Secretion of functional formate dehydrogenase in Pichia pastoris

Abstract

Biofuels are an important tool for the reduction of carbon dioxide and other greenhouse emissions. NAD⁺-dependent formate dehydrogenase has been previously shown to be capable of the electrochemical reduction of carbon dioxide into formate, which can be ultimately converted to methanol. We established that a functional enzyme, tagged for immobilization, could be continuously secreted by Pichia pastoris. The protein can be easily separated from the growth media and its activity remains constant over an extended period of time. This is an important first step in creating a self-sustaining system capable of producing biofuels with minimal resources and space required.

Introduction

Enzymes are capable of facilitating chemical reactions with exceptional versatility and efficiency. While the repertoire of enzymes is vast, many practically useful reactions still cannot be catalyzed with sufficient efficiency. Moreover, even if an enzyme is known to catalyze a particular chemical transformation well, production cost and stability issues may preclude its use in laboratory and industry.

Biofuel production is an important strategy for keeping CO₂ greenhouse gas emissions to a minimum. Most biofuel in the USA is produced by fermentation of sugars obtained from corn (Bhatia et al., 2012; Kondratenko et al., 2013). However, making land useable for these crops requires clearing current flora by either burning or microbial decomposition, both of which release more CO₂ into the atmosphere than the produced biofuels can compensate for (Fargione et al., 2008). While biofuels are important for the future of environmental protection, alternative, more sustainable methods of production are urgently needed.

Capture of CO₂ by employing enzymes provides a very appealing opportunity. Direct capture of carbon dioxide is the ultimate carbon neutral process; however, apart from Rubisco, which is the primary tool for CO₂ capture in plants, there are no other known means for enzymatic capture of carbon dioxide. Initial fixation of carbon dioxide in solution is critically important; once the gas is trapped, it could be fed into multiple one carbon metabolic pathways.

Immobilized nicotinamide adenine dinucleotide (NAD)-dependent formate dehydrogenase (FDH, EC 1.2.1.2)  has been recently used to electrochemically reduce carbon dioxide to formate, which can be in turn enzymatically reduced to methanol, a usable biofuel, by utilizing NAD⁺-dependent formaldehyde dehydrogenase (FlDH, EC 1.2.1.1) and NAD⁺-dependent alcohol dehydrogenase (ADH, EC 1.1.1.1) (Scheme 1) (Kim et al., 2013; Liu et al., 2013). The native function of all these enzymes is to produce NADH by oxidation of their respective substrates: formate, formaldehyde and methanol. The overall strategy is based on the fact that catalysts equally facilitate both forward and reverse reactions, thus given excess of the product and sufficient reducing power, NAD⁺-dependent dehydrogenases can reduce carbon dioxide to methanol.

Scheme 1.

Overview of the proposed bioengineered pathway for the production of methanol from CO₂ utilizing a multi-step enzyme pathway.

While partial hydrogenation of carbon dioxide has been accomplished through multiple pathways such as heterogeneous catalysis, electrocatalysis and photocatalysis, the enzymatic process is advantageous due to its ability to operate under mild conditions (Obert and Dave, 1999; Cazelles et al., 2013; Aresta et al., 2014). However, this multi-step reduction of carbon dioxide faces its own challenges, such as a low pH and elevated temperature required for maximum efficiency, which can cause the enzymes to degrade (Baskaya et al., 2009).

Our goal is to develop an economic and renewable approach to produce active dehydrogenase through continuous secretion in yeast. The yeast will continuously secrete the protein into media, which eliminates the time-consuming step of growing, expressing and extensive purification of the desired proteins from Escherichiacoli. Yeast will also replace enzymes as they degrade and extend the lifetime of any device engineered to capture CO₂. We chose Pichia pastoris, a methylotrophic yeast capable of expressing and secreting non-native proteins, due to its many important benefits: (i) recombinantly produced enzymes (including large proteins with molecular weight >50 kDa) can be secreted; (ii) P. pastoris is tolerant to high concentrations of methanol; (iii) methanol can serve as the only energy source for the culture, eliminating the need for complex media formulations that would increase production costs; (iv) a typical P. pastoris strain would use methanol as an inducer for recombinant production of proteins; (v) P. pastoris cultures grow to high-density improving the costs and diminishing the size of fermentors and (vi) P. pastoris generally does not secrete its own proteins, simplifying downstream purification of the secreted dehydrogenases (Kovar et al., 2010).

