Search

Understanding microbial dispersal is critical to understand the dynamics and evolution of microbial communities. However, microbial dispersal is difficult to study because of uncertainty about their vectors of migration. This applies to both microbial communities in natural and human-associated environments. Here, we studied microbial dispersal along the sourdoughs bread-making chain using a participatory research approach. Sourdough is a naturally fermented mixture of flour and water. It hosts a community of bacteria and yeasts whose origins are only partially known. We analysed the potential of wheat grains and flour to serve as an inoculum for sourdough microbial communities using 16S rDNA and ITS1 metabarcoding. First, in an experiment involving farmers, a miller and bakers, we followed the microbiota from grains to newly initiated and propagated sourdoughs. Second, we compared the microbiota of 46 sourdough samples collected everywhere in France, and of the flour used for their back-slopping. The core microbiota detected on the seeds, in the flour and in the sourdough was composed mainly of microbes known to be associated with plants and not living in sourdoughs. No sourdough yeast species were detected on grains and flours. Sourdough lactic acid bacteria were rarely found in flour. When they were, they did not have the same amplicon sequence variant (ASV) as found in the corresponding sourdough. However, the low sequencing depth for bacteria in flour did not allow us to draw definitive conclusion. Thus, our results showed that sourdough yeasts did not come from flour, and suggest that neither do sourdough LAB.

Recent studies have suggested that species of the Kazachstania genus may be interesting models of yeast domestication. Among these, Kazachstania barnettii has been isolated from various microbially transformed foodstuffs such as sourdough bread and kefir. In the present work, we sequence, assemble, and annotate the complete genomes of two K. barnettii strains: CLIB 433, being one of the two reference strains for K. barnettii that was isolated as a spoilage organism in soft drink, and CLIB 1767, recently isolated from artisan bread-making sourdough. Both assemblies are of high quality with N50 statistics greater than 1.3 Mb and BUSCO score greater than 99%. An extensive comparison of the two obtained genomes revealed very few differences between the two K. barnettii strains, considering both genome structure and gene content. The proposed genome assemblies will constitute valuable references for future comparative genomic, population genomic, or transcriptomic studies of the K. barnettii species.

International audience; Production of leavened bread dates to the second millennium BCE. Since then, the art of bread making has developed, yet the evolution of bread-associated microbial species remains largely unknown. Nowadays, leavened bread is made either by using a pure commercial culture of the yeast Saccharomyces cerevisiae or by propagating a sourdough—a mix of flour and water spontaneously fermented by yeasts and bacteria. We studied the domestication of S. cerevisiae originating from industrial sources and artisanal sourdoughs and tested whether different bread-making processes led to population divergence. We found that S. cerevisiae bakery strains are polyphyletic with 67% of strains clustering into two main clades: most industrial strains were tetraploid and clustered with strains having diverse origins, including beer. By contrast, most sourdough strains were diploid and grouped in a second clade of strains having mosaic genomes and diverse origins, including fruits and natural environments. They harbored a higher copy number of genes involved in maltose utilization, and a high level of gene flow from multiple contributors was detected. Bakery strains displayed higher CO2 production than do strains from other domesticated lineages (such as beer and wine), revealing a specific phenotypic signature of domestication. Interestingly, industrial strains had a shorter fermentation onset than sourdough strains, which were better adapted to a sourdough-like environment, suggesting divergent selection by industrial and artisanal processes. Our results reveal that the domestication of bakery yeast has been accompanied by dispersion, hybridization, and divergent selection through industrial and artisanal processes.

Leavened bread can be made with different wheat varieties and leavening agents. Several studies have now demonstrated that each of these factors can play a role in bread quality. However, their relative impact in artisanal bread making remains to be elucidated. Here, we assessed the impact of two wheat varieties as well as the impact of sourdoughs and yeasts on multiple components of bread organoleptic and nutritional quality. Using a participatory research approach including scientists and bakers, we compared breads leavened with three different sourdoughs and three different commercial yeasts as well as a mix of sourdough and yeast. Breads were made from two wheat varieties commonly used in organic farming: the variety “Renan” and the landrace “Barbu”. Except for bread minerals contents that mostly depended on wheat variety, bread quality was mostly driven by the fermenting agent. Sourdough breads had lower sugar and organic acids contents. These differences were mostly attributable to lower amounts of maltose and malate. They also had a higher proportion of soluble proteins than yeast breads, with specific aroma profiles. Finally, their aroma profiles were specific and more diverse compared to yeast breads. Interestingly, we also found significant nutritional and organoleptic quality differences between sourdough breads. These results highlight the value of sourdough bread and the role of sourdough microbial diversity in bread nutritional and organoleptic quality.

