to different species in a community, and the

266

Chao1 richness (Chao, 1984), which takes into account the rarefaction of species.

267

2.9 Multivariate analysis of microorganism counts and dough biochemical properties

269

A Principal Component Analysis (PCA) was performed in order to highlight the relationships

270

between microorganism counts and biochemical properties of dough. All biochemical

271

variables (pH, lactic acid and acetic acid contents and TTA), yeast and LAB average counts

272

and the LAB:yeast ratio for each dough replicate were analyzed (Supplementary material S5).

273

We tested whether pH and TTA were related to the average LAB (or yeast) count within a

274

bread-making run using simple linear models including the bread making-run and the counts

275

as co-factors.

M
AN
U

SC

RI
PT

268

276

3. Results

278

3.1. Dynamics of LAB and yeast counts

279

Over all the dough samples, the LAB count varied from 7 to 9.7 log10 CFU/g (Figure 1). The

280

bakery effect on the LAB count was not significant (p-value (P) = 0.33, Supplementary

281

material S6). By contrast, the LAB counts depended on the bread-making process steps (P <

282

10-4, Supplementary material S6, and Figure 1). In fact, on average, the LAB count was lower

283

in the dough after kneading than in the other time points (P < 10-3). Variations in LAB counts

284

were also explained by significant variations between bread-making runs (Standard Deviation

285

(SD) =0.59, CI 95% = [0.31; 1.11]) and, to a lesser extent, between dough replicates (SD =

286

0.20, CI 95% [0.16; 0.25]).

287

Over all the samples, the yeast count ranged from 3.2 to 7.6 log10 CFU/g (Figure 1). As for

288

LAB, the bakery effect was not significant (P = 0.21, Supplementary material S6) but the

289

yeast count varied significantly during the bread-making process (P < 10-4, Supplementary

AC
C

EP

TE
D

277

12

ACCEPTED MANUSCRIPT

material S6). In fact, the yeast count was lower in the dough after kneading than in the other

291

time points (P < 10-3). Variations in yeast counts were also explained by significant variations

292

between bread-making runs (SD = 0.68, CI 95% = [0.36; 1.27]) and, to a lesser extent,

293

between dough replicates (SD = 0.22, CI 95% = [0.18; 0.27]).

RI
PT

290

294

3.2 Relationships between microorganism counts and biochemical parameters

296

Relationship between biochemical parameters of sourdoughs and communities of

297

microorganisms were analyzed by PCA (Figure 3). The first two axes of the PCA explained

298

69% of the total variance (39.31% and 29.42%, respectively, Figure 3A). TTA, pH,

299

LAB:yeast ratio and yeast count mainly contributed to the first axis. This axis separated

300

samples from different bread-making runs in the same bakery (symbols, Figure 3B). Three

301

main variables explained the second axis: LAB:yeast ratio, LAB and yeast counts. This axis

302

separated samples from different bakeries (different colors, Figure 3B). The samples carrying

303

the least common yeast and LAB species, T. delbrueckii (Bakery 1, first run, Table 1) and L.

304

plantarum (Bakery 3, first run, Table 1), respectively, were separated on the PCA map. As

305

seen previously with the statistical model, microbial density differed between bread-making

306

runs. Within a bread-making run, pH and TTA varied over time (Figure 3B and

307

Supplementary material S7). For some bread-making runs, pH and TTA decreased linearly

308

with LAB or yeast count (Supplementary material S8). During bread-making process, positive

309

correlations between LAB and yeast counts were observed, reflecting the growth of the

310

microbial population during the bread-making process due to the addition of resources (Figure

311

3B and Supplementary material S9).

AC
C

EP

TE
D

M
AN
U

SC

295

312
313
314

3.3 Species diversity of LAB

3.3.1 LAB species diversity

13

ACCEPTED MANUSCRIPT

The sequencing of 16S rRNA gene of 520 isolates allowed identifying six different species

316

(Table 1). The main one was Lactobacillus sanfranciscensis (79%) followed by L. plantarum

317

(12.8%). Lactobacillus kimchii (2.8%), L. sakei (2.7%), L. hammesii (2.3%) and Lactobacillus

318

pentosus (0.4%) were also found as minor species (Table 1).

319

While L. sanfranciscensis remained the only species identified for Bakeries 1 and 2 in the two

320

bread-making runs, changes between runs were observed for Bakeries 3, 4 and 5. For

321

example, for Bakery 3, L. plantarum was the dominant species during the first run, while the

322

dominance was shared with L. sanfranciscensis in the second run. For Bakery 4, L.

323

sanfranciscensis was the predominant species for the two runs but the minor species changed.

324

L. plantarum and L. hammesii, which were detected as minor species during the first run,

325

were no longer detected in the second run but L. sakei was found instead. Finally, for Bakery

326

5, L. sanfranciscensis was detected as the predominant species during the first run while in the

327

second one, L. sanfranciscensis, L. hammesii and L. kimchii were co-dominant, and L.

328

pentosus was found as a minor species (Table 1).

SC

M
AN
U

TE
D

329

RI
PT

315

3.3.2 Bacterial community dynamics analyzed by PCR-TTGE

In order to examine the bacterial population dynamics throughout the bread-making run,

331

fingerprints of the DNA samples of sourdoughs and doughs of each step of the two bread-

332

making runs were analyzed by PCR-TTGE. One isolate of each species, identified by 16S

333

rDNA sequencing, was used as the control (Figure 2).

334

Each bakery harbored a specific profile and this profile appeared to be clearly stable during

335

the process. For Bakeries 2, 3, and 4 the PCR-TTGE profiles were identical at the two

336

sampling periods, and comparable to those of the L. sanfranciscensis strains isolated from the

337

corresponding sourdoughs (Figure 2B, 2C, 2D). However, for Bakeries 1 and 5, an additional

338

band was observed (band a, Figure 2A and band b, Figure 2E, respectively) suggesting an

339

evolution of the bacterial community between the two sampling periods. In addition, for all

AC
C

EP

330

14

ACCEPTED MANUSCRIPT

340

the bakeries, the profiles of L. sanfranciscensis strains were different suggesting a large

341

intraspecific diversity.

342

3.4 Yeast species diversity

344

The AluI digestions of the NTS2 region were amplified from the 1,675 yeast strains and 9

345

yeast species were identified corresponding to nine AluI digestion patterns (Table 1). Among

346

the whole sample, 48% of the identified yeast species belonged to Kazachstania bulderi, 24%

347

was represented by C. humilis and the other species were Kazachstania unispora (11.4%),

348

Torulaspora delbrueckii (9.1%), Kazachstania exigua (5.9%), Rhodotorula mucilaginosa

349

(1.7%), Candida carpophila (0.2%), S. cerevisiae (0.1%) and Hyphopichia pseudoburtonii

350

(0.1%). One of these species, Kazachstania bulderi, has never been isolated in bakery

351

products before. No yeast species was shared by all bakeries (Table 1).

352

The dominant yeast species appeared to be stable during bread-making process and between

353

runs within the same bakery. The only exceptions were Bakery 2, where a change in the

354

dominant species appeared at the end of one of the bread-making process, and Bakery 1,

355

where the dominant species changed between runs. Only two (from Bakeries 1 and 5) out of

356

1,675 isolates belonged to the S. cerevisiae species, usually used as a leavening agent in

357

bakery products.

SC

M
AN
U

TE
D

EP

AC
C

358

RI
PT

343

359

3.5 Species diversity indexes

360

The number of yeast and LAB species (observed richness) varied from one to five per bakery.

361

Except for Bakery 1 for yeasts and Bakery 5 for LAB, where the Chao1 richness was

362

estimated at 5.5 and thus exceeded the 5 observed species indicating we may have missed

363

species, our sampling appears to give a good representation of the species diversity (Table 2).

