Methanol expression regulator 1 (Mxr1p) promotes xylulose 5-phosphate recycle via increaseing transketolase activity in Pichia pastoris

Background Methanol expression regulator 1 (Mxr1p) is a key transcription factor that plays a vital role in the methanol utilization pathway in Pichia pastoris ( P. pastoris ). Most genes referred to the methanol utilization pathway were regulated by Mxr1p. However, some genes did not show a significant difference between methanol and glycerol even though they play an important role in the methanol utilization pathway. So far, the regulation mechanism about these genes and the relationship with Mxr1p are still unknown. Results Methanol metabolic pathway analysis revealed that most of the methanol-induced genes were upregulated in transcriptional level when cultured in methanol. Whereas some genes like tkl1 (transketolase 1) did not show significant up-regulation in methanol even though it plays a very important role in Xu5P recycle, the reason is still not clear. To clarify this point, firstly, pull-down and MS experiments were performed. The result shows that Tkl1p protein combined with Mxr1p in vitro . Subsequently, this result was further confirmed by yeast two-hybrid in vivo , and the binding region mainly located from 150AA to 400AA. Moreover, Ser215 phosphorylation did not affect this interaction. In addition, Mxr1p-400AA integration into Δmxr1 could rescue cell growth in methanol. All the above results proved that Mxr1p played a post-translational role in the methanol utilization pathway and Mxr1p-400AA may achieved most of Mxr1p functions. Secondly, the function of Mxr1p-Tkl1p complex was expounded by detecting formaldehyde consumption and xylulose production in cell-free systems. Results showed that Mxr1p-Tkl1p mixture could promote formaldehyde consumption and xylulose production in vitro . Conclusion Mxr1p promotes methanol utilization via combining with Tkl1p to accelerate Xu5P recycle and this interaction was not affected by Ser215 phosphorylation.


Abstract
Background Methanol expression regulator 1 (Mxr1p) is a key transcription factor that plays a vital role in the methanol utilization pathway in Pichia pastoris ( P. pastoris ). Most genes referred to the methanol utilization pathway were regulated by Mxr1p. However, some genes did not show a significant difference between methanol and glycerol even though they play an important role in the methanol utilization pathway. So far, the regulation mechanism about these genes and the relationship with Mxr1p are still unknown.
Results Methanol metabolic pathway analysis revealed that most of the methanol-induced genes were upregulated in transcriptional level when cultured in methanol. Whereas some genes like tkl1 (transketolase 1) did not show significant up-regulation in methanol even though it plays a very important role in Xu5P recycle, the reason is still not clear. To clarify this point, firstly, pull-down and MS experiments were performed. The result shows that Tkl1p protein combined with Mxr1p in vitro . Subsequently, this result was further confirmed by yeast two-hybrid in vivo , and the binding region mainly located from 150AA to 400AA. Moreover, Ser215 phosphorylation did not affect this interaction. In addition, Mxr1p-400AA integration into Δmxr1 could rescue cell growth in methanol. All the above results proved that Mxr1p played a post-translational role in the methanol utilization pathway and Mxr1p-400AA may achieved most of Mxr1p functions. Secondly, the function of Mxr1p-Tkl1p complex was expounded by detecting formaldehyde consumption and xylulose production in cell-free systems. Results showed that Mxr1p-Tkl1p mixture could promote formaldehyde consumption and xylulose production in vitro .
Conclusion Mxr1p promotes methanol utilization via combining with Tkl1p to accelerate Xu5P recycle and this interaction was not affected by Ser215 phosphorylation. 3 Background As an important heterologous protein expression system, more than 1,000 proteins have been successfully expressed via P. pastoris, including angiostatin, a-glucosidase [1]and Anti-CEACAM5 nanobody [2,3]. Moreover, as the rare methylotrophic microbes, detailed methanol metabolic pathway especially important intermediates recycle pathway not only significantly promotes other methylotrophic microbes construction but also promotes autotrophic microorganism design [4]. Many results showed that the robust alcohol oxidase 1 promoter (P AOX1 ) contributes most to this successful expression system [5].
Data showed that methanol expression regulation 1 protein (Mxr1p) plays a key role in aox1 and other methanol-induced genes transcription [6]. For example, aox1 was repressed in Dmxr1 mutant even though cells were cultured in methanol [7,8]. The reason is that Mxr1p binds to methanol-induced promoter e.g., aox1, das and pex8 [7,9]  CCA38880) regulates Mxr1p activity in a carbon source-dependent manner. The mechanism is that 14-3-3 protein binds on Mxr1p between residues 212-225. Ser215 phosphorylation promotes the interaction between Mxr1p and the 14-3-3 protein and subsequently leads to methanol metabolic pathway repression [12,13]. However, although phosphorylation of Ser215 is similar in methanol and glucose, cells showed a significant difference in methanol utilization ability. Thus, there may be other proteins interact with Mxr1p to regulate MUT pathway. Based on that, several transcription factors such as Trm1, Prm1 and Mit1 have been identified [14][15][16][17]. Disappointedly, there is no interaction among these transcription factors even though some of them, e.g., Mxr1 and Rop1, shared the same binding site in methanol-induced genes [18].
In this report, methanol-induced genes were identified by methanol metabolic pathway analysis based on transcription level. Furthermore, potential proteins (e.g., Tkl1p) interacting with Mxr1p in vitro were identified via pull-down analysis and Mass spectrometry (MS). Yeast two-hybrid screening showed that Tkl1p interacted with Mxr1p in vivo, the interaction region mainly located between 150AA to 400AA and Ser215 phosphorylation did not affect this interaction. Finally, formaldehyde consumption and xylulose production assay proved that the complex formed by Tkl1p and Mxr1p could promote methanol utilization by accelerating xylulose recycle. All of the above results not only be benefit for clarifying Xu5P recycle route but also would help enrich Mxr1p regulation system.