SummaryDespite bread being one of the most historically and culturally important fermented products, its history and influence on the evolution of associated microbial species remains largely unknown. The first evidence of leavened bread dates to the second millenium BCE in Egypt and since, the art of bread-making developed and spread worldwide. Nowadays, leavened bread is made either by using a pure commercial culture of the yeastSaccharomyces cerevisiaeor by propagating a sourdough, which is a mix of flour and water spontaneously fermented by yeast and bacteria. We studied the domestication ofS. cerevisiaepopulations originating from industry and sourdough and tested whether these different bread-making processes led to population divergence. We found that the origin ofS. cerevisiaebakery strains is polyphyletic with 67 % of strains clustering in two main clades: most commercial strains were tetraploid and clustered with strains having diverse origins, including beer. By contrast, most sourdough strains were diploids and found in a second clade of strains having mosaic genomes and diverse origins including fruits, or clinical and wild environments. When compared to the others, sourdough strains harboured in average a higher copy number of genes involved in maltose utilization, a common sugar produced from dough starch. Overall, a high level of gene flow from multiple contributors was detected. Phenotyping of bakery and non bakery strains further showed that sourdough and industrial bakery populations have undergone human selection for rapidly starting fermentations and for high CO2production. Interestingly, sourdough strains also showed a better adaptation to a sourdough mimicking environment, suggesting that natural selection occurred as well. In summary, our results revealed that the domestication of bakery yeast populations has been accompanied by dispersion, hybridization and divergent selection through industrial and artisanal bakery processes. In addition, they unveiled for the first time a case of fungus domestication where species divergence occurred through autotetraploidisation.

International audience; The metabolism of ferulic acid (FA) was studied during fermentation with different species and strains of lactic acid bacteria (LAB) and yeasts, in synthetic sourdough medium. Yeast strains of Kazachstania humilis, Kazachstania bulderi, and Saccharomyces cerevisiae, as well as lactic acid bacteria strains of Fructilactobacillus sanfranciscensis, Lactiplantibacillus plantarum, Lactiplantibacillus xiangfangensis, Levilactobacillus hammesii, Latilactobacillus curvatus and Latilactobacillus sakei were selected from French natural sourdoughs. Fermentation in presence or absence of FA was carried out in LAB and yeasts monocultures, as well as in LAB/yeast co-cultures. Our results indicated that FA was mainly metabolized into 4-vinylguaiacol (4-VG) by S. cerevisiae strains, and into dihydroferulic acid (DHFA) and 4-VG in the case of LAB. Interactions of LAB and yeasts led to the modification of FA metabolism, with a major formation of DHFA, even by the strains that do not produce it in monoculture. Interestingly, FA was almost completely consumed by the F. sanfranciscensis bFs17 and K. humilis yKh17 pair and converted into DHFA in 89.5 ± 19.6% yield, while neither bFs17, nor yKh17 strains assimilated FA in monoculture

Sourdoughs harbor simple microbial communities usually composed of a few prevailing lactic acid bacteria species (LAB) and yeast species. However, yeast and LAB found in sourdough have been described as highly diverse. Even if LAB and yeast associations have been widely documented, the nature of the interactions between them has been poorly described. These interactions define the composition and structure of sourdough communities, and therefore, the characteristics of the final bread product. In this study, the nature of the interactions between strains of two commonly found sourdough yeast species, Kazachstania humilis and Saccharomyces cerevisiae, and lactic acid bacteria isolated from sourdoughs has been analyzed. Population density analysis showed no evidence of positive interactions, but instead revealed neutral or negative asymmetric interaction outcomes. When in coculture, the yeasts´ population size decreased in the presence of LAB regardless of the strain, while the LAB´s population size was rarely influenced by the presence of yeasts. However, a higher maltose depletion was shown in maltose-negative K. humilis and maltose-positive obligately heterofermentative LAB cocultures compared to monocultures. In addition, tested pairs of obligately heterofermentative LAB and K. humilis strains leavened dough as much as couples of LAB and S. cerevisiae strains, while K. humilis strains never leavened dough as much as S. cerevisiae when in monoculture. Taken together, our results demonstrate that even if higher fermentation levels with increased maltose depletion were detected for K. humilis and obligately heterofermentative LAB pairs, these interactions cannot be ecologically classified as positive, leading us to rethink the established hypothesis of coexistence by facilitation in sourdoughs.

