Transient MutS-Based Hypermutation System for Adaptive Evolution of Lactobacillus casei to Low pH

Bacterial strains and plasmids.The bacterial strains and plasmids used in this work are listed in Table 1. L. casei and E. coli strains were maintained in a laboratory culture collection at −80°C in 15% glycerol stocks. L. casei strains were propagated without aeration at 37°C in MRS broth (Difco Laboratories, Detroit, MI) with 2.5 μg/ml erythromycin (Ery), when appropriate. E. coli strains were propagated with aeration at 200 rpm and 37°C in LB broth (Difco Laboratories), with 50 μg/ml ampicillin (Amp) or 100 μg/ml Ery, when appropriate.

Gene deletions.A list of the PCR primers used to construct recombinant plasmids for gene deletions is provided in Table 2. The isolation of L. casei genomic DNA and partial or complete gene deletions were performed using methods described previously (58). The mutS gene (2.58 kb) was deleted in its entirety from the L. casei 12A and ATCC 334 genomes. Briefly, 5′ and 3′ fragments upstream and downstream, respectively, of the mutS gene were first obtained by PCR amplification of L. casei 12A genomic DNA using the Phusion high-fidelity DNA polymerase kit (New England BioLabs, Ipswich, MA). A 960-bp 5′ fragment was obtained with primers mut5′-f and mut5′-r, which contain XbaI and PstI restriction linkers, respectively, while a 1,210-bp 3′ fragment was obtained with primers mut3′-f and mut3′-r, which contain PstI and XmaI linkers, respectively. The ends of the 5′ and 3′ PCR products were filled with Taq DNA polymerase master mix (5 PRIME, Inc., Gaithersburg, MD), separately blunt-end ligated into the cloning vector pGEM-T (Promega, Inc., Madison, WI), and transformed by electroporation into E. coli DH5α. Transformants were selected on LB with 50 μg/ml ampicillin, 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal), and isopropyl-β-d-thiogalactopyranoside (IPTG), and then plasmid DNA (pDNA) was collected from ampicillin-resistant (Amp^r) colonies and screened by restriction enzyme digestion (XbaI and PstI or PstI and XmaI) to confirm the presence of appropriately sized insert DNA. Inserts from positive clones were extracted from preparatory agarose gels, and then a triple ligation mix, including digested 5′ and 3′ inserts, was assembled into the multiple-cloning site (MCS) of the XbaI- and XmaI-digested vector pBS1 (58).

The ligated DNA was transformed into E. coli EC1000, and transformants were selected on LB with 100 μg/ml Ery. Recombinant plasmid DNA (pDNA) in erythromycin-resistant (Ery^r) colonies was screened by PCR and restriction enzyme digestion and confirmed by DNA sequencing. The recombinant plasmid, designated pBSΔmutS, was transformed as described previously (59) into L. casei 12A and ATCC 334 competent cells, prepared by growth in MRS with 1% glycine and 0.9 M NaCl, respectively, to an optical density at 600 nm (OD₆₀₀) between 0.6 and 0.8, and pretreated with sterile distilled water (dH₂O) for 30 min. Electroporation was performed in 2-mm cuvettes using a Bio-Rad Gene Pulser (Hercules, CA) set at 2,500 V, 25 μF, and 400 Ω. Electroporated cells were recovered in MRS with 0.5 M sucrose and plated onto MRS with 2.5 μg/m Ery (59).

Genomic DNA from L. casei Ery^r colonies was isolated then screened by PCR using the primers mut1 and mut3 to identify merodiploids with pBSΔmutS integrated into the wt mutS locus by a single-crossover event (which produced products for both the wt mutS [4.0-kb amplicon] and the ΔmutS [1.5-kb amplicon] loci). Merodiploids were passaged by 1% (vol/vol) transfers in MRS without antibiotic selection and plated onto brain heart infusion (BHI; Becton-Dickinson) with 7 mM 4-chloro-dl-phenylalanine (ClPhe; Sigma, St. Louis, MO), as previously described (58). To screen for colonies that had lost mutS, PCR was performed with primers mut1 and mut3. DNA sequencing from the 5′-flanking DNA upstream of the deletion site to a point downstream using primers mut1 and mut3, and from the 3′-flanking DNA downstream of the deletion site to a point upstream using primers mut2 and mut7, was used to confirm the deletion of mutS in its entirety from the genomes of L. casei 12A and ATCC 334 (58).

