Strains, media and culture conditions. Bacterial strains used in this study are shown in Supplementary Table 7. E. coli strains were grown in lysogeny broth (LB) medium with 12.5 µg/mL kanamycin added when needed. USYE media were used during the genetic manipulation and fermentation of G. thermoglucosidasius, respectively, as described previously19,21. D-glucose and/or D-xylose were added in USYE medium as declared with the final concentrations of 2% (w/v) glucose, 2% (w/v) xylose, or 1% (w/v) glucose + 1% (w/v) xylose, respectively. The engineered G. thermoglucosidasius strains were cultivated in 250 mL shake flask with 50 mL USYE medium containing 40 mM each (final concentrations) of Bis-Tris, PIPES, and HEPES at pH 7.0 as a seed culture. Then, 0.5 mL seed culture was added to 50 mL USYE medium (Glucose and/or xylose were added as declared) and fermented for designated time at 60°C, 250 rpm.
For yeast culture, the rich medium YPD was used for the routine growth of yeast strain BJ5464-NpgA and its derivatives at 28°C43,44,60. UDMS media were used for the selection of plasmids constructed by yeast homologous recombination as previously described44. The yeast transformants were grown on the solid UDMS medium for 2 − 4 days.
Construction of plasmids and engineered strains. All plasmids used in this work are listed in Supplementary Table 7. All primers and oligonucleotides are listed in Supplementary Table 8. The genomic DNA was extracted from G. thermoglucosidasius DSM 2542 using the Ultraclean® Microbial DNA Isolation kit from Mo Bio Laboratories, Inc. (Cambio Ltd., Cambridge, UK) according to the manufacturer’s protocol. Plasmid extractions were performed using NucleoSpin Plasmid EasyPure kit (Macherey-Nagel). PCR screening for transformants were carried out with Taq 2× Master Mix (TsingKe, China).
To characterize the xylose-inducible promoters, the selected 20 putative promoters were amplified from the genome of G. thermoglucosidasius DSM 2542 using the primer pairs RS10605-F/R, RS03330-F/R, RS11585-F/R, RS19825-F/R, RS08295-F/R, RS11580-F/R, RS07185-F/R, RS05470-F/R, RS07230-F/R, RS06405-F/R, RS11780-F/R, RS01665-F/R, RS07175-F/R, RS11470-F/R, RS07200-F/R, RS16610-F/R, RS07150-F/R, RS11785-F/R, RS01820-F/R and RS03225-F/R, and assembled with the corresponding plasmid backbone amplified from pUCG18-sfgfp using the primer pairs pUCG18-RS10605-F/R, pUCG18-RS03330-F/R, pUCG18-RS11585-F/R, pUCG18-RS19825-F/R, pUCG18-RS08295-F/R, pUCG18-RS11580-F/R, pUCG18-RS07185-F/R, pUCG18-RS05470-F/R, pUCG18-RS07230-F/R, pUCG18-RS06405-F/R, pUCG18-RS11780-F/R, pUCG18-RS01665-F/R, pUCG18-RS07175-F/R, pUCG18-RS11470-F/R, pUCG18-RS07200-F/R, pUCG18-RS16610-F/R, pUCG18-RS07150-F/R, pUCG18-RS11785-F/R, pUCG18-RS01820-F/R and pUCG18-RS03225-F/R by Gibson assembly61, respectively, generating 20 plasmids pUCG18-n-sfgfp (n indicates the name of the selected 20 promoters, see Supplementary Table 7). The 20 plasmids were transformed into G. thermoglucosidasius DSM 2542 to generate the corresponding 20 engineered strains DSM2542-n (n indicates the name of the selected 20 promoters, see Supplementary Table 7).
To construct the plasmid pUCG18-RS11585-Gtribo (Supplementary Table 7), the rib cluster was amplified from the genome of DSM 2542 using the primer pair 11585Gtrib-F/R and assembled with the plasmid backbone amplified from pUCG18-RS11585-sfgfp using the primer pair 11585Gtrib-GJ-F/R by Gibson assembly61. The plasmid pUCG18-RS11585-Gtribo was transformed into DSM 2542 by electroporation to generate the strain DSM2542-DCrib as described previously19. To construct the plasmid XW55-P15, the P15 gene fragment was amplified from plasmid pUC19 using the primer pair P15-F/R, and assembled with the plasmid backbone amplified from plasmid XW5543,44 using the primer pair XW55-P15-GJ-F/R.