Development of an efficient secretion system for functional enzymes is commonly plagued with difficulties related to posttranslational processing (most often glycosylation) of proteins in yeast that may interfere with activity. Therefore, optimization of expression is often necessary. Additionally, we set on determining the optimal position of the affinity tag necessary for attaching the protein to solid support.

We focused our efforts on FDH from yeast Candida boidinii, which has been previously structurally and functionally characterized (Slusarczyk et al., 2000; Schirwitz et al., 2007). In this work, we established that a tagged, functional enzyme could be continuously secreted by P. pastoris. The protein can be easily separated from the growth media using Ni-NTA resin and enzyme's activity remained constant even after prolonged growth at 29°C.

Materials and methods

Cloning of FDH gene into pMCSG49 and pET28a vectors

cDNA of C. boidinii FDH (without His₆-tag) was synthesized by GenScript. To obtain FDH with N-terminal His₆-tag (FDH-N), FDH gene was cloned into pMCSG49 vector (DNASU plasmid repository) through ligation-independent cloning site (Stols et al., 2002). To obtain FDH with C-terminal His₆-tag (FDH-C), the FDH gene was cloned into pET28a (Novagen) using SalI and NcoI restriction enzymes. The DNA sequences were confirmed using Sanger sequencing.

Expression of FDH-N and FDH-C in E. coli and protein purification

The appropriate vectors containing the genes of interest were transformed into E. coli BL21(DE3) cells and plated on Luria-Bertani (LB) agar plates containing ampicillin (FDH-N) or kanamycin (FDH-C).

For the FDH seed culture, LB (50 ml) supplemented with antibiotic (100 μg/μl Amp for FDH-N or 50 μg/μl Kan for FDH-C) was inoculated with a single colony and incubated overnight at 37°C with shaking at 230 r.p.m. The seed culture (10 ml) was then diluted into LB (1 l) supplemented with antibiotic and the culture was grown at 37°C. When the culture reached an OD₆₀₀ of ~0.6, isopropyl β-D-1-thiogalactopyranoside (IPTG) was added to a final concentration of 0.4 mM. FDH was expressed for 4 h at 30°C and the cells were then collected by centrifugation (4000 g × 20 min), flash frozen in liquid nitrogen and stored at −80°C. The typical yield of wet cell paste was 4 g per liter of culture.

Cells were resuspended in working buffer (50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; HEPES, 5% glycerol, pH 7.6) at a ratio of 5 ml of buffer for every 1 g of cells. Phenylmethanesulfonyl fluoride (PMSF) was added to a final concentration of 0.5 mM. Cells were lysed by sonication on ice for 10 min (20 s lyse cycle, 20 s rest), and then the crude cell lysate was centrifuged at 20 000 g for 30 min. To precipitate DNA, a solution of streptomycin sulfate in working buffer was added to the supernatant on ice to a final concentration of 1% w/v and the mixture was allowed to stir for 15 min. Precipitate was pelleted by centrifugation at 20 000 g for 30 min. The supernatant was applied onto a Ni-NTA column (2 ml, Clontech) and washed with 2 column volumes (CV) of working buffer. Column was washed with 30 mM imidazole in working buffer to remove impurities, and then FDH was eluted with 200 mM imidazole in buffer (pH 7.6). Fractions containing protein were identified using Pierce BCA Protein Assay kit (Thermo Scientific). Combined protein fractions were diluted two-fold and applied onto a Q Sepharose Fast Flow column (6 ml, GE Healthcare), then washed with working buffer. Protein was eluted with 300 mM NaCl in working buffer. The purity (>95%) of the final protein was checked on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and the protein was concentrated to 500 µM using a 30 kDa MWCO spin filter (Vivaspin, Sartorius). Protein concentrations were determined by measuring the absorbance at 280 nm using the calculated (ExPASy) extinction coefficient 51 465 M⁻¹ cm⁻¹ (FDH-C) and 52 955 M⁻¹ cm⁻¹ (FDH-N).

