Strains and media
Thermobacillus xylanilyticus XE9/11/91 isolated from a farm soil under a manure heap in northern France was used in this study. The bacterium was cultivated on basal medium composed by three different solutions, a macro-mineral solution, a vitamin solution, and a metallic trace solution complemented with NH4Cl (1 g/L), yeast extract (2 g/L), NaHCO3 and supplemented with 10% CO2 as previously described by . Cultivation volumes are 10 mL or 50 mL of media in sealed contents (100 mL or 500 mL bottles). Various carbon sources were used: xylan from beechwood 5 g/L (Roth), glucose 5 g/L (Sigma Aldrich) or destarched wheat bran 10 g/L (ARD Pomacle-Bazancourt, France).
Successive cultivations of Thermobacillus xylanilyticus
The strain was regenerated from glycerol stock and an overnight non sporulated preculture (OD 600 nm between 1.5 and 2) was prepared on glucose basal medium at 50°C and 150 RPM in glucose basal medium with Multitron shakers (Infors). The xylan media (in 50 mL volume) were inoculated with the preculture to reach a start optical density of 0.1 for the cultivations start. The cultivation was done at 50°C, 150 RPM during 5 hours in xylan basal medium. After 5 hours, a new cultivation was started from the previous one. The successive cultivation was performed every 5 hours (Figure 9). Each 4 cultivations (G20 or generation 23.5, G50 or generation 50.8, G80 or generation 80.3 and G100 or generation 107.7 for xylan), a following of the growth was done by measuring the OD 600 nm. The growth rate was calculated during the exponential phase with the formula: µmax = ln (N2/N1)/t2-t1 with N representing the bacterial population (here the optical density measured) and t the cultivation time. The generation numbers (n) were directly calculated with the formula: n=T/g where g represents the generation time (g=ln2/µmax) and T the total cultivation time. Successive cultivations experiment was performed in triplicate for 17 serial transfers (corresponding to more than 100 generations).
For the followed generations, population analyses were also done by cytometry, xylanase activity production and the xylanase (tx-xyn11) gene expression measured. Figure 9 summarizes the main experiments performed during the successive cultivations.
Measurements of the xylanase enzymatic activity
To evaluate the xylanase activity secreted by T. xylanilyticus, for each followed generation, new 6h (to reach the early stationary phase) cultivations on xylan in 10 mL were prepared. At the end of the cultivations, samples were centrifuged at 3354 × g for 10 minutes (Sorvall ST 8R centrifuge, Thermo Fisher Scientific) at 4°C and the supernatant was recovered. The xylanase activity was determined in triplicate by using the reducing end sugars measurement according to the procedure described by .
Briefly, 0.1 mL of the supernatants (diluted or not) are incubated in 0.9 mL beechwood xylan (Roth) at 0.5% w/v homogeneously suspended in 50 mM sodium acetate buffer (pH of 5.8) at 60°C for 10 minutes. The reducing sugars were measured by following the absorbance at 420 nm on a Specord 200 Plus uv/vis spectrophotometer (Analytik Jena) and by comparing the values with a standard curve done with varying concentrations of xylose. The activity was expressed in IU/mL. One international unit (IU) of enzyme activity was defined as the quantity of enzyme (xylanase) required to liberate one µmol of equivalent xylose per minute at 60°C.
The IU values were normalized with the quantity (in mg) of total protein in the supernatants. The concentrations of total protein in the samples were determined by the Bradford procedure  with a commercial reagent 5× (Serva) as recommended by the supplier. After normalization, the activity was expressed in IU by milligrams of total proteins (IU/mg).