The L. casei 12A genes encoding a two-component signal transduction system (TCS) quorum-sensing histidine kinase (hpk [LCA12A_1895]), phosphate transport ATP-binding protein PstB (pstB [LCA12A_2227]), and NADH dehydrogenase (ndh [LCA12A_1744]) were inactivated by gene replacement using a similar approach. To generate a 1,307-bp deletion that removed the entire hpk ORF, a 550-bp 5′ fragment was obtained with primers hpk2 and hpk3, which contain BamHI and HindIII restriction linkers, respectively, and a 599-bp 3′ fragment was obtained with primers hpk4 and hpk5, which contain HindIII and XmaI linkers, respectively. The 5′ and 3′ fragments were cloned and verified in pGEM-T and assembled in pBS1, and the resulting recombinant plasmid pBSΔhpk was transformed into L. casei 12A, as described for mutS. Screening of DNA from L. casei Ery^r colonies was performed by PCR with primers hpk1 and hpk6 to identify merodiploids (which gave products for both the wt hpk [2.6-kb amplicon] and the Δhpk [1.3-kb amplicon] loci). Nonselective passage and screening for the wt hpk or Δhpk locus in isolates were performed as described for mutS with the primers hpk1 and hpk6. DNA sequencing from the 5′-flanking DNA upstream of the deletion to a point downstream from the 3′-flanking DNA was used to confirm full deletion of the hpk ORF in the L. casei 12A Δhpk mutant.

To excise the entire pstB ORF (774 bp), a 1,064-bp 5′ fragment was generated with primers pst1 and pst2, which contain XbaI and EcoRI restriction linkers, respectively, and a 1,084-bp 3′ fragment was obtained with primers pst3 and pst4, which contain EcoRI and SphI linkers, respectively. The 5′ and 3′ fragments were cloned and verified in pGEM-T and assembled in pBS1, and the recombinant plasmid pBSΔpstB was transformed into L. casei 12A, as described for mutS. Screening of DNA from L. casei Ery^r colonies for merodiploids was performed by PCR with primers pst5 and pst8 (which gave products for both the wt pstB [1-kb amplicon] and the ΔpstB [200-bp amplicon] loci). Nonselective passage and subsequent screening for the wt pstB or ΔpstB locus were performed with primers pst6 and pst7. DNA sequencing from the 5′-flanking DNA upstream of the deletion to a point downstream from the 3′-flanking DNA using primers pst1 and pst6 was used to confirm full deletion of the pstB CDS in the L. casei 12A ΔpstB mutant.

A 561-bp (amino acids 19 to 205) in-frame deletion of the ndh ORF in L. casei 12A was generated using an 824-bp 5′ fragment obtained with primers ndh2 and ndh3, which contain BamHI and HindIII restriction linkers, respectively, and a 742-bp 3′ fragment obtained with primers ndh4 and ndh5, which contain HindIII and EcoRI linkers, respectively. The 5′ and 3′ fragments were cloned into E. coli DH5α using pGEM-T, in E. coli EC1000 using pBS1, and in L. casei 12A, using the recombinant pBSΔndh, as described for mutS. Screening of DNA from L. casei Ery^r colonies was performed by PCR with the primers ndh1 and ndh8 to identify merodiploids (which gave products for both the wt ndh [1.6-kb amplicon] and the Δndh [1.1-kb amplicon] loci). Nonselective passage and subsequent screening for the wt ndh or Δndh locus were performed with the primers ndh1 and ndh8, and DNA sequencing from the 5′-flanking DNA upstream of the deletion to a point downstream from the 3′-flanking DNA confirmed that the desired in-frame deletion was present in the ndh CDS of the L. casei 12A Δndh mutant.