Iron is an essential element for most living organisms, and pathogens are likely to compete with their hosts for the acquisition of this element. The bacterial plant pathogen Dickeya dadantii has been shown to require its siderophore-mediated iron uptake system for systemic disease progression on several host plants, including Arabidopsis thaliana. In this study, we investigated the effect of the iron status of Arabidopsis on the severity of disease caused by D. dadantii. We showed that symptom severity, bacterial fitness and the expression of bacterial pectate lyase-encoding genes were reduced in iron-deficient plants. Reduced symptoms correlated with enhanced expression of the salicylic acid defence plant marker gene PR1. However, levels of the ferritin coding transcript AtFER1, callose deposition and production of reactive oxygen species were reduced in iron-deficient infected plants, ruling out the involvement of these defences in the limitation of disease caused by D. dadantii. Disease reduction in iron-starved plants was also observed with the necrotrophic fungus Botrytis cinerea. Our data demonstrate that the plant nutritional iron status can control the outcome of an infection by acting on both the pathogen's virulence and the host's defence. In addition, iron nutrition strongly affects the disease caused by two soft rot-causing plant pathogens with a large host range. Thus, it may be of interest to take into account the plant iron status when there is a need to control disease without compromising crop quality and yield in economically important plant species.

INRA, MIG, Domaine de Vilvert, F-78352 Jouy en Josas, France : depuis janvier 2015 : INRA, MaIAGE, Domaine de Vilvert, F-78352 Jouy en Josas, France; Iron is an essential element for bacterial growth and virulence. Because of its limited bioavailability in the host, bacteria have adapted several strategies to acquire iron during infection. In the human opportunistic bacteria Bacillus cereus, a surface protein IlsA is shown to be involved in iron acquisition from both ferritin and hemoproteins. IlsA has a modular structure consisting of a NEAT (Near Iron transporter) domain at the N-terminus, several LRR (Leucine Rich Repeat) motifs and a SLH (Surface Layer Homology) domain likely involved in anchoring the protein to the cell surface.METHODS:[b]BACKGROUND:[/b]Isothermal titration calorimetry, UV-Vis spectrophotometry, affinity chromatography and rapid kinetics stopped-flow measurements were employed to probe the binding and transfer of hemin between two different B. cereus surface proteins (IlsA and IsdC).[br/][b]RESULTS:[/b]IlsA binds hemin via the NEAT domain and is able to extract heme from hemoglobin whereas the LRR domain alone is not involved in these processes. A rapid hemin transfer from hemin-containing IlsA (holo-IlsA) to hemin-free IsdC (apo-IsdC) is demonstrated.[br/][b]CONCLUSIONS[/b]:For the first time, it is shown that two different B. cereus surface proteins (IlsA and IsdC) can interact and transfer heme suggesting their involvement in B. cereus heme acquisition.[br/][b]GENERAL SIGNIFICANCE:[/b]An important role for the complete Isd system in heme-associated bacterial growth is demonstrated and new insights into the interplay between an Isd NEAT surface protein and an IlsA-NEAT-LRR protein, both of which appear to be involved in heme-iron acquisition in B. cereus are revealed.

Ferritins are multimeric iron storage proteins encoded by a four-member gene family in Arabidopsis (AtFer1-4). To investigate whether iron sequestration in ferritins is a part of an iron-withholding defense system induced in response to bacterial invasion, we used Arabidopsis thaliana as a susceptible host for the pathogenic bacterium Erwinia chrysanthemi. In this study, we used a T-DNA insertion mutant line to show that the lack of a functional AtFer1 gene resulted in an enhanced susceptibility of Arabidopsis plants to E. chrysanthemi. We found that the AtFer1 gene is upregulated during infection, with a biphasic accumulation of the transcript at critical time points 0.5 and 24 h post-infection (p.i.). The activation of AtFer1 expression observed at 24 h p.i. was independent of the iron-dependent regulatory sequence (IDRS) known to mediate the transcriptional response of the AtFer1 gene to iron excess and to nitric oxide. Upregulation of AtFer1 gene expression was compromised after inoculation with an E. chrysanthemi siderophore null mutant. Infiltration of the purified siderophores chrysobactin and desferrioxamine strongly increased AtFer1 transcript abundance and it did not occur with the iron-loaded forms of these siderophores. We found that neither oxidative stress nor nitric oxide was involved in the plant response to chrysobactin. Our data show that ferritin accumulation during infection of Arabidopsis by E. chrysanthemi is a basal defense mechanism which is mainly activated by bacterial siderophores. The potential role of siderophores in this process is discussed.

Iron is an essential element for most living organisms, and pathogens are likely to compete with their hosts for the acquisition of this element. The bacterial plant pathogen Dickeya dadantii has been shown to require its siderophore‐mediated iron uptake system for systemic disease progression on several host plants, including Arabidopsis thaliana. In this study, we investigated the effect of the iron status of Arabidopsis on the severity of disease caused by D. dadantii. We showed that symptom severity, bacterial fitness and the expression of bacterial pectate lyase‐encoding genes were reduced in iron‐deficient plants. Reduced symptoms correlated with enhanced expression of the salicylic acid defence plant marker gene PR1. However, levels of the ferritin coding transcript AtFER1, callose deposition and production of reactive oxygen species were reduced in iron‐deficient infected plants, ruling out the involvement of these defences in the limitation of disease caused by D. dadantii. Disease reduction in iron‐starved plants was also observed with the necrotrophic fungus Botrytis cinerea. Our data demonstrate that the plant nutritional iron status can control the outcome of an infection by acting on both the pathogen's virulence and the host's defence. In addition, iron nutrition strongly affects the disease caused by two soft rot‐causing plant pathogens with a large host range. Thus, it may be of interest to take into account the plant iron status when there is a need to control disease without compromising crop quality and yield in economically important plant species.

