Engineering Yeast Hexokinase 2 for Improved Tolerance Toward Xylose-Induced Inactivation

Hexokinase 2 (Hxk2p) from Saccharomyces cerevisiae is a bi-functional enzyme being both a catalyst and an important regulator in the glucose repression signal. In the presence of xylose Hxk2p is irreversibly inactivated through an autophosphorylation mechanism, affecting all functions. Consequently, the regulation of genes involved in sugar transport and fermentative metabolism is impaired. The aim of the study was to obtain new Hxk2p-variants, immune to the autophosphorylation, which potentially can restore the repressive capability closer to its nominal level. In this study we constructed the first condensed, rationally designed combinatorial library targeting the active-site in Hxk2p. We combined protein engineering and genetic engineering for efficient screening and identified a variant with Phe159 changed to tyrosine. This variant had 64% higher catalytic activity in the presence of xylose compared to the wild-type and is expected to be a key component for increasing the productivity of recombinant xylose-fermenting strains for bioethanol production from lignocellulosic feedstocks.


Construction of the E. coli library of Hxk2p-variants
Cloning of the HXK2 locus The locus was amplified using primers HXK2_loc_f (5'-GCT TGC ATG CAC GCC ATA GAA GAG CAA TTT CCG TCC-3') and HXK2_loc_r (5'-CCG GGG ATC CGA GAG GGT TAA AAT TGG CGT GCA ATT TTA TGA AG-3'). The high fidelity Phusion Hotstart II polymerase (Thermo Scientific, USA) was used with the following PCR program: 30 s initial denaturation at 98°C, 30 cycles of 10 s denaturation at 98°C, 30 s annealing at 65°C and 1 min elongation at 72°C, and a final 10 min elongation step at 72°C. The resulting DNA fragment was digested with BamHI and SphI (FastDigest, Thermo Scientific, USA) at 37°C for 30 min and ligated into YIplac128, linearized with the same restriction enzymes. The ligation system was subsequently used to transform E. coli NEB5α.
Construction of the mutated HXK2 megaprimer The strategy used to construct the mutated HXK2 megaprimer is outlined in Figure S1 and the primers are listed in Table S3.  Figure S1. Construction of the mutated HXK2 megaprimer Unless stated otherwise the PCR reactions contained 1 HF Buffer, 0.2 mM of each dNTP, 0.01 U µL -1 Phusion Hotstart DNA Polymerase II and 0.5 µM each of forward and reverse primers. The reaction volume was 50 µL. OE-PCR reactions contained 30 fmol of each fragment. DNA fragments amplified by PCR were purified using GeneJet PCR Purification Kit (Thermo Scientific, USA) unless stated otherwise.
PCR I: Mutated fragments were amplified from YIpBB5 using the following PCR program: 30 s denaturation at 98°C, 30 cycles of 10 s denaturation at 98°C, 5 s annealing and extension at 72°C and 1 cycle of 10 min extension at 72°C. Primers were used according to the following scheme: Fragment 2: Group_1_f and Group_2_r Fragment 3: Group_2_f and Group_3_r Fragment 4: Group_3_f and Group_4_r Fragment 5: Group_4_f and Group_5_r Fragment 6: Group_5_f and Group_6_r PCR II: The non-mutated fragments were amplified from YIpBB5 using the following PCR program: 30 s denaturation at 98°C, 20 cycles of 10 s denaturation at 98°C, 15 s annealing at 72°C (-0.5°C cycle -1 ), 30 s extension at 72°C, 16 cycles of 10 s denaturation at 98°C, 15 s annealing at 62.4°C, 30 s extension at 72°C and 1 cycle of 10 min extension at 72°C. Primers were used according to the following scheme: Fragment 1: Yip128-F1 and Group_1_r Fragment 7: Group_6_f and Yip128-R1 OE-PCR I: The two fragments were joined using 1 GC buffer in the reaction mix and the following PCR program: 30 s denaturation at 98°C, 16 cycles of 10 s denaturation at 98°C, 30 s annealing at 70°C (-0.5°C cycle -1 ) and 30 s extension at 72°C. At this point the reaction was halted at 4°C until primers had been added after which the program was repeated. The program was finished with a final 10 min extension at 72°C. The primers were added to a final concentration of 0.52 µM according to the following scheme: Fragment 2+3: Group_1_f and Group_3_r Fragment 3+4: Group_2_f and Group_4_r Fragment 4+5: Group_3_f and Group_5_r Fragment 5+6: Group_4_f and Group_6_r OE-PCR II: The two fragments were joined using the same procedure as in OE-PCR I. The primers were added to a final concentration of 0.52 µM according to the following scheme: Fragment 2+3+4: Group_1_f and Group_4_r Fragment 4+5+6: Group_3_f and Group_6_r OE-PCR III: The two fragments were joined using the same procedure as in OE-PCR I. The primers Group_1_f Group_6_r and were added to a final concentration of 0.52 µM.
