OUP user menu

Formaldehyde dehydrogenase from the recombinant yeast Hansenula polymorpha: isolation and bioanalytic application

Olha M. Demkiv, Solomiya Ya. Paryzhak, Galina Z. Gayda, Volodimir A. Sibirny, Mykhailo V. Gonchar
DOI: http://dx.doi.org/10.1111/j.1567-1364.2007.00255.x 1153-1159 First published online: 1 October 2007


A recombinant yeast clone, a derivative of the recipient Hansenula polymorpha strain NCYC 495, was chosen as an NAD and glutathione-dependent formaldehyde dehydrogenase overproducer. Optimal cultivation conditions for the highest yield of enzyme were established. A simple scheme for the isolation of formaldehyde dehydrogenase from the recombinant strain was proposed, and some characteristics of the purified enzyme were studied. An enzymatic method for formaldehyde assay based on formaldehyde dehydrogenase was developed and used for testing real samples.

  • Hansenula polymorpha (Pichia angusta)
  • formaldehyde dehydrogenase
  • purification
  • enzymatic assay
  • formaldehyde


Formaldehyde is considered to be one of the most important commercial chemicals, due to its broad application in the chemical synthesis of resins. It is also used for the production of different consumer goods (detergents, soaps, and shampoos) and as a sterilizing agent in pharmacology and medicine (Gerberich & Seaman, 1994). Recently, a new health risk factor associated with formaldehyde has been revealed. Some advanced technologies for potable water pretreatment include the ozonation process, during which formaldehyde is generated as a result of the reaction of ozone with humus traces (Schechter & Singer, 1995). At the same time, formaldehyde is a natural metabolite of living organisms. It has been found in fruits, vegetables, flesh, and biological fluids of human origin (Gerberich & Seaman, 1994). In extreme cases, some frozen fish, especially Gadoid species, can accumulate up to 200 mg of formaldehyde per 1 kg wet weight, due to the enzymatic degradation of a natural fish component – trimethylamine oxide (Rehbein, 1995; Pavlishko et al., 2003).

Formaldehyde is classified as a mutagen and possible human carcinogen (Feron et al., 1991), and is one of the chemical mediators of apoptosis, i.e. programmed cell death. These considerations are sufficient to demonstrate the necessity for formaldehyde control in consumer goods and the environment as well as in biological samples. Such control requires the development of simple, cheap, sensitive and selective methods for the analysis of this extremely toxic agent. Among them, enzymatic and biosensor-based approaches are the most promising, due to high selectivity and sensitivity.

Formaldehyde dehydrogenase (FdDH), a key enzyme of formaldehyde metabolism in microorganisms, has been investigated for its possible use in bioanalytic processes (Winter & Cammann, 1989; Korpan et al., 1993, 2000; Herschkovitz et al., 2000; Gonchar et al., 2002; Kawamura et al., 2005; Vastarella & Nicastri, 2005; Ben Ali et al., 2006a, b). The broad use of FdDH in analytic practice is hampered by insufficient activity of the commercial preparations of the enzyme from Pseudomonas putida and Candida boidinii, as well as by relatively high costs of enzyme preparations isolated from the wild-type strains (Sigma-Aldrich Catalogue, 2007).

In this article, we describe the optimization of the cultivation conditions for the gene-engineered strain of the thermotolerant methylotrophic yeast Hansenula polymorpha (Pichia angusta) overproducing thermostable FdDH, the procedure of enzyme isolation from the recombinant cells, some characteristics of the purified FdDH, the development of an FdDH-based method for formaldehyde assay, and its application in analysis of real waste samples.

Materials and methods

Microbial source of FdDH, cultivation conditions, and preparation of cell-free extract (CE)

The stable recombinant strain Tf 11-6 was chosen as an FdDH overproducer shown to have the highest activity of FdDH (up to 4.0 μmol min−1 mg−1 protein) in CE (Demkiv et al., 2005).