Results
Transcription level referred to methanol utilization pathway in P. pastoris Transcriptional levels referred to MUT genes in different media (methanol and glycerol) were detected after 48h. Hierarchical clustering analysis provides a holistic view about MUT genes expression level in methanol or glycerol media (Fig. 1). Results showed that compared with glycerol, aox1 and das1/2 were significantly upregulated in methanol.
Moreover, das2 upregulation was more significant than das1, which indicated that the regulation mechanism between das1 and das2 might be different [19]. As the key intermediate, Xu5P plays a crucial role in formaldehyde fixation and degradation, which means genes related to Xu5P recycle could be promoted in methanol. Results showed that most genes referred to Xu5P recycle route, such as rep1 and tal, were upregulated in methanol. Surprisingly, tkl1 gene did not show significantly improved as other MUT genes, even though it plays a very important role in Xu5P recycle route (Fig.1).

Fig. 2
According to transcriptome analysis, many genes related to MUT pathway and nonoxidative pentose phosphate pathway (NOP) were up-regulated in methanol medium.
However, tkl1 did not show significant up-regulated like das even though it plays a key role in the carbon rearrangement process, which indicates that maybe some other unrevealed regulation pathways participate in tkl1 regulation. To clarify this hypothesis, firstly, the transcriptional and translational level of tkl1 were measured. Results showed that tkl1 did not show a significant difference between methanol and glycerol medium The above results indicated that although Tkl1p participates in Xu5P recycle, its transcriptional and translational level did not improve a lot in methanol, which indicated that Tkl1p might be involved in post-transcriptional modification. Based on that, M1 and M2 proteins were purified and pull-down assay was performed in this research to find some candidate proteins (Fig.3a,b), which may interact with Mxr1p in vitro. Results proved that some proteins could combine with Mxr1p in vitro (Fig.3). Each candidate band was collected and sent for MS analysis, results showed the interesting protein is Tkl1p. This result was further confirmed in vivo by yeast two-hybrid assay (Fig.3b). Moreover, we found the diploid containing M1 and Tkl1p did not grow on selection plate, whereas the diploid containing M2 and Tkl1p grew on selection plate, this result indicated that the region Mxr1p interacting with Tkl1p mainly located between 150AA to 400AA. Our previous results showed that M1 plays a crucial role in the process of binding on P GT1 and P AOX1 [10]. Some other researchers showed that N400 of Mxr1p play the main function between protein interaction [12,13]. Considering all these results, Mxr1p was divided into two parts; the first part contains a DNA binding domain (mainly 1-150AA), which mainly regulates gene transcription. The second is protein interaction domain (mainly 150-400AA), which may be involved in different metabolic pathways via interacting with different proteins.