OE-PCR IV: The three fragments were joined using the following PCR program: 30 s denaturation at 98°C, 16 cycles of 10 s denaturation at 98°C, 30 s annealing at 70°C (-0.5°C cycle -1 ), 30 s extension at 72°C and 1 cycle of 5 min extension at 72°C. At this point the reaction was halted at 4°C until primers had been added after which the program was repeated. The program was finished with a final 10 min extension at 72°C. The primers Yip128-F1 and Yip128-R1 were added to a final concentration of 0.52 µM.
The final megaprimer was purified from 0.8% agarose gel using the QIAquick Gel Extraction Kit (Qiagen, Germany).
Generating the plasmid library using the MEGAWHOP procedure The MEGAWHOP reaction mixture (50 µL) contained the following: 1 HF Buffer, 0.2 mM of each dNTP, 503 ng of HXK2 megaprimer, 50.1 ng of template plasmid YIpBB5 and 0.02 U µL -1 Phusion Hotstart DNA Polymerase II. The whole plasmid amplification was performed using the following PCR program: 30 s denaturation at 98°C, 25 cycles of 10 s denaturation at 98°C, 15 s annealing at 58°C and 3 min extension at 72°C.
After completing the PCR reaction, 0.8 µL DpnI (20 U µL -1 ) was added to 40 µL of the MEGAWHOP reaction mixture and incubated for 2 h at 37°C. AT the same time a negative control (40 µL) containing 1 HF Buffer and 1 ng µL -1 YIpBB5 was treated equally. 2 µL and 5 µL of DpnI-treated MEGAWHOP reaction was used to transform commercial heat shock competent E. coli NEB5α (High Efficiency, New England Biolabs, USA) according to the suppliers instructions. Transformants were selected on solid LB-medium supplemented with 100 mg L -1 ampicillin. Both transformations generated ca. 7700 cfu mL -1 . 5 µL of the negative control did not result in any transformants. Five additional transformation reactions of NEB5α were performed generating a total E. coli library of 57,370  3261 cfu. These transformants were inoculated in 100 mL of LB-medium supplemented with 100 mg mL -1 of ampicillin and grown for 16 h at 37°C. The resulting culture was aliquoted in 25% glycerol and stored at -80°C. Part of the culture was used to purify and sequence the plasmids. This confirmed that all mutations were present in the plasmid mix (Fig. S2). Region 2 Supporting Figure S2. Sequencing of the HXK2-library The nucleotide sequences of the mutated regions are shown to the left. The top sequence is the native sequence and the bottom sequence contains the introduced degeneracy. Codons that were modified are underlined and the native and alternative amino acid residues are indicated with bold and white-onblack letters, respectively. The right panel shows the corresponding region from the electropherogram (note that Region 6 is shown as the reverse complement). Arrows indicate the point of the degeneracy and the dual signals show that the mutations are indeed introduced in the megaprimer.

Construction of strains TMB3460, TMB3461 and TMB3462
Unless stated otherwise the PCR reactions contained 1 HF Buffer, 0.2 mM of each dNTP, 0.01 U µL -1 Phusion Hotstart DNA Polymerase II and 0.5 µM each of forward and reverse primers (Table S4). The reaction volume was 50 µL. OE-PCR reactions contained 200 fmol of each fragment. DNA fragments amplified by PCR were purified from 0.8% agarose gel using the QIAquick Gel Extraction Kit (Qiagen, Germany) unless stated otherwise.
Verification of correct integration and gene deletion was performed using the primers listed in Table S5 and the following reaction mix: 1 DreamTaq Buffer, 0.2 mM of each dNTP, 0.3 µM each of forward and reverse primers and 1 U µL -1 DreamTaq DNA Polymerase.
Construction of TMB3460 (hxk2-∆) The upstream and downstream fragments flanking the HXK2 gene were amplified from genomic DNA from S. cerevisiae CEN.PK2-1C using the following PCR program: 30 s denaturation at 98°C, 30 cycles of 10 s denaturation at 98°C, 30 s annealing at 65°C (-0.5°C cycle -1 ), 15 s extension at 72°C and 1 cycle of 10 min extension at 72°C.