To activate the metabolic pathway from xylose to riboflavin using the selected three xylose-inducible promoters P11585 (S), P16610 (M) and P03225 (W), we constructed a series of plasmids named piprs-pjpurine-pkriboflavin (i, j and k indicate one of S, M and W, respectively. Supplementary Table 7) by yeast-based transformation-associated recombination43,44. Firstly, we prepared eight linear DNA fragments, consisting of (1) the fragment of 2u-ori origin of replication and URA3 amplified from plasmid XW55-P15 using primer pair JM-F/R, (2) the repBST1 origin of replication and kanamycin resistance gene from plasmid pUCG18-sfgfp using primer pair ZLGJ-F/R, (3) the prs gene amplified from the genome of DSM 2542 using primer pair prs-F/R, (4) the purine gene cluster amplified from the genome of DSM 2542 using primer pair purine-F/R, (5) the rib cluster amplified from the genome of DSM 2542 using primer pair Rb-F/R, (6) the promoters S, M and W (amplified from the genome of DSM 2542 using primer pairs S-zl-F/S-pr-R, M-zl-F/M-pr-R and W-zl-F/W-pr-R, respectively) used for linking the DNA fragments generated from (2) and (3), (7) the promoters S, M and W (amplified from the genome of DSM 2542 using primer pairs S-pr-F/S-pu-R, M-pr-F/M-pu-R and W-pr-F/W-pu-R, respectively) used for linking the DNA fragments generated from (3) and (4), (8) the promoters S, M and W (amplified from the genome of DSM 2542 using primer pairs S-pu-F/S-Rb-R, M-pu-F/M-Rb-R, and W-pu-F/W-Rb-R, respectively) used for linking the DNA fragments generated from (4) and (5). Each of the DNA fragments contained a minimum of 35 bp overlapping sequences with the flanking fragments. Then, each combination of the fragments were co-transformed into S. cerevisiae BJ5464-NpgA (MATα ura3-52 his3-Δ200 leu2-Δ1 trp1 pep4::HIS3 prb1Δ1.6R can1 GAL)44 using S. cerevisiae EasyComp Transformation Kit (Invitrogen). The correct yeast transformants were verified by PCR and DNA sequencing. Next, the plasmids were isolated from the yeast transformants using Zymoprep (D2001) Kit (Zymo Research) and transformed into E. coli Top10. The E. coli competent cells were produced and transformed according to the transformation and storage solution protocol62. Finally, the plasmids were transformed into G. thermoglucosidasius as described previously by Cripps19.
Total RNA isolation. The G. thermoglucosidasius DSM 2542 strains cultivated in liquid USYE medium were harvested by fast filtration, flash frozen in liquid nitrogen, and ground into powder for total RNA extraction using TRNzol (Tiangen, China). The integrity and quantity of the isolated RNA were checked by denaturing agarose gel electrophoresis and NanoDrop 2000 spectrophotometer (NanoDrop Technologies, USA), respectively.
RNA-Seq analysis. The cultures of G. thermoglucosidasius DSM 2542 in liquid USYE with 1% xylose or without xylose were sampled at 4 h, 6 h and 8 h, and used for massively parallel cDNA sequencing. The cDNA libraries were prepared and analyzed on Illumina HiSeq 2000. Samples were sequenced twice to obtain appropriate deep sequencing results. Raw data were processed by removing those with low quality (phred quality < 5) and sequencing adaptors. The remaining clean reads were aligned to the genome of DSM 2542. Mapping the total number of reads to each gene was implemented by Picard tools (http://picard.sourceforge.net/). The promoter strength could be reflected by RPKM, which was calculated with our previous reported formula63. The operons were predicted by DOOR 2.064 and Rockhopper65 based on the RNA-Seq data, and manually confirmed.
GFP fluorescence assay. The sfGFP18 was used as a reporter to characterize the performance of these promoters. The G. thermoglucosidasius strains carrying the GFP reporter plasmids were cultured overnight at 60°C in USYE medium with corresponding concentrations of glucose or xylose. Then, 2 µL of these cultures were inoculated into 100 µL of fresh pre-heated media in flat-bottom 96-well microtiter plate (Greiner Bio-One) and sealed airtight with VIEWSeal (In Vitro) to prevent water evaporation. The GFP fluorescence was measured using an ELx808™ Microplate Reader (BioTek) with excitation at 485 nm and emission at 535 nm. Values at the middle of log phase were taken for analysis.
To test the transcription time course of these promoters, fluorescence was measured periodically according to the method described above. To test the dose-dependent activation of these promoters, the GFP fluorescence of cells were measured at different xylose concentrations by flow cytometry (FACS) using a Becton-Dickinson FACSCalibur flow cytometer. For the FACS test, a 100 µL (culture) sample was diluted with PBS to 1 mL, and was analyzed within 10 min with excitation at 488 nm and emission at 580 nm. The cells were assayed at a low flow rate until 20,000 total events were collected. Data were analyzed with the FlowJo FC analysis software 7.6.1.
Analytical Methods. The OD at 600 nm was used to measure cell growth (1 OD600 nm = 107 cells/mL). The xylose and glucose concentrations were measured by HPLC with an Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA) and refractive index detector as descripted previously21,66. The concentration of riboflavin was determined by HPLC as descripted previously67.