Fig. 4
It has been proved that 14-3-3 protein interacted with Mxr1p via phosphorylation of Ser215. In order to examine whether Ser215 phosphorylation would affect the interaction between M2 and Tkl1p. M2-Ser215 was mutant to M2-Asp215 (M3). Tkl1p-His and M3-Flag protein were purified and used for pull-down assay. The targeted band was recognized by Western blotting (Fig.4a), which indicates that phosphorylation of Ser215 or not did not affect the interaction between Tkl1p and Mxr1p. Meanwhile, different strains growth curve were measured in methanol medium (Fig.4b). Results showed that mxr1 deletion (Dmxr1) affect cell grew. Moreover, when M2 or M3 was integrated back into genome respectively, cell grew would be rescued. This indicated that Mxr1p N-400AA take main function of Mxr1p.

Fig. 5
Based on Fig.1, it could be inferred that Tkl1p mainly promotes Xu5P recycle in the carbon rearrangement route. Considering all results, one hypothesis is that the complex formed by Mxr1p and Tkl1p may participate in Xu5P recycle route. In order to test this hypothesis, xylulose production was measured in vitro using crude total protein. Results (Fig.5) showed that xylulose production in tkl1 overexpression strain (32.7 mM, 1% formaldehyde) is much higher (P <0.01) than that in wild type (12.6 mM, 1% formaldehyde), which confirmed that Tkl1p promote Xu5P production. Moreover, when the purified M2 protein was added into this reaction system, higher production of xylulose (36.2 mM, 1% formaldehyde) was detected (P = 0.0009 < 0.01), the above results proved that the complex formed by Mxr1p and Tkl1p promotes xylulose production. In order to further confirm this result, formaldehyde consumption in different reaction systems were detected, results showed that (Fig.5a) more formaldehyde consumption was detected in Tkl1p-Mxr1p reaction system than others. Based on the above results, it concluded that Mxr1p could combine with Tkl1p to promote Xu5P recycle and further accelerate formaldehyde consumption.
Many intermediates, especially formaldehyde, in methanol utilization are toxic for cells, therefore, how to convert these toxic intermediates into useable substrates is very important for cell to survive. It has been proved that high efficient Xu5P recycle route would promote formaldehyde degradation and rescue cells[10]. However, no definite results illustrate Xu5P recycle route in methylotrophic microbes. Based on the above results, it showed that complex formed by Mxr1p and Tkl1p participate in Xu5P recycle route. It has been proved that xylulose production increased in vitro experiments. The reason may be that excessive Xu5P was converted into xylulose by some enzyme. To verify this hypothesis, xylulokinase (XK) expression level was measured in vivo, the result showed that XK expression level was promoted in Tkl1-Mxr1 over-expression cells, meanwhile, methanol consumption was also improved (Fig.5b).

Discussion
As an important renewable carbon source, methanol has drawn lots of attention during the past few years due to its low energy cost, liquid form [20][21][22] and other advantages.
Moreover, compared with other one-carbon sources like CO 2 , methanol could be high efficiency used by native methylotrophic microbes [23]. Besides, both CO 2 fixation pathway and formaldehyde degradation pathway share similar intermediate such as Xu5P and Ru5P [24,25]. Based on the above reasons, lots of attention has been paid to engineer nonnative methylotrophic microbes. Unfortunately, most of the recombinant cells did not grow well in methanol. Many reasons may lead to this problem, but inefficiency intermediates supplement may be an important reason. Therefore, Clarify intermediates recycle regulation mechanism in methanol utilization pathway maybe not only beneficial for other non-native methylotrophic microbes construction but also would promote other one-carbon source utilization [4,26]. In Saccharomyces cerevisiae, Tkl1 has been shown as the ratelimiting factor in NOP pathway [27]. Moreover, in P. pastoris Tkl1p showed similar conserved domain structure with Das1p and Das2p. Some results indicated that cells losing das1 and das2 are still able to grow in methanol [28], which indicated that Tkl1p may be involved in some other pathways. However, Tkl1 did not showed significant difference between methanol and glycerol as DAS1/2, which also indicated that tkl1 and das1/2 regulation mechanism might be different [29] .
In this paper, our results showed that Tkl1p interacted with M2, and the interaction region mainly located between 150AA to 400AA. In addition, the function of Tkl1p-Mxr1p was confirmed to promote Xu5P recycle. However, when we detected XK transcriptional level in different cells, we found that there is no significant difference between Tkl1p overexpression strain and Tkl1p-Mxr1p double overexpression strains (Fig.4), the reason may be that Xu5P could be highly recycled in vivo to support formaldehyde fixation. As a critical transcription factor, Mxr1p is inextricably linked with the methanol utilization pathway both in transcriptional and post-translational [8], which indicated that Mxr1p may be the core of methanol regulation system in P. pastoris. In this study, many efforts have been tried to purify integral Mxr1p protein, unfortunately, all of them were failed ( Fig.S1/2). Some researchers have indicated that overexpressed Mxr1p is toxic to cells [30], but the reason is not clear now. Our results also indicated that Mxr1p overexpression would significantly inhibit cell growth (Fig.S4). Based on our results, it inferred that Mxr1p might be involved in many different metabolic pathways besides MUT pathway.
Therefore, Mxr1p overexpression may inhibit other important metabolic pathways, not 'toxic' for cells.