Yeast cells Tf 11-6 were cultivated in flasks on a shaker (200 r.p.m.) at 28°C until the middle of the exponential growth phase (c. 24 h) in a medium containing 3.5 g L−1 (NH4)2SO4, 1.0 g L−1 KH2PO4, 0.5 g L−1 MgSO4·7H2O, 0.1 g L−1 CaCl2, 0.04 g L−1 leucine and 0.5 g L−1 yeast extract, supplemented with standard microelements (Demkiv et al., 2005). As carbon sources, 1% methanol, 1% glucose or 1% ethanol was used.

After washing, the cells were suspended in 0.05 M K, Na-phosphate buffer (pH 8.0) (PB), containing 0.4 mM phenylmethylsulfonyl fluoride and 1.0 mM EDTA, frozen, and kept at −20°C. To obtain the CE, cells were disrupted with glass beads (diameter=0.45–0.5 mm) in a planetary disintegrator at 110 g at 4°C for 6 min. The cell debris was removed by centrifugation at 19 800 g at 4°C for 40 min, and the supernatant (CE) was used for testing the activity and isolation of the enzyme.

Assay of FdDH and alcohol oxidase (AOX) activities

The activity of FdDH was determined by the rate of NADH formation monitored spectrophotometrically at 340 nm (Schutte et al., 1976). One unit (1 U) of enzyme activity (A+FA) was defined as the amount of the enzyme that forms 1 μmol NADH min−1 under standard conditions of the assay: 25°C, 1 mM formaldehyde, 1 mM NAD+ and 2 mM glutathione in PB (50 mM phosphate buffer, pH 8.0). Nonspecific activity (A−FA) was measured in the same way but without addition of formaldehyde. Instant activity of FdDH (A) was calculated as the difference between A+FA and A−FA.

The activity of AOX was determined by the rate of hydrogen peroxide formation in the reaction with methanol, as monitored by the peroxidative oxidation of o-dianisidine in the presence of horseradish peroxidase (Shleev et al., 2006). The millimolar extinction coefficient of the colored product in acidic solution (2.5 M HCl) at 525 nm was shown to be 13.38 mM−1 cm−1.

Isolation and purification of FdDH

FdDH was isolated from the CE of H. polymorpha recombinant strain Tf 11-6 by two-step column chromatography on the anion-exchange sorbent DEAE-Toyopearl 650 M. In the first step, CE was applied to the sorbent, equilibrated by PB (pH 8.0). The fraction of unabsorbed proteins, which contained FdDH, was diluted with water (1 : 3), Tris base was added to adjust the pH to 8.8, and the final solution was applied to the same column (the second step), previously washed with 1 M NaCl and equilibrated with 40 mM Tris buffer, pH 8.8 (TB). Enzyme was eluted with 0.1 M NaCl in the initial buffer (TB), and the specific activity of FdDH was assayed in each fraction. The concentration of protein was determined by the Lowry method. The fractions of eluate with enzyme activity higher than 10 U mg−1 and devoid of AOX activity were combined, and then dithiothreitol (DTT) up to 2 mM and ammonium sulfate (up to 80% saturation; pH 8.0; at 0°C) were added. After incubation at 0°C for 1 h, the enzyme was collected by centrifugation (19 800 g, 30 min, 4°C), and the pellet was resuspended in a minimal volume of ammonium sulfate solution (80% saturation) in 40 mM TB with 2 mM DTT. The FdDH preparation was kept at −10°C.


The molecular weights of the FdDH subunits were calculated from the electrophoretic mobility values of FdDH and a set of standard proteins after sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) in 5–20% polyacrylamide gel (PAAG). The SDS PAAG was stained with Coomassie Blue R-250. To visualize enzyme bands in native PAAG (Maidan et al., 1997), we used a modified mixture for the FdDH assay: 1.0 mM formaldehyde, 1.0 mM NAD, 2.0 mM glutathione, 0.05 mM nitrotetrazolium blue (NTB) and 0.003 mM phenazine methosulfate (PMS) in 50 mM PB (pH 8.0). The PAAG was washed with water after the appearance of blue–violet FdDH bands.