Measurements of the tx-xyn11 gene expression
The cultivation samples for gene expression (5 mL) were centrifuged at 3354 × g for 10 minutes (Sorvall ST 8R centrifuge, Thermo Fisher Scientific) at 4°C. Cell pellets were kept in 1.5 mL of RNAlater at -80°C before the analyses. The cells pellets were used for RNA extraction with RNeasy Mini Kit (Qiagen) by following the supplier recommendations. The obtained RNA solutions were treated with RNase-Free DNase Set (Qiagen) to prevent the presence of DNA in the samples. Absence of DNA was verified by polymerase-chain reaction with DreamTaq™ Hot Start Green PCR Master Mix (Thermo Fisher Scientific) by using specific primers for tx-xyn11 gene, followed by electrophoresis migrations on RNase-free TBE buffer 1× agarose gel 1% (Mupid® One Electrophoresis System, Eurogentec). The RNA concentrations were determined using a QubitTM Fluorometer and Qubit® RNA Assay Kit (Invitrogen).
First strand cDNAs were synthesized by using SuperScript® IV First-Strand cDNA Synthesis Reaction kit (Thermo Fisher Scientific) with 500 ng of total RNA in presence of random hexamer primers (50 µM), DNTP mix (0.5 mM of each), 1× SuperScript IV buffer, DTT (100 mM), RNAseOUTTM Recombinant RNase Inhibitor (2 U/µL) and 800 U (or 40 U/µL) of Superscript IV reverse transcriptase in a final volume of 20 µL. The synthesis was done at 23°C for 10 minutes followed by a step at 52.5°C for 10 minutes. An inactivation was done at 80°C for 10 minutes. The generated cDNAs were kept at -80°C until utilization.
Quantitative PCR was realized with a QuantStudio™ 3 Real-Time PCR System (Applied BiosystemsTM) to determine the tx-xyn11 gene expression. Specific primers (forward primer: GACGGCACGCAGACGTTCCA, reverse primer: GCCTTCGGTTGCGAGCACCT) previously described  was used and yielded a specific 162 bp long product. Before utilization, the specificity of the primers was tested and confirmed.
The amplifications were performed in 15 µL final volume containing 7.5 µL of Absolute Blue qPCR SYBR green low ROX mix (Thermo Fisher Scientific), 1.4 µL of primers mix (280 nM), 1.1 µL of DNase/RNase-free water and 5 µL template cDNA (10-fold diluted).
The PCR program was the following: an initial denaturation of 95°C for 15 s followed by 40 cycles of 95°C for 10 s, 60°C for 45 s with a single fluorescence measurement before an elongation step of 72°C for 30 s. The specificity of the PCR products was confirmed by melting curve analysis (after a step at 95°C for 15 s, the melting curve analysis was performed between 60–95°C with a heating rate of 0.1 C°/s). Different non-template controls and positive controls (with genomic DNA of T. xylanilyticus) were also included to confirm the specificity of the reactions. The tx-xyn11 transcripts were normalized by using the expression of T. xylanilyticus 16S rRNA gene amplified by using the specific primers (forward primer: CGCGAGCGACGCAATCCCA, reverse primer: CGGTTACCCCACCGGCTTCG). For the relative expression of tx-xyn11 gene, calculation was done using the 2−∆∆Ct method . For the quantification of tx-xyn11 transcripts over generations, the expressions of tx-xyn11 at the beginning of the successive cultivations (generation 0) were used as reference samples. In order to compare, the expression of tx-xyn11 gene between glucose and xylan successive cultivations, the reference sample was defined as the first cultivation on glucose. The results were expressed as the fold increase of mRNA level over the reference samples. Quantitative PCR was realized in triplicates for each sample of successive cultivations.
Population analyses by flow cytometry
Population analyses at the different generations described above (for the successive cultivations) were done by flow cytometry with a BD Accuri™ C6 (BD Biosciences) using the MOBICYTE core facilities (University of Reims Champagne-Ardenne). The analytical parameters were flow rate of 35 µL/minute, core size of 16 µm and the threshold were down to 30000 in FSC-H. 20 000 cells were collected for each sample.
Samples taken at different generations (5 mL) were filtered with 5 µm cellulose syringe filters and diluted with PBS 1× to reach maximum 2500 detected events/s (and events/µL) to prevent doublet reading. Samples were analyzed first unstained, and populations were detected by the forward and scatter signals (respectively FSC-A and SSC-A) to determine cell percentage of each detected population with a gating at 30 000 in FSC-A. The BD CSampler Software was used to acquire and treat the cytometric data.