Conclusion
In this paper, Tkl1p regulation mechanism was clarified via different experiments, results showed that Tkl1p could bind on Mxr1p specific region (mainly located from 150AA to 400AA), and the complex would promote Xu5P recycle then further promote methanol utilization.
Electroporation was used to transform P. pastoris. Transformation and recombinant DNA operations were performed as described previously [11]. Selection of marker-resistant colonies was performed using LB with 50 µg/mL ampicillin or kanamycin, and YPD with 0.3 mg/mL G418 or 0.1 mg/mL zeocin.
Subsequently, the PCR products were digested with NdeI/XhoI, and the digested fragments were inserted into the pSVT7 plasmid. The recombinant plasmid was chemically transformed into E. coli BL21(DE3) cells.

M1/2-pGBKT7 recombinant plasmid construction
M1/2 were amplified with Pfu DNA polymerase and the P. pastoris genome was used as the template. The NdeI/XhoI-digested PCR products were ligated into the pGBKT7 plasmid.
Recombinant plasmids M1-pGBKT7 and M2-pGBKT7 were transformed into S. cerevisiae following the protocol provided by TAKARA.

AD-pGADT7 recombinant plasmid construction
The general procedure was the same as above. Target genes (like tkl1) were amplification and the PCR products were double-digested with the same restriction enzymes, and the digested fragments were ligated into the appropriate plasmid. Recombinant plasmid AD-pGADT7 was transformed into S. cerevisiae using the protocol provided by TAKARA.

M1/2/3 pull-down assay
His-M1/2 and His-M3 fusion proteins were immobilized to beads (Dynabeads Ò His-Tag Isolation) and washed four times with washing buffer (50 mM sodium phosphate, 300 mM NaCl, 0.01% Tween-20, pH 8.0). A total protein (or purified Tkl1p-His protein) from a yeast culture in BMMY or BMGY medium was added (Reaction solution). The mixture was incubated for 30 min and the beads washed four times with A solution (3.25 mM sodium phosphate, 70 mM NaCl, 0.01% Tween-20, pH 7.4) (Washing solution). Finally, 50 mL His elution buffer (300 mM imidazole, 50 mM sodium phosphate, 300 mM NaCl, 0.01% Tween-20, pH 7.4) was added to the samples and these were stirred for 2 min. The His-tagged protein was eluted (Elution solution), and proteins that interacted with the His-tag protein were transferred to sample tubes. All mixtures were subjected to SDS-PAGE and MS analysis.

M1/2 yeast two-hybrid
The protocol from TAKARA was used. Y2HGold cells transformed with the M1 or M2-pGBKT7 recombinant plasmids (SD) were cultured in YPD medium. Y187 with the AD-pGADT7 recombinant plasmid (AD) were also cultured in YPD medium. Equal volumes of the cultures containing SD and AD were then mixed and added to 0.5 mL 2´ YPDA. The resuspended cells were mixed thoroughly. This mixture was incubated overnight (20-24 h) at 30 °C, 200 rpm. Finally, a 100 mL aliquot of the incubated mating culture was incubated in SD/Try/Leu/His/Ade medium at 30 °C for 3-5 d until visible single colonies appeared.