Chemical assay of formaldehyde

The assay was performed by two methods, using chromotropic acid (Polska Norma PN-71 C-04568, 1988) and 3-methyl-2-benzothiazoline hydrazone hydrochloride (MBTH) (Sawicki et al., 1961).

Enzymatic assay of formaldehyde

Model and real samples (0.5 mL) (water for the blank sample) were treated at room temperature with 0.5 mL of FdDH-containing reagent of the following composition: 23 mU mL−1 FdDH, 0.63 mM glutathione, 0.31 mM NAD+, 1.0 mM NTB, 0.024 mM PMS and 0.01% Triton X-100 in 60 mM PB (pH 8.0). The reaction mixture was incubated for 30 min at room temperature. To terminate the reaction, 3 mL of 0.3 M HCl was added, and the OD of the sample at 570 nm was measured. The formaldehyde content was calculated from the calibration curve.

Results and discussion

Optimization of cultivation conditions

As reported by us previously (Demkiv et al., 2005), to construct the FdDH-overproducing recombinant strain, the FLD1 gene from H. polymorpha with its own promoter was recloned into LEU2-containing integrative plasmid pYT1 devoid of the autonomously replicating sequence (ARS) to be used for multicopy integration of the gene into chromosome of the recipient strain NCYC 495 (leu1-1). The transformation of the cells by electroporation was performed with linear and cyclic forms of the plasmid, and 50 clones among Leu+ transformants with elevated resistance to formaldehyde were selected and tested according to their FdDH activity in CE. The stable recombinant strain Tf 11-6 was shown (Demkiv et al., 2005) to have the highest activity of FdDH in CE (up to 4.0 μmol min−1 mg−1 protein), so it was chosen as an FdDH overproducer.

To optimize the cultivation conditions for the highest enzyme yield, the effect of growth medium composition on FdDH level for strain Tf 11-6 was investigated (Fig. 1). As shown in Fig. 1a, FdDH activity in CE is dependent on the carbon source. Cultivation on 1% methanol provided the highest level of enzyme synthesis for recipient and overproducing strains. Addition of formaldehyde to methanol medium stimulated synthesis of FdDH. As shown in Fig. 1b, FdDH specific activity in CE of Tf 11-6, cultivated in methanol medium supplemented with 10 mM formaldehyde, equaled 7.0 U mg−1; that is, it was twice as high as the activity of FdDH in the cells cultivated without addition of formaldehyde.

Figure 1

FdDH activity in CE of the recipient and recombinant strains cultivated in different media. (a) Visualization of FdDH activity in 8% PAAG after native electrophoresis of CE. Carbon sources in a growth medium: 1% ethanol (EtOH), 1% methanol (MeOH) or 1% glucose (Glc). Each CE sample contained 0.1 mg of protein: 1, Tf 11-6 (0.12 U for EtOH, 0.35 U for MeOH, and 0.05 U for Glc); 2, leu 1-1 (0.01 U for EtOH and Glc, 0.15 U for MeOH); K, 0.01 mg of purified FdDH preparation (0.17 U). (b) FdDH activity in CE of cells cultivated in a medium containing 1% methanol and different concentrations of formaldehyde (mM).

Enzyme purification and characterization

For isolation of FdDH, cells of the recombinant overproducer strain cultivated in 1% methanol medium in the presence of 5 mM formaldehyde were used. The characteristics of the purification procedure, which included two-step ion-exchange chromatography and ammonium sulfate precipitation, are shown in Table 1. The molecular mass of the FdDH subunit, estimated by SDS-PAGE, was shown to be about 40 kDa (Fig. 2a), which is the same as the value found for C. boidinii−41 kDa (Melissis et al., 2001). It was reported that the predicted FLD1 product (Fld1p) is a protein of 380 amino acids (c. 41 kDa) (Baerends et al., 2002). Taking into account that the molecular masses of native FdDHs from various methanol-utilizing yeasts were estimated to range from 80 to 85 kDa (Allais et al., 1983; Schutte et al., 1976), the isolated enzyme is presumed to be dimeric.