Finally, the hpk, pstB, and ndh genes were deleted in combination to generate a total of seven L. casei 12A double- and triple-gene-deletion mutants (Table 1). The pstB gene was inactivated in the L. casei 12A Δhpk and L. casei 12A Δndh mutants, resulting in L. casei 12A Δhpk ΔpstB and L. casei 12A Δndh ΔpstB mutants, respectively. The ndh gene was inactivated in the L. casei 12A Δhpk and L. casei 12A Δhpk ΔpstB mutants, resulting in L. casei 12A Δhpk Δndh and L. casei 12A Δhpk ΔpstB Δndh mutants, respectively.

Adaptive evolution.The approach in this study used for adaptive evolution of L. casei to low pH was based upon the work of Zhang et al. (22). The adjustment of pH in MRS broth below the standard pH 6.5 was attained by addition of 85% lactic acid (MP Biomedicals LLC, Solon, OH). Working cultures of L. casei 12A, 12A ΔmutS mutant, ATCC 334, and 334 ΔmutS mutant strains were prepared by direct inoculation of 10 ml of MRS broth from glycerol stocks with incubation at 37°C for 17 h, followed by transfer (1% [vol/vol]) into 10 ml of MRS (pH 5.5) (day 0 cultures). Cultures were incubated at 37°C and continuously transferred at mid-exponential phase (OD₆₀₀, 2.0 to 3.0) to fresh MRS (pH 5.5) over a period of 10 days, and then MRS at pH 5.0 for 15 days, MRS at pH 4.5 for 20 days, MRS at pH 4.2 for 25 days, and finally, MRS at pH 4.0 for 30 days. During the 100-day adaptive evolution process, samples were collected approximately every 5 days and preserved at −80°C as glycerol stocks, with occasional PCR screening of mutS as described for ΔmutS construction to verify culture purity during the experiment.

At the end of the adaptive evolution process, all four cultures were streaked onto MRS plates at pH 4.0 titrated with 85% lactic acid and incubated in BBL GasPak system with a GasPak EZ anaerobe container system with indicator (Becton, Dickinson and Company, Dublin, Ireland) at 37°C for 20 days. After incubation, the seven largest colonies from each culture were collected and subjected to growth studies in MRS broth at pH 4.0 (see below), and the isolate from each adaptive evolution culture with the highest growth rate at pH 4.0 was selected for further characterization.

Growth at pH 4.0.Working cultures of the seven isolates collected from each of the final four adaptive evolution cultures were prepared by a 1% (vol/vol) inoculation of 1 ml of MRS broth from glycerol stocks, with incubation at 37°C for 17 h to allow cultures to attain early stationary phase. Working cultures of L. casei ATCC 334, 334-AE, 334-MAE, 12A, 12A-AE, 12A-MAE and 12A Δhpk, 12A ΔpstB, 12A Δndh, 12A Δhpk ΔpstB, 12A Δhpk Δndh, 12A Δndh ΔpstB, and 12A Δhpk ΔpstB Δndh mutants were prepared from glycerol stocks, followed by successive 1% (vol/vol) transfer in 10 ml of MRS and 1 ml of MRS, respectively, with incubation at 37°C for 17 h. Cells were harvested from 1-ml cultures by centrifugation (14 × 10³ rpm, 5 min), washed once with 1 ml of phosphate-buffered saline (PBS), and suspended in 1 ml of MRS, adjusted to pH 4.0. These suspensions were used to inoculate (1% [vol/vol]) 0.2 ml of MRS broth (pH 4.0), in Costar 96-well flat-bottom assay plates with a low-evaporation lid (Corning, Inc., Kennebunk, ME). At least three wells were left uninoculated to serve as blanks throughout the growth curve experiments. The outer wells of the 96-well plates contained 0.2 ml of sterile dH₂O. After inoculation, the plates were capped with radiation-sterilized adhesive seals (Nunc, Roskilde, Denmark) and incubated at 37°C. The OD₆₀₀ was measured every 24 h for 17 days using a SpectraMax Plus 384 microplate reader (Molecular Devices LLC, Sunnyvale, CA) at 37°C with 1 min automixing before each measurement. Growth curves of L. casei strains were assembled using data from a minimum of two independent biological experimental replicates.