Dmxr1 mutant construction
The P. pastoris strain carrying the mxr1 gene deletion was constructed by homologous recombination using kan as a marker. The upstream region of the mxr1 gene was amplified initially by PCR using Pfu DNA polymerase (Thermo Scientific) and the P. pastoris genome as the template. The primers for this PCR, mxr1s-1 and mxr1s-2, included SphI and BamHI restriction sites, respectively. The 0.6-kb PCR-amplified fragment was inserted into SphI/BamHI-digested pMD™19-T plasmid (TAKARA) to create the pMXR1UP plasmid.
The downstream region of the mxr1 gene was also amplified with primers mxr1x-1 and mxr1x-2, carrying restriction sites for KpnI and EcoRI, respectively. This 0.5-kb PCR fragment was inserted into pMD19-T to yield the pMXR1Down plasmid. Next, the G418 resistance gene with its own promoter and terminator (1556 bp) was amplified by PCR, using pFA6a-KanMX6 as the template and primers kan-1 and kan-2, which carried BamHI and KpnI restriction sites, respectively, and the fragment was cloned into the SphI/BamHIdigested plasmid pMXR1UP to give the pMXR1UP-Kan plasmid. This plasmid was digested with KpnI/EcoRI to generate a 2.2-kb fragment that was then inserted into KpnI/EcoRIdigested pMXR1Down plasmid, yielding a P. pastoris mxr1 deletion plasmid, pMD19-T-MXR1-del. The deletion cassette was released from pMD19-T-MXR1-del as a 2.7-kb EcoRI/SphI-digested fragment and transformed by electroporation into wild-type P. pastoris (strain X-33). G418-resistant transformants were isolated on YPD supplemented with 1 mg/mL G418. The correct integration of the deletion cassette into the genome and replacement of the mxr1 open reading frame (ORF) in the transformants was confirmed by PCR analysis and Sanger sequencing.

mxr1 and tkl overexpression strain construction
The mxr1 gene and tkl were amplified by PCR using genomic DNA as the template, and different primers. The fragment was ligated into the pMD™19-T plasmid (TAKARA) and sequenced. The recombinant plasmid and pGAPZB plasmid (Invitrogen) were then digested with PmlI/XhoI, and the mxr1 fragment was inserted into pGAPZB to yield pGM. Finally, pGM was digested with AvrII and electro-transformed into the P. pastoris X-33 Dmxr1 mutant and P. pastoris X-33 wild-type, yielding Mxr1 overexpression strains mxr1-Dmxr1 and mxr1-wt, respectively. tkl1 were amplified by PCR using genomic DNA as the template. Purified PCR products were PmlI/KpnI-digested. This was followed by ligation into pGAPZB to create pGAPZB-tkl1. Then all the above plasmids were transformed into X-33 respectively by electro-transformation.

Real-time PCR
Total RNA extraction was carried out according to the standard procedure described in the Quantscrip TR kit (TAKARA). The reaction consisted of 5´ gDNA Eraser Buffer (2 μL

Western blotting detection
Total protein samples were taken from cells cultured in YPD. Cultivating E. coli BL21 cells were transferred from 1× YPD after 16 h to 2× YPD. Cells were harvested ~6 h later and the cell pellet washed twice with ice-cold 50 mM potassium phosphate buffer (pH 7.0). An electronic oscillator was used to disrupt cells to obtain soluble proteins. Before oscillation, 200 μL of ice-cold PEBF (0.7882% Tris-HCl, 0.0585% EDTA and 2 μL of 100 mM PMSF) and glass beads were also added to ensure efficient cell disruption. The lysed cells were centrifuged at 10,000 g for 1 min, and the supernatant was collected and used for SDS-PAGE analysis.

Xylulose detection
Wild-type (X-33) cells and Tkl1p over-expression strain cells were harvested when the OD 600 reached ~6.0. The harvested cells were transferred from YPD to BMMY and incubated overnight at 30 °C and 230 rpm. Total protein extraction was the same as described above. The reaction system included formaldehyde (1-7%), xylulose (20 mM), and PEB and Mxr1 proteins (25 mL). The reaction was initiated by the addition of the crude cell extract (70 mL) and terminated with 20 mL H 2 SO 4 -Na 2 SO 3 (2 M, pH 2.0). Highperformance liquid chromatography (HPLC) was performed to quantify the xylulose level.
The HPLC parameters were as follows: an Aminexâ HPX-87H ion exclusion column, column temperature 60 °C, the flow rate of 0.6 mL/min and 5 mM H 2 SO 4 as eluent [32,33].

Formaldehyde and Methanol detection
Cells were incubated overnight in BMGY medium (0.5% methanol) and then transformed into fresh BMMY medium incubate for 12h, 30℃. Cells were centrifuged and supernatant was collected for methanol and formaldehyde detection. HPLC was used for detection

Author contributions
Chunjun Zhan and Yankun Yang designed the research; Chunjun zhan and Yingyue Pan performed most of the experiments; Xiuxia Liu, Chunli Liu, Jinling Zhan, and Dinghua Xu analysed the data; Chunjun Zhan wrote the paper and Zhonghu Bai edited the manuscript.

Ethical approval
This article does not contain any studies with human participants performed by any of the authors.

Consent for publication
Not applicable.