View this table:
Table 1

Isolation and purification of FdDH from the recombinant strain Tf 11-6

StepTotal activity (U)Specific activity (U mg−1)Yield (%)Purification factor (fold)
Chromatography 1, unabsorbed proteins1848.8962
Chromatography 2, eluate3912.1212.7
Precipitation by 80% (NH4)2SO4, precipitate3017.015.73.8
  • 1 and 2, chromatography on DEAE-Toyopearl-650 M at pH 8.0 and pH 8.8.

  • CE, cell-free extract.

Figure 2

Some characteristics of the purified FdDH preparation (17 U mg−1). (a) Estimation of molecular mass of the FdDH subunit (kDa) using SDS-PAGE (5–20%): left, protein standards; right, μg of FdDH. (b) pH optimum (1) and pH stability (2) of the enzyme. (2) FdDH was incubated in different buffers (50 mM PB, pH 5.25–8.0; 50 mM borate buffer, pH 9.2–10.0) for 60 min and tested for residual activity at standard assay conditions. (c) Temperature optimum (1) and thermostability (2) of FdDH. (1) Before addition of FdDH, the reaction mixture was preincubated for 10 min at a fixed temperature (23°C, 30°C, 35°C, 40°C, …, 70°C); FdDH was then added, and its activity was determined in a thermostated cuvette at the same temperature for at least 5 min. (2) FdDH solution in PB (pH 8.0) was heated for 10 min at a fixed temperature (23°C, 30°C, 35°C, 40°C, …, 70°C), cooled, and tested for residual activity under standard assay conditions (25°C).

Some physicochemical characteristics of the isolated FdDH are shown in Fig. 2b and c. In Fig. 2b, the evaluation of the pH optimum and pH stability of the enzyme (incubation in the appropriate buffer at room temperature for 60 min) is presented. The pH optimum was found to be in the range 7.5–8.5, and the highest stability of FdDH was observed at pH 7.0–8.5.

The optimal temperature for enzyme activity was 50°C (Fig. 2c). At 65°C, the enzyme retained about 60% of its highest activity (the assay time was c. 5 min), i.e. equal to the level of FdDH activity at 30°C. As shown in Fig. 2c, the enzymatic activity at 37°C was 1.6-fold higher than with the standard conditions for FdDH activity assay (at 25°C). Study of the thermal stability of the enzyme (Fig. 2c) demonstrated that its activity was completely preserved after 10 min of incubation at 40°C, and was partially preserved at 55°C (up to 70%) and 60°C (25%). Complete inactivation occurred after heating of the enzyme solution at 70°C for 5 min. This apparent high thermostability makes FdDH potentially useful for bioanalytic purposes, namely, for formaldehyde assay in food products, wastewater, and pharmaceuticals, and for biotransformation of formaldehyde to formic acid.

Thus, thermostable NAD+- and glutathione-dependent FdDH from the overproducing recombinant strain of the thermotolerant methylotrophic yeast H. polymorpha was isolated, and some physicochemical properties of the enzyme were studied. The specific activity of the final preparation of FdDH was 17.0 U mg−1 protein at 25°C and about 27 U mg−1 at 37°C. For comparison, the specific activities (determined at 37°C) of commercially available FdDH preparations from Ps. putida are 1–6 U mg−1, and those of preparations from the yeast C. boidinii are 17–20 U mg−1 (Sigma-Aldrich Catalogue, 2007).

Application of FdDH for enzymatic assay of formaldehyde

We propose to use FdDH for the enzymatic assay of formaldehyde. The method developed is based on the photometric detection of a colored product, formazan, formed from NTB in a reaction coupled with FdDH-catalyzed oxidation of formaldehyde (Fig. 3).

Figure 3

The scheme of reactions for the developed FdDH-based enzymatic assay of formaldehyde.