364

Over all the -diversity indexes, a larger diversity was described within a bakery than

15

ACCEPTED MANUSCRIPT

between bakeries for LAB (Simpson index = 0 to 0.44 versus 0.36, Shannon index = 0 to 0.80

366

versus 0.76). By contrast, more diversity was observed between bakeries than within a bakery

367

for yeasts (Simpson index = 0.69 versus 0 to 0.48; Shannon index = 1.43 versus 0 to 0.78). It

368

appears that a high LAB diversity index is associated to a low yeast diversity index and

369

conversely (Table 2). A low LAB diversity index corresponded to L. sanfranciscensis alone

370

associated to different yeast species and high LAB diversity index characterized a microbiota

371

where one dominant yeast species was associated to several LAB species (Table 2).

SC

RI
PT

365

372

4. Discussion

374

The LAB and yeast counts varied extensively from 7 log10 to 9.7 log10 CFU/g and 3.2 log10 to

375

7.6 log10 CFU/g of dough, respectively and the LAB:yeast ratio ranged from 1:1 to 10,000:1

376

as previous found in sourdoughs (Scheirlink et al., 2007, Vrancken et al., 2010, Minervini et

377

al., 2012a, Lattanzi et al., 2013). The LAB and yeast counts changed significantly between

378

bread-making runs. This could not be easily explained by the bread-making practice as

379

practices did not change between runs except for Bakery 3, where rye and wheat flour were

380

used for the first bread-making run and only wheat flour for the second run. Thus, the origin

381

of the flour may make a difference although the number of LAB in rye sourdough is assumed

382

to be similar to that in wheat sourdough (Ercolini, 2013; Ercolini et al., 2013; Weckx et al.,

383

2010). Alternatively, external factors such as temperature fluctuation could change microbial

384

counts between bread-making runs (Kurtzman and Robnett, 2003; Meroth et al., 2003a;

385

Vogelmann and Hertel, 2011; Vrancken, et al., 2011b). Finally, the microbial composition of

386

the community may affect the microbial counts, as suggested by the low counts of yeasts in

387

presence of T. delbruecki observed for Bakery 1 and also in a Belgium sourdough (Scheirlinck

388

et al., 2007). The consequences of a low count are not well documented, but may not be a low

389

leavening ability (Reale et al., 2013).

AC
C

EP

TE
D

M
AN
U

373

16

ACCEPTED MANUSCRIPT

We also found count variations during the bread-making process but this variation was ten

391

times lower than between bread-making runs. For both LAB and yeasts, the counts after

392

kneading decreased significantly compared to the other steps, as expected after the addition of

393

flour and water during kneading. In addition, we found significant count variation between

394

replicates of the same dough. Despite sourdough mixing, this result suggests that sourdough is

395

not completely homogeneous. However, unlike for cheese (Ercolini et al., 2003), we found

396

that sample position in the dough affected the microbial count but not the microbial species

397

diversity.

398

The analysis of biochemical properties did not allow for detecting main factors influencing

399

dough properties. The acidity of the dough did not seem to be determined by microbial count

400

alone; it increased with microbial count during only some bread-making runs. Acidity

401

variation is probably a complex interplay between the addition of flour, which increases pH

402

and decreases TTA, and the fermentation process, which acidifies the dough due to the

403

microorganisms

404

Complementary phenotypic descriptions of the LAB and yeast isolated in the sampling are

405

required to better understand the relationship between these two types of organism.

406

The LAB species L. sanfranciscensis unquestionably dominated the sourdough ecosystem. In

407

previous studies of French sourdoughs, L. sanfranciscensis was either not found (Valcheva et

408

al., 2006; Vera et al., 2011) or detected but not predominantly (Ferchichi et al., 2007; Robert

409

et al., 2009; Valcheva et al., 2005). Yet, with the exception of reports by Groenewald et al.

410

(2006), L. sanfranciscensis has only been isolated from sourdough and is known to be

411

dominant (Lattanzi et al., 2013; Minervini et al., 2012a; Scheirlinck et al., 2008). Moreover,

412

the analysis of the sequenced L. sanfranciscensis TMW 1.1304 strain has shown that its

413

metabolism is specifically adapted to the sourdough environment. Hence, the genome of L.

414

sanfranciscensis enables it to grow easily in this environment and contributes to its ability to

fermentative

metabolism

without

significant

growth.

AC
C

EP

TE
D

throughout

M
AN
U

SC

RI
PT

390

17

ACCEPTED MANUSCRIPT

out-compete other bacteria in traditional sourdough (Vogel et al., 2011). Unlike L.

416

sanfranciscensis, the other LAB species detected in this study, L pentosus, L. sakei, L.

417

plantarum and L. kimchii, have already been detected in sourdough and are also frequently

418

isolated from other habitats (Botta and Cocolin, 2012; Chaillou et al., 2013; Corsetti and

419

Gobbetti, 2002; Yoon et al., 2000).

420

Compared to other extensive sourdoughs studies in Italy and Belgium (Lattanzi et al. 2013,

421

Minervini et al. 2012a, Minervini et al. 2012b, Vrancken et al. 2010), we found a high

422

diversity for yeast species whereas only five bakeries were sampled. There was nine yeast

423

species over all five bakeries. Five species belonged to the Saccharomycotina subphylum and

424

among them most were in the Kazachstania clade: K. unispora, K. exigua, K. bulderi and C.

425

humilis (Kurtzman and Robnett, 2003). All the dominant yeast species, except K. bulderi,

426

have already been described in bakery products (Minervini, et al., 2012a; Nuobariene et al.,

427

2012; Palomba et al., 2010; Valmorri et al., 2010; Vrancken et al., 2010; Zhang et al., 2011).

428

Among 69 natural sourdoughs previously analyzed in Italy, Belgium, Denmark and China,

429

more than 90% contained S. cerevisiae, 23% C. humilis and 3% K. unispora (Gullo et al.,

430

2003; Minervini, 2012a; Nuobariene et al., 2012; Palomba et al., 2010; Scheirlinck et al.,

431

2007; Valmorri et al., 2010; Vrancken et al., 2010; Zhang et al., 2011). In contrast with the

432

literature, S. cerevisiae and C. humilis were not the most common species in our sampling (for

433

a review, see De Vuyst et al. (2014)). In fact, despite the use of bakers yeast for another type

434

of bread product in Bakery 5, S. cerevisiae was only identified once in 398 isolates. This

435

might be contamination from bakers yeast (Minervini et al., 2012a; Venturi et al., 2012;

436

Vrancken et al., 2010) or an introduction from the flour (Meroth et al., 2004). Our results

437

suggest that S. cerevisiae does not compete with other sourdough and bread dough yeast

438

species and that Kazachstania species may be as well or better adapted to these environments.

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

415

18

ACCEPTED MANUSCRIPT

The species K. bulderi, which we found for the first time in bread dough, has been isolated in

440

a 3-month-old maize silage (Middelhoven et al., 2000) and in liquid food used to feed piglets

441

(Gori et al., 2011). The recent description of this species, as well as its genetic proximity with

442

Kazachstania barnettii (Middelhoven et al., 2000), could explain why it has never been

443

previously identified in sourdough. Surprisingly, we also found that R. mucilaginosa could

444

become the dominant identified yeast species at the last step of the bread-making process. The

445

invasion of R. mucilaginosa might be related to the textile covering the bread at the end of the

446

bread-making process and could play a role in the final bread taste and flavor, though not in

447

the bread rise (Zhang et al., 2011). Overall, we found that organic French sourdoughs harbor

448

special yeast communities in which K. bulderi (48% of the isolates) and C. humilis (24%) are

449

common and S. cerevisiae (less than 0.2%) is rare. Whether this high yeast species diversity is

450

related to organic flour acting as a reservoir or is related to the sampled bakeries, which are on

451

farm or closely related to farmers (except for Bakery 5), should be investigated.

452

The dominant yeast and LAB species appeared to be stable during bread-making process. The

453

dominant yeast and LAB species were also stable over bread-making runs except in one case,

454

where the yeast species T. delbruecki was replaced by K. exigua. This confirmed previous

455

findings by Venturi et al. (2012) and Rosenquist and Hansen (2000). The LAB PCR-TTGE

456

fingerprints did not fully agree with the data obtained using the culture-dependent approach.