Siderophores (ferric ion chelators) are secreted by organisms in response to iron deficiency. The pathogenic enterobacterium Erwinia chrysanthemi produces two siderophores, achromobactin and chrysobactin (CB), which are required for systemic dissemination in host plants. Previous studies have shown that CB is produced in planta and can trigger the up-regulation of the plant ferritin gene AtFER1. To further investigate the function of CB during pathogenesis, we analyzed its effect in Arabidopsis (Arabidopsis thaliana) plants following leaf infiltration. CB activates the salicylic acid (SA)-mediated signaling pathway, while the CB ferric complex is ineffective, suggesting that the elicitor activity of this siderophore is due to its iron-binding property. We confirmed this hypothesis by testing the effect of siderophores structurally unrelated to CB, including deferrioxamine. There was no activation of SA-dependent defense in plants grown under iron deficiency before CB treatment. Transcriptional analysis of the genes encoding the root ferrous ion transporter and ferric chelate reductase, and determination of the activity of this enzyme in response to CB or deferrioxamine, showed that these compounds induce a leaf-to-root iron deficiency signal. This root response as well as ferritin gene up-regulation in the leaf were not compromised in a SA-deficient mutant line. Using the Arabidopsis-E. chrysanthemi pathosystem, we have shown that CB promotes bacterial growth in planta and can modulate plant defenses through an antagonistic mechanism between SA and jasmonic acid signaling cascades. Collectively, these data reveal a new link between two processes mediated by SA and iron in response to microbial siderophores.

International audience; AtNRAMP3 and AtNRAMP4 are two Arabidopsis metal transporters sharing about 50% sequence identity with mouse NRAMP1. The NRAMP1/Slc11A1 metal ion transporter plays a crucial role in the innate immunity of animal macrophages targeted by intracellular bacterial pathogens. AtNRAMP3 and AtNRAMP4 localize to the vacuolar membrane. We found that AtNRAMP3 is upregulated in leaves challenged with the bacterial pathogens Pseudomonas syringae and Erwinia chrysanthemi, whereas AtNRAMP4 expression is not modified. Using single and double nramp3 and nramp4 mutants, as well as lines ectopically expressing either of these genes, we show that AtNRAMP3 and, to a lesser extent, AtNRAMP4 are involved in Arabidopsis thaliana resistance against the bacterial pathogen E. chrysanthemi. The susceptibility of the double nramp3 nramp4 mutant is associated with the reduced accumulation of reactive oxygen species and ferritin (AtFER1), an iron storage protein known to participate in A. thaliana defense. Interestingly, roots from infected plants accumulated transcripts of AtNRAMP3 as well as the iron-deficiency markers IRT1 and FRO2. This finding suggests the existence of a shoot-to-root signal reminiscent of an iron-deficiency signal activated by pathogen infection. Our data indicate that the functions of NRAMP proteins in innate immunity have been conserved between animals and plants.

Ferritins are multimeric iron storage proteins encoded by a four-member gene family in Arabidopsis (AtFer1-4). To investigate whether iron sequestration in ferritins is a part of an iron-withholding defense system induced in response to bacterial invasion, we used Arabidopsis thaliana as a susceptible host for the pathogenic bacterium Erwinia chrysanthemi. In this study, we used a T-DNA insertion mutant line to show that the lack of a functional AtFer1 gene resulted in an enhanced susceptibility of Arabidopsis plants to E. chrysanthemi. We found that the AtFer1 gene is upregulated during infection, with a biphasic accumulation of the transcript at critical time points 0.5 and 24 h post-infection (p.i.). The activation of AtFer1 expression observed at 24 h p.i. was independent of the iron-dependent regulatory sequence (IDRS) known to mediate the transcriptional response of the AtFer1 gene to iron excess and to nitric oxide. Upregulation of AtFer1 gene expression was compromised after inoculation with an E. chrysanthemi siderophore null mutant. Infiltration of the purified siderophores chrysobactin and desferrioxamine strongly increased AtFer1 transcript abundance and it did not occur with the iron-loaded forms of these siderophores. We found that neither oxidative stress nor nitric oxide was involved in the plant response to chrysobactin. Our data show that ferritin accumulation during infection of Arabidopsis by E. chrysanthemi is a basal defense mechanism which is mainly activated by bacterial siderophores. The potential role of siderophores in this process is discussed.