Samples were also analyzed after staining with BacLight™ RedoxSensor™ Green (RSG, 1 µM) or propidium iodide (PI, 10 mg/L) (Invitrogen) to detect the metabolic activity and the membrane permeability respectively [43, 44]. After 488nm laser excitation, green fluorescence were collected at 525 nm ± 30 nm for RSG (FL1-A) and the red fluorescence signal at 670LP (FL3-A) for PI.
Cells were also analyzed at different growth phases (exponential phase, stationary phase and sporulation) on different carbon sources (glucose, xylan and destarched wheat bran) with the same cytometric parameters.
The cell sorting was realized from T. xylanilyticus cultivated on xylan basal medium at the generation 0 but also along generations (G20 and G50). The cell sorting was performed with a BD FACSAria™ II Cell Sorter coupled with the BD Accuri™ C6 Plus flow cytometer from the URCACyt technical platform facilities.
In order to standardize the signals of the BD Accuri™ C6 Plus flow cytometer and BD FACSAria™ II cell sorter, an analysis of 2.5 µm microbeads (BD Biosciences) was done with both systems.
5 × 106 events were collected for each population and the rate of sorting was around 2500 events/second with a 70 µm nozzle. Each cell sorting was performed in triplicate. The events were collected in 10 mL of PBS 1×. To check for a correct cell sorting, an analysis of each population collected was done with BD Accuri™ C6 plus flow cytometer. Each population solution was then centrifuged at 12108 × g (Sorvall ST 8R centrifuge, Thermo Scientific) for 30 minutes at room temperature. The cell pellet conserved in 100 µL of 1× PBS was solubilized with 1 mL of xylan basal medium. All the cell solution was used to inoculate a new cultivation on xylan and characterize the separated populations.
Cultivations of the sorted populations
Population cultivations were performed as described in the previous parts for 19 hours. The growth was followed with regular measurement of the OD600 nm. At the different growth phases determined (lag, beginning and end of the exponential phase and stationary phase), samples of 5 mL were taken and centrifuged at 3354 × g for 10 minutes (Sorvall ST 8R centrifuge, Thermo Scientific). The supernatants were used for the measurements of xylanase activity and protein contents according to the procedures described above.
Scanning electron microscopy analyses
During the followed successive cultivations, samples were taken before and after the cell sorting for analyses. Before the cell sorting, drops of 20 µL were dried on glass coverslip. After the cell sorting, 10 mL of different sorted populations (at 500 cells by µL corresponding to 5 × 106) were concentrated by centrifugation at 16000 × g for 40 minutes. The cell pellets were solubilized in 500 µL and 22.5 µL drops were dried on glass slide (corresponding to 225000 cells).
After drying, glass slides were washed 2-times in 1× PBS, then fixed in 2.5% (W/V in PBS) glutaraldehyde (Sigma-Aldrich) at room temperature for 1 hour. After 2 distilled water washing, cells were dehydrated at room temperature in graded ethanol solutions (50, 70, 90, and twice with absolute ethanol) for 10 minutes and in a solution of ethanol (100%)/hexamethyldisilazane (V/V) for 5 minutes. Glass slides were finally desiccated with one final drop of HMDS. After air-drying at room temperature, samples were sputtered with a thin gold-palladium film using a JEOL ion sputter JFC 1100 instrument. Samples were then observed using a Schottky Field Emission Scanning Electron Microscope (JEOL JSM-7900F).
Successive cultivation on glucose
Successive cultivations were also performed by using glucose as carbon source during 80 generations. For the generations 23.9, 42.7, 62.9 and 79.9 (due to a different growth rate than on xylan), a switch of the carbon source between glucose and xylan was performed and the same analyses (flow cytometry and xylanase activity) as those for xylan successive cultures were done.