Cloning of FDH into pPIC9

Sequences coding for FDH-N and FDH-C were cloned into pPIC9 vector (Invitrogen) using EcoRI and NotI restriction sites. This procedure inserts FDH gene in frame with the secretion signal open reading frame. Gene product is a fusion of α-factor with Glu-Lys-Arg*Glu-Ala-Glu-Ala at the end, followed by FDH protein. The preliminary cleavage of the signal sequence occurs between arginine and glutamine and is denoted with an asterisk. Glu-Ala repeats can be further cleaved off. FDH gene flanked by EcoRI and NotI restriction sites and empty pPIC9 vector were digested using EcoRI and NotI restriction enzymes (New England Biolabs; NEB). The digested products were purified using Nucleospin Extract II (BioBasic) kit.

The digested pPIC9 vector and FDH gene were mixed in 1:8 vector to gene ratio (based on DNA concentration in ng/μl) and then ligated using T4 DNA Ligase (NEB). The mixture was incubated at room temperature for 30 min and transformed into E. coli XL-10 cells. DNA was purified using EZ-10 Spin Column Plasmid DNA kit (BioBasic) and the sequence confirmed by Sanger sequencing.

Transformation of pPIC9 into P. pastoris

Transformation into P. pastoris was done using previously published protocol with minor modifications (Wu and Letchworth, 2004). The pPIC9 vector containing the genes of interest was cut using SalI restriction enzyme (NEB). Cut vector was purified using Nucleospin Extract II kit protocol.

Pichia pastoris GS115 yeast culture was started from a glycerol stock by inoculating 2 ml of yeast extract peptone dextrose (YPD) media and the culture was grown at 28.5°C with shaking at 230 r.p.m. overnight. Next day this culture was transferred into 100 ml of YPD and grown until OD₆₀₀ reached 1.5. The cells were pelleted at 3000 g and then resuspended in ice-cold sterile LiAc-DTT buffer (100 mM lithium acetate, 10 mM dithiothreitol; DTT, 600 mM sorbitol, 10 mM Tris pH 7.5) and allowed to incubate at room temperature for 40 min. Cells were centrifuged at 3000 g and the pellet was washed twice with 1 ml of ice-cold sterile 1 M sorbitol. Cells were resuspended in 400 µl of 1 M sorbitol to reach OD₆₀₀ between 100 and 200. Digested plasmid (3.5 μg) was mixed with GS115 cells (200 µl), the mixture was incubated on ice for 10 min and then transferred into Electroporation cuvette (2 mm electrode gap, 450 µl, Bulldog Bio). The electroporation pulse was applied to the cells at 1.5 kV using an Eppendorf 2510 electroporator. After transformation, the culture was immediately diluted 10-fold using 1 M sorbitol and variable amounts were plated on Regeneration Dextrose Base (RDB-) plates.

The same transformation procedure was repeated using empty pPIC9 digested with SalI. This step generated P. pastoris strain that does not have any proteins attached to α-factor. The strain was used to evaluate background secretion.

Screening for P. pastoris Mut⁺ phenotype (ability to grow on methanol as a sole carbon source)

Colonies that successfully grow on RDB- plates are assumed to have undergone successful homologous recombination upon transformation. To determine if transformed P. pastoris was capable of growing in the presence of methanol, several colonies from RDB- plates were streaked on Minimal Methanol (MM) plates and the plates were allowed to incubate for 3–4 days at 30°C. A colony from MM plate was used to prepare glycerol stock, 50 µl aliquots were flash frozen using acetone/dry ice bath and kept at −80°C until needed.