The auxotrophic marker cassette TRP1 was amplified from plasmid p424 using the following PCR program: 30 s denaturation at 98°C, 30 cycles of 10 s denaturation at 98°C, 30 s annealing at 65°C (-0.5°C cycle -1 ), 1 min extension at 72°C and 1 cycle of 10 min extension at 72°C.
The deletion cassette HXK2_US-TRP1-HXK2_DS was created by OE-PCR using the following PCR program: 30 s denaturation at 98°C, 16 cycles of 10 s denaturation at 98°C, 30 s annealing at 68°C (-0.5°C cycle -1 ) and 1 min extension at 72°C. At this point the reaction was halted at 4°C until primers HXK2_US_f and HXK2_DS_r had been added after which the program continued: 30 s denaturation at 98°C, 20 cycles of 10 s denaturation at 98°C, 30 s annealing at 63°C (-0.4°C cycle -1 ) and 1 min extension at 72°C. The program was finished with a final 10 min extension at 72°C.
The amplified deletion cassette was purified and used to transform TMB3042. Transformants were selected on solid YNB medium with 2% glucose, 50 mg L -1 uracil and 220 mg L -1 leucine. Correct integration was verified by PCR amplification from chromosomal DNA of randomly selected colonies using primers HXK2_823US_f and TRP1_71_r. One positive clone was named TMB3460.

Construction of TMB3461 (hxk2-∆ hxk1-∆1)
The upstream and downstream fragments flanking the HXK1 gene were amplified from genomic DNA from S. cerevisiae CEN.PK2-1C. The auxotrophic marker cassette URA3 was amplified from plasmid p426. All fragments were amplified using the following PCR program: 30 s denaturation at 98°C, 30 cycles of 10 s denaturation at 98°C, 30 s annealing at 65°C (-0.5°C cycle -1 ), 1 min s extension at 72°C and 1 cycle of 10 min extension at 72°C.
The deletion cassette HXK1_US-URA3-HXK1_DS was created by OE-PCR using the following PCR program: 30 s denaturation at 98°C, 16 cycles of 10 s denaturation at 98°C, 30 s annealing at 68°C (-0.5°C cycle -1 ) and 1 min extension at 72°C. At this point the reaction was halted at 4°C until primers HXK1_US_f and HXK1_DS_r had been added after which the program continued: 30 s denaturation at 98°C, 20 cycles of 10 s denaturation at 98°C, 30 s annealing at 67°C (-0.4°C cycle -1 ) and 1 min extension at 72°C. The program was finished with a final 10 min extension at 72°C.
The amplified deletion cassette was purified and used to transform TMB3460. Transformants were selected on solid YNB medium with 2% galactose and 220 mg L -1 leucine. Correct integration was verified by PCR amplification from chromosomal DNA of randomly selected colonies using primers HXK1_1510US_f and URA3_120_r. One positive clone was named TMB3461.
The deletion cassette GLK1_US-[LoxP-KanMX-LoxP]-GLK1_DS was created by OE-PCR using the following PCR program: 30 s denaturation at 98°C, 16 cycles of 10 s denaturation at 98°C, 30 s annealing at 68°C (-0.5°C cycle -1 ) and 1 min extension at 72°C. At this point the reaction was halted at 4°C until primers GLK1_US_f and GLK11_DS_r had been added after which the program continued: 30 s denaturation at 98°C, 20 cycles of 10 s denaturation at 98°C, 30 s annealing at 68.5°C (-0.4°C cycle -1 ) and 1 min extension at 72°C. The program was finished with a final 10 min extension at 72°C.
The amplified deletion cassette was purified and used to transform TMB3461. Transformants were selected on solid YNB medium with 2% galactose, 220 mg L -1 leucine and 200 mg L -1 geneticin. Correct integration was verified by PCR amplification from chromosomal DNA of randomly selected colonies using primers GLK1_1050US_f and kanMX_32_r. One positive clone was named TMB3462.