The terminal pH and d/l-lactic acid concentration were measured after the 17-day experiment for each L. casei strain tested and compared to the measured blank values. At the end of the growth experiment, cultures were transferred from wells to 1.7-ml microtubes and centrifuged at 14 × 10³ rpm for 10 min. The supernatant was transferred to 0.5-ml PCR tubes and stored at −20°C until the assay was performed. At that time, the solution was thawed, and the pH was measured using an Orion 3 Star benchtop pH meter (Thermo Fisher, Waltham, MA) with an EasyFerm Plus K8 160 pH probe (Hamilton Robotics, Reno, NV). Quantitative d/l-lactic acid analyses were performed using the R-Biopharm AG (Darmstadt, Germany) d/l-lactic acid UV-method test kit, as directed by supplier, except that the total assay volume was reduced from 1 to 0.5 ml while maintaining the proportions of reagents described in manufacturer's instructions. A Ryan-Einot-Gabriel-Welch q-test (REGWQ) for multiple pairwise comparisons was used to identify statistically significant differences in pH values and lactic acid yields from each treatment (60).

Restoration of mutS.The mutS lesion in each of the L. casei 12A ΔmutS and 334 ΔmutS acid-evolved representative isolates was repaired to restore DNA replication fidelity. A 4.8-kb fragment spanning 1 kb upstream to 1.2 kb downstream of mutS was obtained by Phusion PCR of L. casei 12A genomic DNA with the primers mut5′-f and mut3′-r, with XbaI and XmaI linkers, respectively, and cloned into E. coli DH5α using pGEM-T, and in E. coli EC1000 using pBS1, as previously described (58). The recombinant plasmid, pBS1:mutS, was transformed into L. casei 12A ΔmutS and 334 ΔmutS mutants, and screening of L. casei Ery^r colonies for merodiploids, nonselective passage, and screening for the presence of the restored mutS gene were performed essentially as described for the construction of knockout mutants. DNA sequencing from the 5′-flanking DNA upstream of mutS to a point 2.3 kb downstream, using primers mut1 and mut4, from the 3′-flanking DNA downstream of mutS to a point 3 kb upstream, using primers mut5 and mut7, and an overlapping 2.9-kb region, across mutS using primers mut6 and mut3, was used to confirm complete restoration of mutS in the genome of each strain.

Acid resistance.Working cultures of L. casei strains were prepared as described for the growth studies and then inoculated (1% [vol/vol]) into 15 ml of MRS broth and incubated at 37°C for 11 h until the cells reached early stationary phase. Cells were harvested by centrifugation (9 × 10³ × g, 10 min, 25°C), washed once with 15 ml of 0.1% Bacto peptone-buffered solution (pH 6.5), and then suspended in 1.45 ml of Bacto peptone (approximately 10¹⁰ CFU/ml). A 0.1-ml aliquot of the cell suspension was transferred to 10 ml of MRS broth without dextrose at pH 3.0, 2.5, and 2.0 (adjusted with HCl) and incubated for 1 h at 37°C. Before and after the acid challenge, cells were serially diluted and plated on MRS agar and incubated for 48 h at 37°C, and the difference in cell count was used to calculate the percent survival. Two independent biological experimental replicates of the acid resistance experiment were conducted. An REGWQ for multiple pairwise comparisons was used to identify statistically significant differences in the percent survival at each pH value.