To confirm the presence of FDH gene, a single colony was used to inoculate 20 μl of sterile H₂O. The yeast cells were lysed by freezing the inoculated solution for 1 min in liquid N₂ followed by thawing for 1 min in a 42°C water bath. This process was repeated 15–20 times for best results. For PCR screening, 10 μl of lysate was mixed with 1.2 μl α-factor forward primer (Integrated DNA Technologies; IDT), 1.2 μl 3′AOX1 reverse primer (IDT) and 12 μl GoTaq DNA Polymerase master mix (Promega). After PCR, samples were analyzed using a 1% agarose DNA gel run in TAE buffer. A DNA band for the FDH gene was present at ~1.2 kb.

Expression and purification of FDH secreted from P. pastoris

A small volume of YPD (2 ml) was inoculated with glycerol stock (10 μl). After growing the culture for 16 h, the cells were pelleted at 2000 g and resuspended in 20 ml of buffered methanol-complex medium (BMMY; 1% yeast extract, 2% peptone, 1.34% yeast nitrogen base, 4 × 10⁻⁵% biotin,100 mM sodium phosphate pH 6.0, 0.5% methanol) and allowed to grow for 120 h at 28.5°C. Every 24 h, an aliquot was taken for analysis and methanol was added to BMMY to a final concentration of 0.5%.

The cells were removed by centrifugation at 5000 g for 2 min and the media was decanted. After pH of the media was raised to approximately pH 7–8 using NaOH, it was added to Ni-NTA resin (2 ml) pre-equilibrated with working buffer (50 mM HEPES, 5% glycerol, pH 7.6). The mixture was then rocked on ice for 2 h. Media-resin mixture was washed with working buffer (~100 ml buffer for every 1 ml of resin). Protein was eluted with 300 mM imidazole in working buffer (pH 7.6). The purity (>95%) of the final protein was checked on SDS-PAGE. The protein was concentrated using a 30 kDa MWCO spin filter (Vivaspin, Sartorius) and the imidazole concentration adjusted to ~3 mM. Protein concentrations were determined by measuring the absorbance at 280 nm using the calculated (ExPASy) extinction coefficient 54 445 M⁻¹ cm⁻¹ (FDH-C and FDH-N).

Kinetic assays

All kinetic measurements were done at room temperature (22°C) on an Agilent 8453 UV-Vis Spectrophotometer monitoring absorbance at 340 nm with a background absorbance at 800 nm. Enzyme activity was characterized in a buffer containing 100 mM sodium phosphate (pH 7), 1.0 mM NAD⁺, 160 mM formate. Unless specifically stated, the final concentration of proteins in the enzymatic assay was 0.5 μM FDH and kinetic measurements were recorded every 20 s over a period of 5 min. Change in absorbance at 340 nm was fit to a linear model. One unit of activity (U) is defined as the production/consumption of 1 µmol of NADH per min. The specific activity of enzyme was calculated with the following equation:

$$\mathrm{specific\; activity}=\frac{\mathrm{slope}\ast V_f}{6220\ast m_f},$$

where V_f is the final volume of the enzymatic assay and m_f is the mass (mg) of FDH enzyme.

Tryptic digestion

For tryptic digestion, 50 µl of protein stock (~0.1 µg protein) was mixed with 50 µl buffer (6 M urea, 100 mM Tris, pH 7.8) and 5 µl of reducing agent (200 mM DTT, 100 mM Tris, pH 7.8) and allowed to sit at room temperature for 1 h. To alkylate cysteines, 20 µl of alkylating agent (200 mM iodoacetamide, 100 mM Tris, pH 7.8) was added and protein mixture was allowed to sit at room temperature for 1 h. To consume unreacted iodoacetamide, 20 µl of reducing agent was added to the mixture. The sample was then diluted with water (775 µl) and 100 µl of porcine trypsin (SAFC, 200 ng/µl final concentration, 100 mM Tris, pH 7.8) was added. Protein mixture was allowed to digest overnight at 37°C. To stop the reaction, acetic acid was added to the mixture until pH was below six.