457

This discord was expected because PCR-TTGE analyzes the total DNA of a matrix (Ercolini,

458

2004) and thus does not distinguish between viable and non-viable microorganisms. On the

459

other hand, when bacteria are isolated by plating but not detected with PCR-TTGE, it seems

460

they are not a major component of the microbial community being investigated (Alfonzo et

461

al., 2013). However, overall, the profiles were identical even in the case of Bakery 3, which

462

switched from rye/wheat flour to wheat flour. These results are consistent with previous data

463

(Scheirlinck et al. 2008, Minervini et al. 2012b; Gnzle, 2014; Weckx et al., 2010) showing

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

439

19

ACCEPTED MANUSCRIPT

that the LAB diversity is mainly influenced by the bakery environment, bakery practices and

465

back-slopping parameters rather than by the flour used.

466

The stable presence of L. sanfranciscensis in the different bakeries suggests a high fitness of

467

this species in natural sourdoughs as previously reported by other authors (Venturi et al. 2012,

468

Meroth et al. 2003, Siragusa et al., 2009, Vrancken et al., 2011a). The stable presence of K.

469

unispora and K. bulderi was never reported before and suggests that these species are fit in

470

natural sourdoughs. By contrast, the low yeast count of T. delbruecki, and L. plantarum as

471

well as the fact that they have been replaced over time completely or partly, respectively by K.

472

exigua and L. sanfranciscensis may reflect a low competitiveness of T. delbruecki and L.

473

plantarum compared to K. exigua and L. sanfranciscensis.

474

To conclude, this study is the first to highlight the overall stability of the microbial species

475

composition during the bread-making process and to show a significant variation in microbial

476

counts during and between bread-making runs. Our results reveal the predominance of L.

477

sanfranciscensis and the dominance of Kazachstania species in organic sourdoughs of five

478

French bakeries and suggest that S. cerevisiae is not a good competitor in natural organic

479

sourdough. It will be interesting to study the intraspecific diversity and functional potential of

480

different strains of L. sanfranciscensis and Kazachstania species. Further investigations

481

should also be carried out to test whether organic sourdoughs are a reservoir of microbial

482

diversity.

SC

M
AN
U

TE
D

EP

AC
C

483

RI
PT

464

484

Aknowledgements

485

This work was supported by grants from Rgion Ile-de-France, Oniris and ANR Bakery. We

486

wish to thank Jean-Franois Berthelot, Christian Dalmasso, Michel Perrin, and the bakers

487

from Pain Virgule and La Boulangeoise who shared with us their bread making

488

knowledge and pleasure of good bread. We also thank Serge Casaregola, Christelle Louis-

20

ACCEPTED MANUSCRIPT

Mondsir and Huu-Vang Nguyen for providing strains and for their kind assistance. Many

490

thanks to Sylvie and Isabelle from Oniris for their technical assistance. We are grateful to

491

Professor Herv Prvost for providing us the internal sequences of the 16S rDNA primers for

492

the lactic acid bacteria isolates identification.

RI
PT

489

AC
C

EP

TE
D

M
AN
U

SC

493

21

ACCEPTED MANUSCRIPT

References

495

Alfonzo, A., Ventimiglia, G., Corona, O., Di Gerlando, R., Gaglio, R., Francesca, N.,

496

Moschetti, G., Settanni, L., 2013. Diversity and technological potential of lactic acid

497

bacteria of wheat flours. Food Microbiology 36, 343-354.

RI
PT

494

Ammor, S., Dufour, E., Zagorec, M., Chaillou, S., Chevallier, I., 2005. Characterization and

499

selection of Lactobacillus sakei strains isolated from traditional dry sausage for their

500

potential use as starter cultures. Food Microbiology 22, 529-538.

SC

498

Botta, C., Cocolin, L., 2012. Microbial dynamics and biodiversity of table olives

502

fermentations: the different culture-dependent and independent approaches. Frontiers in

503

Microbiology 3, 245.

504
505

M
AN
U

501

Burke, D., Dawson, D.S., Stearns, T., 2000. Methods in yeast genetics: A Cold Spring Harbor
Laboratory course manual, Burke, D, New York, 2000 ed.

Chaillou, S., Champomier-Verges, M.-C., Cornet, M., Crutz-Le Coq, A.-M., Dudez, A.-M.,

507

Martin, V., Beaufils, S., Darbon-Rongere, E., Bossy, R., Loux, V., Zagorec, M., 2005. The

508

complete genome sequence of the meat-borne lactic acid bacterium Lactobacillus sakei

509

23K. Nat Biotech 23, 1527-1533.

TE
D

506

Chaillou, S., Lucquin, I., Najjari, A., Zagorec, M., Champomier-Verges, M.C., 2013.

511

Population genetics of Lactobacillus sakei reveals three lineages with distinct

512

evolutionary histories. PLoS ONE 8, e73253.

514
515
516
517
518

AC
C

513

EP

510

Chao, A., 1984. Nonparametric estimation of the number of classes in a population.

Scandinavian Journal of statistics, 265-270.

Corsetti, A., Gobbetti, M., 2002. Lactobacillus plantarum, in: Roginski, H. (Ed.),
Encyclopedia of Dairy Sciences. Elsevier, Oxford, pp. 1501-1507.
Corsetti, A., Settanni, L., 2007. Lactobacilli in sourdough fermentation. Food Research
International 40, 539-558.

22

ACCEPTED MANUSCRIPT

519

Daniel, H.M., Meyer, W., 2003. Evaluation of ribosomal RNA and actin gene sequences for

520

the identification of ascomycetous yeasts. International Journal of Food Microbiology 86,

521

61-78.

523

De Vuyst, L., Neysens, P., 2005. The sourdough microflora: biodiversity and metabolic
interactions. Trends in Food Science & Technology 16, 43-56.

RI
PT

522

De Vuyst, L., Van Kerrebroeck, S., Harth, H., Huys, G., Daniel, H.M., Weckx, S., 2014.

525

Microbial ecology of sourdough fermentations: Diverse or uniform? Food Microbiology

526

37, 11-29.

528

De Vuyst, L., Vancanneyt, M., 2007. Biodiversity and identification of sourdough lactic acid

M
AN
U

527

SC

524

bacteria. Food Microbiology 24, 120-127.

529

De Vuyst, L., Vrancken, G., Ravyts, F., Rimaux, T., Weckx, S., 2009. Biodiversity, ecological

530

determinants, and metabolic exploitation of sourdough microbiota. Food Microbiology

531

26, 666-675.

534
535

TE
D

533

Ercolini, D., Hill, P.J., Dodd, C.E.R., 2003. Bacterial community structure and location in
Stilton cheese. Applied and Environmental Microbiology 69, 3540-3548.
Ercolini, D., 2004. PCR-DGGE fingerprinting: novel strategies for detection of microbes in
food. Journal of Microbiological Methods 56, 297-314.

EP

532

Ercolini, D., 2013. High-throughput sequencing and metagenomics: moving forward in the

537

culture-independent analysis of food microbial ecology. Applied and Environmental

538

AC
C

536

Microbiology 79, 3148-3155.

539

Ercolini, D., Pontonio, E., De Filippis, F., Minervini, F., La Storia, A., Gobbetti, M., Di

540

Cagno, R., 2013. Microbial ecology dynamics during rye and wheat sourdough

541

preparation. Applied and Environmental Microbiology 79, 7827-7836.

23

ACCEPTED MANUSCRIPT

542

Ferchichi, M., Valcheva, R., Prvost, H., Onno, B., Dousset, X., 2007. Molecular

543

identification of the microbiota of French sourdough using temporal temperature gradient

544

gel electrophoresis. Food Microbiology 24, 678-686.