Investigation of the glucose phosphorylating activity in strain TMB3462 revealed a significant level of activity despite confirmation of all deletion cassettes. Amplification of each gene using specific primers ( TRP1_71_r

Construction of the screening strain TMB3463
Generation of plasmid pUG62AUR To facilitate the use of pUG6AUR, a new multiple cloning site was created near the LoxP site upstream of the aureobasidin A resistance gene using linkers. These linkers consisted of two long oligonucleotides, 37 and 35 bases, respectively, and were complimentary except for the sticky ends. These were instead complementary to the restriction sites SalI and NdeI present in the pUG6AUR plasmid. All in all, the linkers included six different restriction sites: SalI, KpnI, SmaI, SphI, AvrII and NdeI (Fig. S4). The linker oligomers were annealed to each other by mixing 500 pmol of each oligonucleotide with 6 µmol NaCl in 1 DreamTaq Buffer (final volume 30 µL) and incubated at 100°C in a Thermocycler for 5 min. The temperature was then lowered slowly at 1°C min -1 until a temperature of 4°C was reached. The salt was removed through alcohol precipitation. 1 mL 99.5 % ethanol was added to the 30 μL mix and incubated at -80°C for 10 min. The tube was centrifuged for 12 min at 4°C and the supernatant was discarded. The pellet was washed once with 500 μL of 70% ethanol after which the pellet was dried in room temperature for 10 min. The small pellet was finally dissolved in 25 μL TE buffer giving an estimated concentration of 20 µM (ca. 475 ng µL -1 ).
The annealed linkers were ligated to pUG6AUR, previously digested with SalI and NdeI, in a 20 µL reaction containing the following: 1 Fast Digest Green Buffer, 50 ng µL -1 linker, 12.5 ng µL -1 linear pUG6AUR, 5% (w v -1 ) PEG 4000, 0.5 mM ATP and 0.125 U µL -1 T4 DNA Ligase. The ligation mixture was incubated at room temperature for 60 min after which the ligase was inactivated at 70°C for 5 min. 2 μL of the ligation mix was removed and diluted to a final concentration of 2.5 ng plasmid μL -1 . 5 µL of the diluted ligation mix was used to transform heat shock competent E. coli NEB5. Transformants were selected on LB-medium with 100 mg L -1 ampicillin. Successful ligation was determined by digesting the purified plasmids with KpnI/SphI or KpnI/AvrII which are specific for the presence of the linker. The resulting plasmid was named pUG62AUR (Fig. S5).
Supporting Figure S5. The pUG62AUR plasmid with a multiple cloning site consisting of the AvrII, SphI, SmaI and KpnI restriction sites.
Verification of correct integration and gene deletion was performed using the primers listed in Table S7 and the following reaction mix: 1 DreamTaq Buffer, 0.2 mM of each dNTP, 0.3 µM each of forward and reverse primers and 1 U µL -1 DreamTaq DNA Polymerase.
The additional upstream and downstream fragments flanking the HXK1 gene were amplified from genomic DNA from S. cerevisiae CEN.PK2-1C using the following PCR program: 30 s denaturation at 98°C, 5 cycles of 10 s denaturation at 98°C, 15 s annealing at 60.9°C, 8 s extension at 72°C, 25 cycles of 10 s denaturation at 98°C, 20 s annealing and extension at 72°C and 1 cycle of 10 min extension at 72°C.
The purified upstream and downstream fragments were digested with SphI/AvrII and KpnI/SphI, respectively. The digested fragments were ligated into pUG62AUR, previously digested with KpnI/AvrII, using a 3:1 molar ratio of insert to vector and an incubation time of 2 h at room temperature. 5 µL of the ligation mix was used for transformation of heat shock competent E. coli NEB5. Transformants were selected on LB-medium with 100 mg L -1 ampicillin. Successful ligation was determined by colony PCR using primers HXK1_DS2_f and HXK1_US2_r and the resulting plasmid was named pUG62AUR-HXK1USDS (Fig. S6).
TMB3462 was transformed with pUG62AUR-HXK1USDS, linearized with FspI, according to High efficiency protocol. Before plating the transformants the cells were resuspended in 1 mL of YP with 2% galactose and incubated at 30°C for 2.25 h. Transformants were selected on solid YNB medium containing 2% galactose, 220 mg L -1 leucine and 0.15 mg L -1 aureobasidin A. Correct integration was verified by colony PCR on randomly selected colonies using primers HXK1_US_f and AmpR-R130. Attempts to amplify the HXK1 gene using primers HXK1_547_f and HXK1_1282_r did not result in any amplification, indicating the gene had been deleted. This was confirmed by a very low glucose phosphorylating activity in one positive clone which was named TMB3463 (Fig. S7).  Tables   Supporting Table S1. Maximum specific consumption rates, production rates and yields in anaerobic batch fermentation of 20 g L -1 glucose and 50 g L -1 xylose by TMB3492 (Hxk2p-wt) and TMB3493 (Hxk2p-Y).
Values are given as mean  standard deviation of two independent experiments. Supporting Table S2. Overall yields and production rates in anaerobic batch fermentation of 20 g L -1 glucose and 50 g L -1 xylose by TMB3492 (Hxk2p-wt) and TMB3493 (Hxk2p-Y).
Values are given as mean  standard deviation of two independent experiments.