Mass spectrometry

The molecular weights of the purified proteins were confirmed by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry using a Bruker Autoflex III Smartbeam MALDI-TOF mass spectrometer. Purified protein samples and tryptic digest samples (~10 µM concentration) were mixed with saturated solution of sinapic acid in a 1:2 volume ratio. Ovalbumin (molecular weight; MW of 44.3 kDa) was used as an external standard.

Protein sequences of the constructs used in this work

FDH-N in pPIC9 vector (MW = 43 915 Da ε₂₈₀ = 54 445 M⁻¹)

EAEAYVEFHH HHHHSSGVDL GTENLYFQSN AMKIVLVLYD AGKHAADEEK LYGCTENKLG IANWLKDQGH ELITTSDKEG ETSELDKHIP DADIIITTPF HPAYITKERL DKAKNLKLVV VAGVGSDHID LDYINQTGKK ISVLEVTGSN VVSVAEHVVM TMLVLVRNFV PAHEQIINHD WEVAAIAKDA YDIEGKTIAT IGAGRIGYRV LERLLPFNPK ELLYYDYQAL PKEAEEKVGA RRVENIEELV AQADIVTVNA PLHAGTKGLI NKELLSKFKK GAWLVNTARG AICVAEDVAA ALESGQLRGY GGDVWFPQPA PKDHPWRDMR NKYGAGNAMT PHYSGTTLDA QTRYAEGTKN ILESFFTGKF DYRPQDIILL NGEYVTKAYG KHDKK.

FDH-N in pMCSG49 vector (MW = 43 108 Da ε₂₈₀ = 52 955 M^-1)

HHHHHHSSGV DLGTENLYFQ SNAMKIVLVL YDAGKHAADE EKLYGCTENK LGIANWLKDQ GHELITTSDK EGETSELDKH IPDADIIITT PFHPAYITKE RLDKAKNLKL VVVAGVGSDH IDLDYINQTG KKISVLEVTG SNVVSVAEHV VMTMLVLVRN FVPAHEQIIN HDWEVAAIAK DAYDIEGKTI ATIGAGRIGY RVLERLLPFN PKELLYYDYQ ALPKEAEEKV GARRVENIEE LVAQADIVTV NAPLHAGTKG LINKELLSKF KKGAWLVNTA RGAICVAEDV AAALESGQLR GYGGDVWFPQ PAPKDHPWRD MRNKYGAGNA MTPHYSGTTL DAQTRYAEGT KNILESFFTG KFDYRPQDII LLNGEYVTKA YGKHDKK.

FDH-C in pET28a vector (MW = 42 104 Da ε₂₈₀ = 51 465 M^-1)

MKIVLVLYDA GKHAADEEKL YGCTENKLGIANWLKDQGHEL ITTSDKEGET SELDKHIPDA DIIITTPFHP AYITKERLDK AKNLKLVVVA GVGSDHIDLD YINQTGKKIS VLEVTGSNVV SVAEHVVMTM LVLVRNFVPA HEQIINHDWE VAAIAKDAYD IEGKTIATIG AGRIGYRVLE RLLPFNPKEL LYYDYQALPK EAEEKVGARR VENIEELVAQ ADIVTVNAPL HAGTKGLINK ELLSKFKKGA WLVNTARGAI CVAEDVAAAL ESGQLRGYGG DVWFPQPAPK DHPWRDMRNK YGAGNAMTPH YSGTTLDAQT RYAEGTKNIL ESFFTGKFDY RPQDIILLNG EYVTKAYGKH DKKVDKLAAA LEHHHHHH.