Ferchichi, M., Valcheva, R., Prvost, H., Onno, B., Dousset, X., 2008. Rapid identification of

546

Lactobacillus nantensis, Lactobacillus spicheri and Lactobacillus hammesii species using

547

species-specific primers. International Journal of Food Microbiology 123, 269-276.

550
551
552
553

SC

549

Galle, S., Arendt, E.K., 2013. Exopolysaccharides from sourdough lactic acid bacteria.
Critical Reviews in Food Science and Nutrition 54, 891-901.

Gnzle, M.G., 2014. Enzymatic and bacterial conversions during sourdough fermentation.
Food Microbiology 37, 2-10.

M
AN
U

548

RI
PT

545

Gnzle, M.G., Vermeulen, N., Vogel, R.F., 2007. Carbohydrate, peptide and lipid metabolism
of lactic acid bacteria in sourdough. Food Microbiology 24, 128-138.
Gnzle, M.G., Vogel, R.F., 2003. Contribution of reutericyclin production to the stable

555

persistence of Lactobacillus reuteri in an industrial sourdough fermentation. International

556

Journal of Food Microbiology 80, 31-45.

558

Gobbetti, M., De Angelis, M., Corsetti, A., Di Cagno, R., 2005. Biochemistry and physiology
of sourdough lactic acid bacteria. Trends in Food Science & Technology 16, 57-69.

EP

557

TE
D

554

Gori, K., Bjorklund, M.K., Canibe, N., Pedersen, A.O., Jespersen, L., 2011. Occurrence and

560

identification of yeast species in fermented liquid feed for piglets. Microbial Ecology 61,

561

AC
C

559

146-153.

562

Granato, D., de Arajo Calado, V.M., Jarvis, B., 2014. Observations on the use of statistical

563

methods in Food Science and Technology. Food Research International 55, 137-149.

564

Groenewald, W.H., Van Reenen, C.A., Todorov, S.D., Toit, M.D., Witthuhn, R.C., Holzapfel,

565

W.H., Dicks, L.M.T., 2006. Identification of lactic acid bacteria from vinegar flies based

24

ACCEPTED MANUSCRIPT

566

on phenotypic and genotypic characteristics. American Journal of Enology and

567

Viticulture 57, 519-525.

Gullo, M., Romano, A.D., Pulvirenti, A., Giudici, P., 2003. Candida humilisdominant

569

species in sourdoughs for the production of durum wheat bran flour bread. International

570

Journal of Food Microbiology 80, 55-59.

RI
PT

568

Hammes, W.P., Brandt, M.J., Francis, K.L., Rosenheim, J., Seitter, M.F.H., Vogelmann, S.A.,

572

2005. Microbial ecology of cereal fermentations. Trends in Food Science & Technology

573

16, 4-11.

575

Huys, G., Daniel, H.-M., De Vuyst, L., 2013. Taxonomy and biodiversity of sourdough yeasts

M
AN
U

574

SC

571

and lactic acid bacteria, Handbook on Sourdough Biotechnology. Springer, pp. 105-154

576

Iacumin, L., Cecchini, F., Manzano, M., Osualdini, M., Boscolo, D., Orlic, S., Comi, G.,

577

2009. Description of the microflora of sourdoughs by culture-dependent and culture-

578

independent methods. Food Microbiology 26, 128-135.

Jaffrs, E., Sohier, D., Leroi, F., Pilet, M.F., Prvost, H., Joffraud, J.J., Dousset, X., 2009.

580

Study of the bacterial ecosystem in tropical cooked and peeled shrimps using a

581

polyphasic approach. International Journal of Food Microbiology 131, 20-29.

584
585
586

EP

583

Kan, V.L., 1993. Polymerase chain reaction for the diagnosis of candidemia. Journal of
Infectious Diseases 168, 779-783.
Kurtzman, C.P., Robnett, C.J., 2003. Phylogenetic relationships among yeasts of the

AC
C

582

TE
D

579

Saccharomyces complex determined from multigene sequence analyses. FEMS Yeast

Research 3, 417-432.

587

Lattanzi, A., Minervini, F., Di Cagno, R., Diviccaro, A., Antonielli, L., Cardinali, G.,

588

Cappelle, S., De Angelis, M., Gobbetti, M., 2013. The lactic acid bacteria and yeast

589

microbiota of eighteen sourdoughs used for the manufacture of traditional Italian sweet

590

leavened baked goods. International Journal of Food Microbiology 163, 71-79.

25

ACCEPTED MANUSCRIPT

Marcon, E., Morneau, F., 2010. Mesures de la biodiversit.

592

Meroth, C.B., Hammes, W.P., Hertel, C., 2003a. Identification and population dynamics of

593

yeasts in sourdough fermentation processes by PCR-denaturing gradient gel

594

electrophoresis. Applied and Environmental Microbiology 69, 7453-7461.

RI
PT

591

Meroth, C.B., Hammes, W.P., Hertel, C., 2004. Characterisation of the microbiota of rice

596

sourdoughs and description of Lactobacillus spicheri sp. nov. Systematic and Applied

597

Microbiology 27, 151-159.

SC

595

Meroth, C.B., Walter, J., Hertel, C., Brandt, M.J., Hammes, W.P., 2003b. Monitoring the

599

bacterial population dynamics in sourdough fermentation processes by using PCR-

600

denaturing gradient gel electrophoresis. Applied and Environmental Microbiology 69,

601

475-482.

602
603

M
AN
U

598

Middelhoven, W., Kurtzman, C., Vaughan-Martini, A., 2000. Saccharomyces bulderi sp. nov.,
a yeast that ferments gluconolactone. Antonie van Leeuwenhoek 77, 223-228.
Minervini, F., Di Cagno, R., Lattanzi, A., De Angelis, M., Antonielli, L., Cardinali, G.,

605

Cappelle, S., Gobbetti, M., 2012a. Lactic acid bacterium and yeast microbiotas of 19

606

sourdoughs used for traditional/typical Italian breads: interactions between ingredients

607

and microbial species diversity. Applied and Environmental Microbiology 78, 1251-1264.

608

Minervini, F., Lattanzi, A., De Angelis, M., Di Cagno, R., Gobbetti, M., 2012b. Influence of

609

artisan bakery- or laboratory- propagated sourdoughs on the diversity of lactic acid

611

EP

AC
C

610

TE
D

604

bacterium and yeast microbiotas. Applied and Environmental Microbiology 78, 53285340.

612

Nguyen, H.-V., Gaillardin, C., 1997. Two subgroups within the Saccharomyces bayanus

613

species evidenced by PCR amplification and restriction polymorphism of the non-

614

transcribed spacer 2 in the ribosomal DNA unit. Systematic and Applied Microbiology

615

20, 286-294.

26

ACCEPTED MANUSCRIPT

Nguyen, H.-V., Gaillardin, C., Neuvglise, C., 2009. Differentiation of Debaryomyces

617

hansenii and Candida famata by rRNA gene intergenic spacer fingerprinting and

618

reassessment of phylogenetic relationships among D. hansenii, C. famata, D. fabryi, C.

619

flareri (=D. subglobosus) and D. prosopidis: description of D. vietnamensis sp. nov.

620

closely related to D. nepalensis. FEMS Yeast Research 9, 641-662.

623
624

active yeasts from sourdoughs. LWT - Food Science and Technology 48, 190-196.

SC

622

Nuobariene, L., Hansen, .S., Arneborg, N., 2012. Isolation and identification of phytase-

ODonnell, K., 1993. Fusarium and its near relatives. The fungal holomorph: mitotic, meiotic
and pleomorphic speciation in fungal systematics, 225-233.

M
AN
U

621

RI
PT

616

625

Palomba, S., Blaiotta, G., Ventorino, V., Saccone, A., Pepe, O., 2010. Microbial

626

characterization of sourdough for sweet baked products in the Campania region (southern

627

Italy) by a polyphasic approach. Annals of Microbiology 61, 307-314.