The optimal composition of the reaction mixture and the optimal assay conditions were estimated as described in ‘Materials and methods’. The assay was performed in a mode of incomplete conversion of the analyte (c. 10%), using a limited concentration of the enzyme (23 mU mL−1) in the reagent. These conditions are economic and reasonable, because of a high content of formaldehyde in the tested samples. Under conditions of complete oxidation of formaldehyde (excess of the enzyme), the sensitivity of the assay was determined as 2.5 μM (in the final reaction mixture) or 20 μM (in the tested samples).

The reliability of the developed method was tested on the real wastewater samples containing formaldehyde. As shown in Table 2, the comparison of formaldehyde content values obtained by the FdDH-based method and by two routinely used chemical methods (chromotropic acid and MBTH) showed a good correlation between both approaches. Only in some cases (samples DK5 and DK7) with a lower formaldehyde content was the difference between the compared methods higher than 15–41% and 26%, respectively. A relatively large difference was also observed between the two chemical methods for the above-mentioned samples −37% and 21%. This can be explained by a higher error in measurement of low OD values obtained for samples with a low formaldehyde content. On the other hand, it is worth emphasizing that the chemical approaches used are not free from possible error due to interfering effects of the coimpurities usually present in the real samples, e.g. phenol, which is an attendant pollutant of formaldehyde-containing wastes (Polska Norma PN-71 C-04568, 1988). To evaluate the possible interfering effect of the components of real samples on formaldehyde assay by FdDH-based and chromotropic acid methods, we used a standard addition test for sample WW-A (Table 2 and Fig. 4a and b). It can be clearly seen from Fig. 4 that the chromotropic method is more sensitive to the interfering effect of real sample components than the enzymatic method: the slope values for calibration curves obtained for formaldehyde in water and in the background of real sample (WW-A) differed by 24% (2.986 and 2.255, respectively). The respective values obtained for the enzymatic method were 0.761 and 0.729, and the difference was only 4.2%, which is inside the limit of statistical deviation. Thus, we can conclude that analytic data obtained by the FdDH-based method are more reliable than those obtained by the chemical methods. Because of this very important analytic feature of the enzymatic method, it can be recommended for practical applications in preference to the chemical methods, which are labor- and time-consuming, needing distillation of the samples or performance of a standard addition test (in the case of phenol contamination).

View this table:
Table 2

Mutant HpPyc1p proteins obtained by pentapeptide insertions

Chemical methodsSamplesChromotropic acidMBTHFdDH-based method
DK19.3 ± 0.619.56 ± 0.517.89 ± 0.59
DK28.7 ± 0.508.06 ± 0.326.66 ± 0.26
DK37.2 ± 0.337.84 ± 0.366.88 ± 0.41
DK47.1 ± 0.366.3 ± 0.467.58 ± 0.32
DK51.65 ± 0.351.2 ± 0.152.32 ± 0.08
DK64.64 ± 0.244.99 ± 0.0595.73 ± 0.32
DK71.62 ± 0.171.96 ± 0.202.47 ± 0.15
WW-A112 ± 4.5 (standard addition test)116 ± 5.1 (standard addition test)
84.4 ± 6.5 (routine test)111 ± 6.1 (routine test)
  • MBTH, 3-methyl-2-benzothiazoline hydrazone hydrochloride.

Figure 4

Standard addition test for the formaldehyde assay by the chromotropic acid method (a) and the FdDH-based method (b). Curve 1 corresponds to the calibration experiment performed for water solutions of formaldehyde (traditional calibration), and curve 2 corresponds to the standard addition calibration (formaldehyde was added at different concentrations to the diluted real sample; WW-A). Some statistical data are presented on the graphs: parameters of linear regression (coefficients of the equation Y=A+BX, where Y=OD, X=formaldehyde concentration (mM), A=OD of the variant without addition of exogenous formaldehyde, and B=slope value); R=linear regression coefficient.


This work was financially supported by the project INTAS OPEN CALL 03-51-6278, NATO Linkage Grant LST.NUKR.CLG 980621, and Polish scientific project KBN 3PO4B00323.


  • Editor: Marten Veenhuis


View Abstract