FDH-C in pPIC9 (MW = 43 071 Da ε₂₈₀ = 54 445 M^-1)

EAEAYVEFMK IVLVLYDAGK HAADEEKLYG CTENKLGIAN WLKDQGHELI TTSDKEGETS ELDKHIPDAD IIITTPFHPA YITKERLDKA KNLKLVVVAG VGSDHIDLDY INQTGKKISV LEVTGSNVVS VAEHVVMTML VLVRNFVPAH EQIINHDWEV AAIAKDAYDI EGKTIATIGA GRIGYRVLER LLPFNPKELL YYDYQALPKE AEEKVGARRV ENIEELVAQA DIVTVNAPLH AGTKGLINKE LLSKFKKGAW LVNTARGAIC VAEDVAAALE SGQLRGYGGD VWFPQPAPKD HPWRDMRNKY GAGNAMTPHY SGTTLDAQTR YAEGTKNILE SFFTGKFDYR PQDIILLNGE YVTKAYGKHD KKGENLYFQS HHHHHH.

Results and Discussion

FDH from C. boidinii is fairly small (43 kDa) and thermostable up to 50°C (Slusarczyk et al., 2000; Tishkov and Popov, 2004). It has been structurally characterized (Schirwitz et al., 2007; Guo et al., 2016), and its specific activity was reported to be 4.4 U/mg (Krahulec et al., 2008) and 6.5 U/mg at pH 7.5 at 30°C (Slusarczyk et al., 2000). The location of affinity tags in the sequence of the protein can interfere with enzymatic activity, thus we have cloned the FDH gene into the pMCSG49 vector and into the pET28a vector to be able to obtain a protein with N-terminal His₆-tag (FDH-N) and C-terminal His₆-tag (FDH-C), respectively. Both constructs were successfully expressed in E. coli upon induction with 0.4 mM IPTG, which was visualized in a well-defined protein band at ~45 kDa. The proteins were successfully purified using a two-step purification (Ni-NTA affinity chromatography followed by ion exchange on Q Sepharose), as seen in Fig. 1.

Fig. 1.

SDS-PAGE analysis (10% acrylamide gel) of FDH-N at various stages of purification. FDH-N was expressed in E. coli BL21(DE3). Lane 1− before IPTG induction; Lane 2− after IPTG induction; Lane 3− cell debris; Lane 4− after streptomycin sulfate precipitation; Lane 5− Ni-NTA column flow through; Lane 6− Ni-NTA column wash; Lane 7− before Q Sepharose column; Lane 8− after Q Sepharose column, final protein.

Specific activity of FDH was dependent upon the placement of the His₆-tag, where FDH-N was more active than FDH-C in the reduction of NAD⁺ to NADH and consequently the conversion of formate to carbon dioxide (Fig. 2). Both FDH-C and FDH-N exhibited slightly lower activities compared to previously reported value (Krahulec et al., 2008) measured at higher temperature (30°C) and higher pH (7.5).

Fig. 2.

The kinetic assay graph of FDH-C (0.5 μM, left) and FDH-N (0.1 μM, right) proteins measuring the conversion of NAD⁺ to NADH during the oxidation of sodium formate to carbon dioxide at pH 7. The specific activity of FDH-C is 1.1 U/mg and the specific activity of FDH-N is 2.7 U/mg.

Next, we introduced the genes encoding FDH-C and FDH-N into the genome of P. pastoris. While FDH-N had a higher specific activity in bacteria, we determined the expression yield of FDH-N in yeast was much lower when compared to FDH-C, thus we focused our efforts on characterization of FDH-C. In both cases, FDH was expressed with α-factor at the N-terminus to induce extracellular secretion in P. pastoris. The α-factor signal derived from Saccharomyces cerevisiae consists of a 19 amino acid sequence (pre-signal) responsible for signaling that the expressed protein must be transferred to the endoplasmic reticulum after translation, and a 67 residue pro-signal, which is responsible for transferring the protein into the Golgi compartment where it can then be secreted outside the cell (Cregg et al., 2000; Lin-Cereghino et al., 2012; Ahmad et al., 2014).