Parayre, S., Falentin, H., Madec, M.-N., Sivieri, K., Le Dizes, A.-S., Sohier, D., Lortal, S.,

629

2007. Easy DNA extraction method and optimisation of PCR-temporal temperature gel

630

electrophoresis to identify the predominant high and low GC-content bacteria from dairy

631

products. Journal of Microbiological Methods 69, 431-441.

634
635
636

EP

633

Poutanen, K., Flander, L., Katina, K., 2009. Sourdough and cereal fermentation in a
nutritional perspective. Food Microbiology 26, 693-699.
Reale, A., Di Renzo, T., Succi, M., Tremonte, P., Coppola, R., Sorrentino, E., 2013.

AC
C

632

TE
D

628

Microbiological and fermentative properties of baker's yeast starter used in breadmaking.

Journal of Food Science 78, M1224-1231.

637

Robert, H., Gabriel, V., Fontagn-Faucher, C., 2009. Biodiversity of lactic acid bacteria in

638

French wheat sourdough as determined by molecular characterization using species-

639

specific PCR. International Journal of Food Microbiology 135, 53-59.

27

ACCEPTED MANUSCRIPT

640
641

Rosenquist, H., Hansen, A., 2000. The microbial stability of two bakery sourdoughs made
from conventionally and organically grown rye. Food Microbiology 17, 241-250.
Scheirlinck, I., Van der Meulen, R., Van Schoor, A., Vancanneyt, M., De Vuyst, L.,

643

Vandamme, P., Huys, G., 2007. Influence of geographical origin and flour type on

644

diversity of lactic acid bacteria in traditional Belgian sourdoughs. Applied and

645

Environmental Microbiology 73, 6262-6269.

RI
PT

642

Scheirlinck, I., Van der Meulen, R., Van Schoor, A., Vancanneyt, M., De Vuyst, L.,

647

Vandamme, P., Huys, G., 2008. Taxonomic structure and stability of the bacterial

648

community in Belgian sourdough ecosystems as assessed by culture and population

649

fingerprinting. Applied and Environmental Microbiology 74, 2414-2423.

651

M
AN
U

650

SC

646

Shannon, C.E., Weaver, W., 1949. 77 mathematical theory of communication. Urbana:

University of Illinois Press.

Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688.

653

Siragusa, S., Di Cagno, R., Ercolini, D., Minervini, F., Gobbetti, M., De Angelis, M., 2009.

654

Taxonomic structure and monitoring of the dominant population of lactic acid bacteria

655

during wheat flour sourdough type I propagation using Lactobacillus sanfranciscensis

656

starters. Applied and Environmental Microbiology 75, 1099-1109.

EP

658

Stackebrandt, E., Ebers, J., 2006. Taxonomic parameters revisited: tarnished gold standards.
Microbiology today 33, 152.

AC
C

657

TE
D

652

659

Torriani, S., Felis, G.E., Dellaglio, F., 2001. Differentiation of Lactobacillus plantarum, L.

660

pentosus, and L. paraplantarum by recA gene sequence analysis and multiplex PCR

661
662

assay with recA gene-derived primers. Applied and Environmental Microbiology 67,
3450-3454.

663

Valcheva, R., Ferchichi, M.F., Korakli, M., Ivanova, I., Gnzle, M.G., Vogel, R.F., Prvost,

664

H., Onno, B., Dousset, X., 2006. Lactobacillus nantensis sp. nov., isolated from French

28

ACCEPTED MANUSCRIPT

665

wheat sourdough. International Journal of Systematic and Evolutionary Microbiology 56,

666

587-591.
Valcheva, R., Korakli, M., Onno, B., Prvost, H., Ivanova, I., Ehrmann, M.A., Dousset, X.,

668

Gnzle, M.G., Vogel, R.F., 2005. Lactobacillus hammesii sp. nov., isolated from French

669

sourdough. International Journal of Systematic and Evolutionary Microbiology 55, 763-

670

767.

RI
PT

667

Valmorri, S., Tofalo, R., Settanni, L., Corsetti, A., Suzzi, G., 2010. Yeast microbiota

672

associated with spontaneous sourdough fermentations in the production of traditional

673

wheat sourdough breads of the Abruzzo region (Italy). Antonie van Leeuwenhoek 97,

674

119-129.

M
AN
U

SC

671

Venturi, M., Guerrini, S., Granchi, L., Vincenzini, M., 2012. Typing of Lactobacillus

676

sanfranciscensis isolates from traditional sourdoughs by combining conventional and

677

multiplex RAPD-PCR profiles. International Journal of Food Microbiology 156, 122-

678

126.

TE
D

675

Vera, A., Ly-Chatain, M.H., Rigobello, V., Demarigny, Y., 2011. Description of a French

680

natural wheat sourdough over 10 consecutive days focussing on the lactobacilli present in

681

the microbiota. Antonie van Leeuwenhoek 101, 369-377.

EP

679

Vernocchi, P., Valmorri, S., Gatto, V., Torriani, S., Gianotti, A., Suzzi, G., Guerzoni, M.E.,

683

Gardini, F., 2004. A survey on yeast microbiota associated with an Italian traditional

684

AC
C

682

sweet-leavened baked good fermentation. Food Research International 37, 469-476.

685

Vogel, R., Pavlovic, M., Ehrmann, M., Wiezer, A., Liesegang, H., Offschanka, S., Voget, S.,

686

Angelov, A., Bocker, G., Liebl, W., 2011. Genomic analysis reveals Lactobacillus

687

sanfranciscensis as stable element in traditional sourdoughs. Microbial Cell Factories 10,

688

S6.

29

ACCEPTED MANUSCRIPT

689
690

Vogelmann, S.A., Hertel, C., 2011. Impact of ecological factors on the stability of microbial
associations in sourdough fermentation. Food Microbiology 28, 583-589.
Vogelmann, S.A., Seitter, M., Singer, U., Brandt, M.J., Hertel, C., 2009. Adaptability of lactic

692

acid bacteria and yeasts to sourdoughs prepared from cereals, pseudocereals and cassava

693

and use of competitive strains as starters. International Journal of Food Microbiology

694

130, 205-212.

RI
PT

691

Vrancken, G., Rimaux, T., Weckx, S., Leroy, F., De Vuyst, L., 2011a. Influence of temperature

696

and backslopping time on the microbiota of a type I propagated laboratory wheat

697

sourdough fermentation. Applied and Environmental Microbiology 77, 2716-2726.

M
AN
U

SC

695

698

Vrancken, G., De Vuyst, L., Rimaux, T., Allemeersch, J., Weckx, S., 2011b. Adaptation of

699

Lactobacillus plantarum IMDO 130201, a wheat sourdough isolate, to growth in wheat

700

sourdough simulation medium at different pH values through differential gene

701

expression. Applied and Environmental Microbiology 77, 3406-3412.

Vrancken, G., De Vuyst, L., Van der Meulen, R., Huys, G., Vandamme, P., Daniel, H.-M.,

703

2010. Yeast species composition differs between artisan bakery and spontaneous

704

laboratory sourdoughs. FEMS Yeast Research 10, 471-481.

TE
D

702

Weckx, S., Van der Meulen, R., Maes, D., Scheirlinck, I., Huys, G., Vandamme, P., De Vuyst,

706

L., 2010. Lactic acid bacteria community dynamics and metabolite production of rye

707

sourdough fermentations share characteristics of wheat and spelt sourdough

709
710

AC
C

708

EP

705

fermentations. Food Microbiology 27, 1000-1008.

Weisburg, W.G., Barns, S.M., Pelletier, D.A., Lane, D.J., 1991. 16S ribosomal DNA
amplification for phylogenetic study. Journal of Bacteriology 173, 697-703.

711

Yoon, J.H., Kang, S.S., Mheen, T.I., Ahn, J.S., Lee, H.J., Kim, T.K., Park, C.S., Kho, Y.H.,

712

Kang, K.H., Park, Y.H., 2000. Lactobacillus kimchii sp. nov., a new species from kimchi.

713

International Journal of Systematic and Evolutionary Microbiology 50, 1789-1795.