Cell morphology.The morphologies of the L. casei 12A, 12A-MAE, 12A Δndh ΔpstB mutant, and 12A Δhpk ΔpstB mutant strains were investigated by scanning electron microscopy. Working cultures were prepared as described for adaptive evolution. Batch cultures in 1 liter of MRS with 3% glucose broth in Sartorius Biostat B-plus dual-controlled biofermenters (Sartorius AG, Göttingen, Germany) were prepared using a 1% (vol/vol) inoculum of working cultures standardized to an OD₆₀₀ of 1.0 in MRS. Cultures were incubated at 37°C with an agitation rate of 100 rpm, and when pH 3.8 was reached, it was maintained at that level by automatic addition of 15% (vol/vol) NH₄OH. A 1-ml sample of cells was collected after 30 h and centrifuged at 800 rpm for 25 min, and the supernatant was discarded. For biological fixation of the samples, 0.75 ml of buffered glutaraldehyde (2% glutaraldehyde, 0.1 M HEPES) was added to the cell pellets, followed by vortexing and incubation overnight at room temperature. Cells were transferred to glass slides (120 mm) coated with lysine, rinsed with 0.1 M HEPES three times for 5 min, and serially dehydrated with ethanol (2 times for 10 min in 50% ethanol, 2 times for 10 min in 70% ethanol, 2 times for 10 min in 95% ethanol, and then 3 times for 15 min in 100% ethanol). After rinsing, the slides were chemically dried with ethanol (100%):Bis(trimethylsilyl)amine (HMDS) solutions (2:1 for 15 min, 1:1 for 15 min, 1:2 for 15 min, and 0:1 3 times for 15 min). The HMDS in the final step was removed by evaporation overnight at room temperature, and then the fixed cells were mounted onto aluminum stubs and Au/Pd sputter-coated with an EMS-150 ES carbon/metal coater (Electron Microscopy Sciences, Hatfield, PA) as a service by the USU Microscopy Core Facility. Cells were imaged and photographed using a FEI Quanta FEG 650 scanning electron microscope (Hillsboro, OR) under high vacuum.

For analysis of cell size, ImageJ (61) was used to measure the width and length of 35 individual L. casei 12A, 12A-MAE, 12A Δndh ΔpstB mutant, and 12A Δhpk ΔpstB mutant cells, using methods described by Kokkinosa et al. (62). An REGWQ for multiple pairwise comparisons was used to identify statistically significant differences in cell length and width from each treatment.

Whole-genome resequencing.Genomic DNA was extracted using the MasterPure Gram-positive DNA purification kit (Epicentre, Madison, WI), following the provided procedure, with the addition of 5 U/ml mutanolysin from Streptomyces globisporus ATCC 21553 (Sigma) to the Ready-Lyse lysozyme solution step (37°C, 1 h). Genomic DNA was sequenced as a service by the Utah State University Center for Integrated Biosystems (USU CIB) using an Ion Torrent system (Life Technologies, Carlsbad, CA). Sequence contigs were assembled against the respective wt L. casei ATCC 334 and 12A reference genomes (GenBank accession numbers CP000423 and CP006690 , respectively) and analyzed for single-nucleotide polymorphisms (SNPs), insertions and deletions, and copy number variations using the Ion Reporter software, as a service by the USU CIB. Identified mutations were referenced against 12A population sequencing data with regions previously determined to be hypervariable in the L. casei wt genome (T. Overbeck, J. L. Steele, and J. R. Broadbent, unpublished data) and were excluded from the mutational analysis if they occurred within these regions. Mutations with a coverage depth greater than 10× and a frequency greater than 90% were considered genuine. Mutations occurring within annotated open reading frames (ORFs) were subsequently analyzed for their putative effect on translation. Six nonsense mutations identified in L. casei 12A-MAE (in hypI, srtA, pstB, hypII, hpk, and ndh genes) were verified in L. casei 12A-MAE and in 100-day L. casei 12A ΔmutS culture by sequencing PCR products generated with the primers hypI-f and hypI-r, srtA-f and srtA-r, pstB-f and pstB-r, hypII-f and hypII-r, 3hpk-f and 3hpk-r, and ndh-f and ndh-r (Table 2).