At pH 6, the P. pastoris strain encoding FDH-C was able to secrete FDH upon the addition of 0.5% methanol to the media. Over time, the amount of the enzyme in the media increased, as visualized by SDS-PAGE (Fig. 3). After 48 h, the approximate concentration of protein in media was 8–10 mg/l. Upon purification of FDH-C from the BMMY media, the major product had the same size as the enzyme isolated from E. coli, therefore it was not degraded by natural yeast proteases (Fig. 3). Specific activity of secreted FDH was measured using the purified protein to ensure an accurate reading of concentration. On occasion, the purified protein would have a slight yellow color leftover after Ni-NTA column, however, this did not affect enzymatic activity of the protein. Overall, the secreted protein showed about a 10-fold decrease in activity compared to protein expressed in E. coli (Table I). Secreted FDH retains its activity in media despite the prolonged incubation at elevated temperature (29°C) and proteases. A kinetic assay of purified protein isolated from media grown for 24, 48 and 120 h shows little difference in specific activity for the reduction of NAD⁺ (Fig. 4, Table I).

Fig. 3.

SDS-PAGE (10% acrylamide) analysis of media after growing yeast with empty pPIC9 for 120 h (Lane 1); media after growing yeast with FDH-C for 0 h (Lane 2), 8 h (Lane 3), 24 h (Lane 4), 48 h (Lane 5), 72 h (Lane 6), 96 h (Lane 7) and 120 h (Lane 8).

Table I.

Activity of various FDH constructs expressed in E. coli and P. pastoris

Construct	Expression host	Specific Activity, U/mg
FDH-N	E. coli	2.7
FDH-C	E. coli	1.1
FDH-C, 24 h	P. pastoris	0.11
FDH-C, 48 h	P. pastoris	0.13
FDH-C, 120 h	P. pastoris	0.10

Fig. 4.

The kinetic assay graphs measuring the conversion of NAD⁺ to NADH and the oxidation of sodium formate to carbon dioxide at pH 7 as catalyzed by secreted FDH-C protein (0.5 μM). The specific activity of FDH at 24 h is 0.11 U/mg, at 48 h is 0.13 U/mg and at 120 h is 0.10 U/mg.

Puzzled by the significant drop in enzymatic activity of the FDH secreted from P. pastoris, we hypothesized that the protein produced in yeast may be glycosylated. However, the MALDI-TOF analysis of the protein expressed in P. pastoris (Fig. 5) shows no signs of significant posttranslational modifications. Another possibility for the reduced activity is imprecise cleavage of the prepro-signal on the N-terminus. The placement of linkers or fusion proteins in relation to the protein of interest can affect secretion efficiency and yeast-secreted proteins can be degraded and/or proteolyzed on their N- or C-terminus (Park et al., 1997; Moua et al., 2016). To determine the possibility of the incomplete or impartial cleavage of the α-factor, we digested the secreted enzyme with trypsin. Indeed, we have observed a peak with a mass of 7624 Da in the MALDI spectrum of the tryptic digest, which corresponds to a large fragment of α-factor (Fig. 6). The presence of α-factor in secreted FDH-C protein suggests that this sequence was not completely removed during secretion process. This observation could explain, at least in part, reduced activity of secreted FDH-C and the presence of additional bands on SDS-PAGE gel (Fig. 3).

Fig. 5.

MALDI-TOF spectrum of FDH-C secreted from P. pastoris. The calculated mass for the protein is 43.07 kDa.

Fig. 6.

MALDI-TOF spectrum of FDH-C secreted from P. pastoris after tryptic digestion. A peak at 7624 Da corresponds to a large fragment of α-factor.

In summary, we have shown that functional FDH can be continuously secreted from an engineered P. pastroris strain. Moreover, while the activity of FDH is diminished relative to the enzyme produced in E. coli, the FDH secreted by P. pastoris maintains its activity in the culture medium for a prolonged period of time. Enzyme secretion by yeast is a promising method for creating multi-enzyme devices for biofuel production. A yeast that only needs methanol for growth will continuously secrete functional enzyme to replenish the FDH inactivated by the effect of pH, temperature and prolonged electrochemical cycling.