30

ACCEPTED MANUSCRIPT

714

Zhang, J., Liu, W., Sun, Z., Bao, Q., Wang, F., Yu, J., Chen, W., Zhang, H., 2011. Diversity of

715

lactic acid bacteria and yeasts in traditional sourdoughs collected from western region in

716

Inner Mongolia of China. Food Control 22, 767-774.

RI
PT

717

AC
C

EP

TE
D

M
AN
U

SC

718

31

ACCEPTED MANUSCRIPT

14

11
15

9
12 152

66

14

2
2

79.0 12.8 2.7 0.4 2.8 2.3

K. exigua

S. cerevisiae

Rhodotorula sp.

C. humilis

SC

M
AN
U

7
17
13
18
5
2
1
2

79

RI
PT

K. bulderi

K. unispora

C. carpophila

T. delbruecki

L. hammessii

H. pseudoburtonii
1
1

40
39
40
12
60

EP

11
13
19
2
7
12
29
25
29
20
10
411

TE
D

L. kimchii

L. pentosus

L. sakei

40
35
38
39

25
24
23
15
22
17
22
25
22
36

AC
C

B1_BMR1_CS
B1_BMR1_LD
B1_BMR1_AK
B1_BMR1_BB
B1_BMR2_LD
B2_BMR1_CS
B2_BMR1_LD
B2_BMR1_AK
B2_BMR1_BB
B2_BMR2_LD
B3_BMR1_CS
B3_BMR1_LD
B3_BMR1_AK
B3_BMR1_BB
B3_BMR2_LD
B4_BMR1_CS
B4_BMR1_LD
B4_BMR1_AK
B4_BMR1_BB
B4_BMR2_LD
B5_BMR1_CS
B5_BMR1_LD
B5_BMR1_AK
B5_BMR1_BB
B5_BMR2_LD
Total isolates
Percentage
(%) of LAB
and yeasts
isolates

L. plantarum

Table 1: Number of yeasts and LAB isolates for each bakery (B) from the two bread-making
runs (BMR1 and BMR2) according to the process step (CS: chief sourdough; LSD: final
leavened dough; AK: after kneading; BB: before baking).
L. sanfranciscensis

1
2
3
4

28
20
80
80
80
80
80
80
80
80
80
80

79
80
79
80
79
191 800 397

28

9.1 0.2 0.1 11.4 47.8 23.7 1.7

99

1
2

5.9 0.1

The genus names were abbreviated as L: Lactobacillus; C: Candida; T: Torulospora; H:

Hyphopichia; K: Kazachstania; S: Saccharomyces

ACCEPTED MANUSCRIPT

Table 2: diversity indexes of the LAB and yeast communities

3
Observed richness

Chao1 richness

Shannon diversity

RI
PT

1
2

Simpson diversity

LAB

Yeast

LAB

Yeast

LAB

Bakery 1

5.5

0.78

Bakery 2

0.64

0.34

Bakery 3

0.49

0.31

Bakery 4

Bakery 5

5.5

Total

EP
AC
C

LAB

0.48

M
AN
U
0

0.80

0.44

0.02

0.71

0.01

0.35

1.43

0.76

0.69

0.36

TE
D

Yeast

SC

Yeast

Time (hours)

20

25

10

15

Time (hours)

20

25

10
10

15

20

10

Time (hours)

20

8
7

25

10

15

20

25

10
CS

LD AK

BB

BB

LD AK

CS

10
15

6
0

Time (hours)

4
5

25

9
BB

BB

4
5

3
0

LD AK

Time (hours)

LD AK

CS

3
0

CS

LAB count (log 10 CFU/g)

10
8
6
5
4
3
25

4
5

RI
PT

20

3
0

LAB count (log 10 CFU/g)

SC

EP
8

Yeast count (log 10 CFU/g)

6
4

15

15

AC
C

BB

3
10

10

TE
D

10
9

10

LD AK

CS

Yeast count (log 10 CFU/g)

6
5

Time (hours)

10
9

BB

4
0

8
0

Time (hours)

LD AK

25

BB

Yeast count (log 10 CFU/g)

20

LD AK

15

CS

10

Bakery 5

10
5

Time (hours)

CS

6
4
3

Yeast count (log 10 CFU/g)

25

M
AN
U

20

15

BB

10

LD AK

CS

LAB count (log 10 CFU/g)

BB

3
0

Bakery 4

10
8
7

LD AK

5
3

CS

BB

LAB count (log 10 CFU/g)

10
9
8
7

LD AK

CS

ACCEPTED
MANUSCRIPT
Bakery
3

Bakery 2

LAB count (log 10 CFU/g)

Yeast count (log 10 CFU/g)

Bakery 1

10

15

Time (hours)

20

25

10

15

Time (hours)

20

25

ACCEPTED MANUSCRIPT

AK1 BB1 CS2 LD2 AK2 BB2

1A1 1A2

CS1 LD1

AK1

BB1 CS2 LD2

AK2

BB2

1B

CS1 LD1 AK1 BB1 CS2 LD2 AK2

BB2

SC

RI
PT

CS1 LD1

M
AN
U

BB1

CS2 LD2 AK2 BB2

1D

CS1 LD1 AK1 BB1 CS2 LD2 AK2 BB2

AC
C

EP

CS1 LD1 AK1

TE
D

Figure 2

1E

1C

B)

LAB.count

AC
C
0.5

0.0
Dim 1 (39.38%)

0.5

1.0

Yeast.count

TE
D
EP

0.5

pH

1.0

M
Dim.2
AN
0
U2S
C

0.0

1.0

Dim 2 (29.24%)

0.5

lactic.acid
TTA
acetic.acid

ratio

RI
PT

A)
4

1.0

ACCEPTED MANUSCRIPT

0
Dim.1

ACCEPTED MANUSCRIPT

Highlights

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

First description of microbial dynamics during sourdough bread-making process

Predominance of Lactobacillus sanfranciscensis in organic sourdough analyzed
Higher diversity in yeast species than in lactic acid bacteria species
High occurrence of Kazachstania species

ACCEPTED MANUSCRIPT

Supplementary material S1: Characteristics of the sourdough from the five selected bakeries and sampling periods.

Second bread-making run date

PoitouCharentes

Lot-et-Garonne

Rhne-Alpes

Pays de la Loire

Picardie

February 2012

March 2012

March 2012

April 2012

May 2012

November 2012

December 2012

January 2013

January 2013

December 2012

NF

NF

Wheat

Wheat (50%) +
rye (50 %)d

Wheat

Wheat

75

82

73

58

21

23

22.6

24.1

Farmer (F) / Non farmer (NF)

Flour in chief sourdough

Wheat

Hydration of sourdough (%)

87

No. of back-slopping steps before use

2
22.5

6h30

6h

11h

14h

17h30

43

16

15.1

33

9.9

37

44

49.4

39.4

53.8

19

38

35

27

34.8

0.9

1.3

0.8

0.8

1.5

Me

Me

Ma

Me

Me

4h30

6h50

4h45

4h20

5h50

Sourdough (%)
Flourb (%)
b

Water (%)
b

Salt (%)
c

Kneading method

Total period of fermentation

a

EP

TE
D

Temp (C) of back-slopping

Period (h) of back-slopping

RI
PT

First bread-making run date

SC

Location

M
AN
U

Bakery

AC
C

1
2
3

Mature sourdough, after the last back-slopping

Percentage of total dough weight
c
Mechanically (Me) or manually (Ma)
d
For the second bread-making run, only wheat flour was used
b

ACCEPTED MANUSCRIPT

Chief sourdough

sampling

Back-sloppings

M
AN
U

9h30 19h05*

SC

RI
PT

0 day 4 days*

Chief sourdough (CS)

TE
D

Final leavened
dough

sampling

Final leavened dough (LD)

directly - 15min *

sampling

Dough after kneading (AK)

EP

Kneading

AC
C

1h50 - 3h30*

Shaping

1h30 - 4h*

sampling

Baking
Supplementary material S2

Dough before baking (BB)

ACCEPTED MANUSCRIPT

Supplementary material S2: Bread making process and experimental design for sampling.

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

* Range of time depends on each bakerys practices

ACCEPTED MANUSCRIPT

1
2

Supplementary Material S3: List of the reference strains used in this study
Lactobacillus brevis CIP 102806T
Lactobacillus hammesii CIP 108546T
Lactobacillus kimchii DSM13961T
Lactobacillus plantarum ATCC 14917T
Lactobacillus sanfranciscensis ATCC 43322
Lactobacillus sanfranciscensis CIP 103252T
Saccharomyces cerevisiae CLIB 227 T
Kazachstania unispora CLIB 234 T

SC

Kazachstania exigua CLIB 179 T

Kazachstania bulderi CLIB 596 T
Candida humilis CLIB 1323 T

M
AN
U

Torulaspora delbrueckii CLIB 230 T

Candida carpophila CLIB 1330 T
Hyphopichia pseudoburtonii CLIB 35 T

RI
PT

Lactobacillus paraplantarum 1888

CIP : Collection of Institut Pasteur ; DSM : Deutsche Stammsammlung fr Mikroorganismen ;

ATCC : American Type Culture Collection ; CLIB : Collection de Levures d'Intrt

Biotechnologique, collection of CIRM-Levures

AC
C

EP

TE
D

ACCEPTED MANUSCRIPT

4
5
6
7
8

Supplementary material S4: Sequences of oligonucleotide primers used for PCR amplification and sequencing
Oligonucleotide sequence (5'
3')

Sequence length (bp)

Target genomic region

Reference

fD1
rD1

AGAGTTTGATCCTGGCTCAG
TAAGGAGGTGATCCAGCC

1500

16S rRNA gene

Weisburg et al. (1991)

SP1
SP2

ACTCCTACGGGAGGCAGCA
ACCGCGGCTGCTGGCACG

SP3
SP5

GATACCCTGGTAGTCCACG
GGTACCTTGTTACGACTT

16S rRNA gene

This study

planF
pentF

CCGTTTATGCGCAACACCTA
CAGTGGCGCGGTTGATAT

318 for Lb. plantarum

218 for Lb. pentosus

paraF
preV

GTCACAGGCATTACGAAAAC
TCGGGATTACCAAACATCAC

107 for Lb. paraplantarum

recA

Torriani et al. (2001)

Lham
23S/p7

TTGACCGCAAGGTCATCATTTT
GGTACTTAGATGTTTCAGTTC

196

16S-23S rDNA intergenic spacer region

(ISR)

Ferchichi et al. (2006)

702-F
310-R

AATTGCCTTCTTCCGTGT A
AGTTGCGCACAATTATTT TC

407

katA

Ammor et al. (2005)

V3P1
V3P3

CCTACGGGAGGCAGCAG
ATTACCGCGGCTGCTGG with GC-clamp

194

16S rRNA gene

Parayre et al. (2007)

SR21

CTTAATCTTTGAGACAAGC

LR13

CGATCTGCTGAGATTAAG

intergenic spacer rDNA

Nguyen et al. (2009)

NL1

GCATATCAATAAGCGGAGGAAAAG

NL4

GGTCCGTGTTTCAAGACGG

26S gene D1/D2 region ribosomal gene

O'Donnell (1993)

CA5R

GTGAACAATGGATGGACCAGATTCGTCG

CA14

AACTGGGATGACATGGAGAAGATCTGGC

TE
D

SC

M
AN
U

1280

RI
PT

Primer set

EP

2540

AC
C

1
2
3

550

800

ACT1 gene
Kan (1993)

ACCEPTED MANUSCRIPT
Supplementary Material S5: Physico-chemical characterization and microbial counts for each
Bakery (B) from the two bread-making runs (BMR1 and BMR2) according to the sample step
(CS: chief sourdough; LSD: final leavened dough; AK: after kneading; BB: before baking)
and the dough replicate (R1 = dough replicate 1 or R2 = dough replicate 2).
Results are means standard deviation (n = 3).

14.17 0.96 7.96 0.52

10.57 3.11 5.19 ND
8.00 0.70
6.32 ND
9.90 1.82
4.32 ND
27.30 3.90 9.72 1.00
31.30 4.10 11.30 1.44
16.43 0.75 6.98 0.20
15.50 0.71 7.00 1.37
19.10 3.38 8.45 ND
19.50 2.55 8.45 0.13
21.50 2.76 8.19 ND
17.35 0.92 7.95 ND
15.77 0.40 11.34 0.56
11.57 0.75 6.59 ND
7.90 0.17
4.21 ND
12.87 4.21 9.35 ND
25.35 2.76 6.14 ND
16.43 0.75 3.47 ND
15.40 0.85 4.27 ND
19.10 3.38 4.61 ND
22.30 5.17 6.64 ND
21.50 2.76 5.26 ND
18.67 2.37 6.32 ND
19.00 0.00 11.44 0.10
16.23 0.40 9.34 ND
11.80 1.01 6.34 ND
15.30 0.00 9.12 ND
19.10 1.71 6.06 ND
21.30 2.00 3.29 ND
12.00 3.25 5.28 ND
16.00 1.30 5.29 ND
11.13 2.02 4.17 ND
10.67 2.37 3.86 ND
14.73 2.30 4.52 ND
14.23 1.36 4.31 ND
17.80 0.35 9.98 ND
14.67 0.06 9.14 ND
11.57 0.75 5.00 ND
15.53 0.81 8.20 ND
16.43 0.38 6.95 ND

0.93 0.05
0.46 ND
0.51 ND
0.43 ND
1.92 ND
2.03 ND
0.66 ND
0.81 ND
0.73 ND
0.84 ND
0.64 ND
0.66 ND
1.18 0.56
0.78 ND
0.55 ND
0.98 ND
1.17 ND
0.67 ND
0.68 ND
0.74 ND
0.44 ND
0.90 ND
0.42 ND
1.94 0.04
1.27 ND
1.09 ND
1.05 ND
1.65 ND
0.66 ND
0.64 ND
0.67 ND
0.76 ND
0.44 ND
0.51 ND
0.71 ND
0.99 ND
0.46 ND
0.62 ND
0.55 ND
1.70 ND

AC
C

Log10 LAB
count
(CFU/g)

Log10 Yeast
count
(CFU/g)

9.25 0.04
8.85 0.11
8.99 0.12
8.93 0.01
9.00 0.04
9.05 0.09
9.21 0.05
9.16 0.07
9.14 0.09
9.22 0.06
9.24 0.04
9.57 0.12
9.01 0.13
8.81 0.06
8.80 0.07
9.10 0.01
8.43 0.11
8.46 0.09
8.41 0.06
8.33 0.07
8.34 0.05
8.58 0.08
8.59 0.01
7.55 0.09
7.71 0.14
7.06 0.15
7.42 0.11
8.93 0.07
9.03 0.10
8.99 0.03
9.04 0.04
8.67 0.02
8.75 0.01
8.98 0.10
9.13 0.03
8.50 0.06
8.24 0.04
7.24 0.04
7.84 0.05
7.87 0.30

4.74 0.11
3.66 0.40
4.08 0.09
4.18 0.04
5.66 0.02
5.67 0.08
5.55 0.04
5.56 0.06
5.57 0.07
5.67 ND
5.66 0.06
5.56 0.03
5.34 0.09
5.27 0.03
4.67 0.10
5.59 0.10
6.04 0.07
6.52 0.09
6.51 0.02
6.04 0.03
5.97 0.07
6.09 0.08
6.06 0.04
7.17 0.03
7.39 0.01
6.76 0.04
7.15 0.01
7.56 0.04
7.52 0.04
7.52 0.03
7.49 0.04
6.69 0.05
6.63 0.07
7.16 0.04
7.15 0.03
7.18 0.08
7.17 0.08
6.88 0.15
6.99 0.02
6.00 0.03

RI
PT

[acid acetic]
g/kg)

SC

3.90 0.00
4.30 0.00
4.50 0.00
4.30 0.00
3.74 0.04
3.69 0.06
3.79 0.00
3.84 0.01
3.76 0.03
3.70 0.03
3.74 0.02
3.76 0.03
3.83 0.06
4.26 0.06
4.74 0.06
4.02 0.04
3.74 0.04
3.79 0.00
3.84 0.01
3.76 0.03
3.73 0.03
3.76 0.03
3.74 0.02
3.85 0.07
3.90 0.00
4.13 0.06
3.90 0.00
3.76 0.04
3.89 0.02
3.92 0.11
3.84 0.04
4.81 0.45
4.59 0.01
3.87 0.06
4.01 0.03
3.90 0.00
4.00 0.00
4.20 0.00
4.00 0.00
3.76 0.10

[lactic acid]
g/kg

M
AN
U

B1_BMR1_CS_R1
B1_BMR1_LD_R1
B1_BMR1_AK_R1
B1_BMR1_BB_R1
B1_BMR2_CS_R1
B1_BMR2_CS_R2
B1_BMR2_LD_R1
B1_BMR2_LD_R2
B1_BMR2_AK_R1
B1_BMR2_AK_R2
B1_BMR2_BB_R1
B1_BMR2_BB_R2
B2_BMR1_CS_R1
B2_BMR1_LD_R1
B2_BMR1_AK_R1
B2_BMR1_BB_R1
B2_BMR2_CS_R1
B2_BMR2_LD_R1
B2_BMR2_LD_R2
B2_BMR2_AK_R1
B2_BMR2_AK_R2
B2_BMR2_BB_R1
B2_BMR2_BB_R2
B3_BMR1_CS_R1
B3_BMR1_LD_R1
B3_BMR1_AK_R1
B3_BMR1_BB_R1
B3_BMR2_CS_R1
B3_BMR2_CS_R2
B3_BMR2_LD_R1
B3_BMR2_LD_R2
B3_BMR2_AK_R1
B3_BMR2_AK_R2
B3_BMR2_BB_R1
B3_BMR2_BB_R2
B4_BMR1_CS_R1
B4_BMR1_LD_R1
B4_BMR1_AK_R1
B4_BMR1_BB_R1
B4_BMR2_CS_R1

TTA (mL
NaOH 0.1 /
10g)

TE
D

pH

EP

Sample

ACCEPTED MANUSCRIPT
12.23 0.40
12.60 1.55
13.50 1.91
8.67 3.78
7.27 0.23
11.60 0.40
13.77 0.40
25.67 0.81
22.60 ND
16.60 2.17

5.01 ND
4.47 ND
4.32 ND
6.46 ND
4.25 ND
6.46 ND
5.22 ND
5.86 ND
6.14 ND
5.55 ND

0.75 ND
0.73 ND
0.56 ND
0.48 ND
0.57 ND
0.73 ND
0.59 ND
0.7 ND
0.79 ND
0.44 ND

8.02 0.21
7.81 0.07
7.71 0.29
7.66 0.01
7.77 0.09
8.16 0.07
8.23 0.14
8.49 0.08
8.30 0.03
8.37 0.24

M
AN
U
TE
D
EP
AC
C

5.87 0.38
5.69 0.04
6.06 0.05
5.70 0.05
5.71 0.03
6.18 0.08
6.11 0.01
7.37 0.04
7.23 0.02
6.64 0.03

RI
PT

3.73 0.08
3.91 0.06
3.80 0.04
4.20 0.03
4.26 0.04
3.78 0.09
3.87 0.06
3.61 0.05
3.46 0.06
3.66 0.07

SC

B4_BMR2_CS_R2
B4_BMR2_LD_R1
B4_BMR2_LD_R2
B4_BMR2_AK_R1
B4_BMR2_AK_R2
B4_BMR2_BB_R1
B4_BMR2_BB_R2
B5_BMR2_LD_R1
B5_BMR2_LD_R2
B5_BMR2_BB_R2

ACCEPTED MANUSCRIPT

Statistical
value

p-value

Method

0.42

0.51

Process step

25.18

<10

LAB

Bakery

0.25
[0.008;8.38]

-3.98x10-8

0.33

Bread
making run

0.59
[0.31;1.11]

74.98

<10

Dough
replicate

0.20
[0.16;0.25]

96.51

<10

4.82

0.03

0.11
[0.10;0.13]

Residuals

-5

-5

36.40

<10-4

Bakery

0.46
[0.08;2.79]

0.02

0.21

Bread
making run

0.68
[0.36;1.27]

85.46

<10-5

Dough
replicate

0.22
[0.18;0.27]

171.04

<10-4

0.09
[0.08;0.10]

AC
C

Residuals

EP

yeast

Process step

TE
D

Method

-4

SC

Standard
deviation [CI
95%]

Factors

RI
PT

Supplementary material S6: Results of the statistical test of the effect of factors on the variations in
LAB and yeast counts.

M
AN
U

1
2
3

ACCEPTED MANUSCRIPT

M
AN
U

Dim.2

AC
C

EP

TE
D

SC

RI
PT

0
Dim.1

ACCEPTED MANUSCRIPT

EP

TE
D

M
AN
U

SC

RI
PT

Supplementary Material S7: Individuals factor map on 1& 2 axes colored by steps: black = Chief Sourdough; red = final leavened dough; green =
dough after kneading; blue = dough before baking; Gray lines join the first dough replicate steps for each bread-making run.

AC
C

1
2

ACCEPTED MANUSCRIPT

TE
D

EP

AC
C

3.6 3.8 4.0 4.2 4.4 4.6 4.8

Yeast count (log10 CFU/g)

C)

Yeast count (log10 CFU/g)

7.5

7.0

SC

pH

M
AN
U

10

10

RI
PT

20

15

15

8.0

8.5

9.0

LAB count (log10 CFU/g)

3.6 3.8 4.0 4.2 4.4 4.6 4.8

TTA

B)

TTA

20

25

pH

25

A)

D)

7.0

7.5

8.0

8.5

LAB count (log10 CFU/g)

9.0

ACCEPTED MANUSCRIPT

EP

TE
D

M
AN
U

SC

RI
PT

Supplementary material S8: A) Plot of TTA according to yeast count (log10 CFU/g); B) Plot of TTA according to LAB count (log10 CFU/g); C)
Plot of pH according to yeast count (log10 CFU/g); D) Plot of pH according to LAB count (log10 CFU/g).
Plots represented the dough samples 1 and were colored according to the bakery (B) and the bread-making (BMR) run : red= B1_BMR1;
magenta = B1_BMR2; dark magenta = B2_BMR1; yellow = B2_BMR2; green = B3_BMR1; orange = B3_BMR2; pervenche = B4_BMR1;
brown = B4_BMR2.
A) Significant decreasing linear relation between yeast count and pH was determined (p-value < 10-5, estimation: -0.30). B) In 3 of 8 cases,
linear negative relation was determined between pH and LAB count within bread-making run. C) Significant increasing linear relation between
yeast count and TTA was determined (p-value < 0.04, estimation: 2.12). D) Significant increasing linear relation between LAB count and TTA
was determined (p-value < 0.05, estimation: 2.35).

AC
C

1
2
3
4
5
6
7
8
9

ACCEPTED MANUSCRIPT

M
AN
U

8.5

TE
D

7.5

AC
C

EP

8.0

7.0

LAB count (log10 CFU/g)

RI
PT

9.0

SC

9.5

Yeast count (log10 CFU/g)

ACCEPTED MANUSCRIPT

AC
C

EP

TE
D

M
AN
U

SC

RI
PT

Supplementary material S9: Plot of the LAB count according to the yeast count according to
the bread-making run (BMR) for each bakery (B): red = B1_BMR1; magenta = B1_BMR2;
dark magenta = B2_BMR1; yellow = B2_BMR2; green = B3_BMR1; orange = B3_BMR2;
light blue = B4_BMR1; brown = B 4_BMR2; gray= B5_BMR2.