Applied Biochemistry and Biotechnology

, Volume 160, Issue 6, pp 1685–1698

Toxicity of Methanol and Formaldehyde Towards Saccharomyces cerevisiae as Assessed by DNA Microarray Analysis

Authors

    • Department of Food BiotechnologyHokkaido Food Processing Research Center
  • Satomi Murata
    • Japan Pulp & Paper Research Institute, Inc.
  • Yumiko Iwahashi
    • National Food Research Institute (NFRI)National Agriculture and Food Research Organization (NARO)
  • Emiko Kitagawa
    • Applied Science Business Unit, Technical Support Team, Array Group, Genomics Sequencing&Array Biz DepartmentRoche Diagnostics K. K.
  • Ryoji Nakagawa
    • Department of Food BiotechnologyHokkaido Food Processing Research Center
  • Tazusa Hashido
    • Department of Food BiotechnologyHokkaido Food Processing Research Center
  • Hitoshi Iwahashi
    • Health Technology Research Center (HTRC)National Institute of Advanced Industrial Science and Technology (AIST)
Article

DOI: 10.1007/s12010-009-8684-y

Cite this article as:
Yasokawa, D., Murata, S., Iwahashi, Y. et al. Appl Biochem Biotechnol (2010) 160: 1685. doi:10.1007/s12010-009-8684-y

Abstract

To assess the toxicity of the C1 compounds methanol and formaldehyde, gene expression profiles of treated baker’s yeast were analyzed using DNA microarrays. Among approximately 6,000 open reading frames (ORFs), 314 were repressed and 375 were induced in response to methanol. The gene process category “energy” comprised the greatest number of induced genes while “protein synthesis” comprised the greatest number of repressed genes. Products of genes induced by methanol were mainly integral membrane proteins or were localized to the plasma membrane. A total of 622 and 610 ORFs were induced or repressed by formaldehyde, respectively. More than one-third of the genes found to be strongly repressed by formaldehyde belonged to the “protein synthesis” functional category. Conversely, genes in the subcategory of “nitrogen, sulfur, and selenium metabolism” within “metabolism” and in the category of “cell rescue, defense, and virulence” were up-regulated by exposure to formaldehyde. Our data suggest that membrane structure is a major target of methanol toxicity, while proteins were major targets of formaldehyde toxicity.

Keywords

MicroarrayMethanolFormaldehydeToxicitySaccharomyces cerevisiae

Introduction

Methanol is a transparent, colorless liquid produced in very large volume. It is widely used as an industrial solvent and used in many consumer products, in adhesives, antifreeze, paints, and cleaning solutions, as an ingredient. Moreover, methanol has been proposed to be use as an alternative to fossil fuels. It is also detected as a by-product in sewage treatment, fermentation, and paper production. Vegetables and fruits contain methanol in some degree. There are some incident on surrogate alcohols (i.e., nonbeverage alcohols and illegally produced alcohols) and appears to be a major public health problem in some countries. Methanol is reported to be toxic even in low dose, like inhalation of carburetor cleaning fluid fumes [1]. Thus, all of us share some risks to expose to high dose of methanol accidentally, occupationally, or may be intentionally. It is reported that daily low level (at 1/4 LD50) exposure to methanol also influence a normal immune response on rat [2].

Similarly, formaldehyde is a common environmental agent found in exhaust gas and as a component in paint, glue, and many other medicinal and industrial products. Formaldehyde is also produced endogenously and exists in cells and tissues of our body; however, exposure to large amounts of formaldehyde vapor can irritate the nasal mucosa and may potentially be carcinogenic. It also gives a positive Ames test.

Methanol and formaldehyde are intermediates in the oxidation of methane and are referred to as C1 compounds. Formaldehyde is much more toxic than methanol [3]. LD50 values for formaldehyde and methanol by oral administration in the rat is 800 and 5,628 mg/kg, respectively. However, cytotoxicity of methanol has not been fully elucidated and it is not clear if the toxicity of methanol is mainly due to its possible conversion to formaldehyde by exposed cells.

Recently, toxicogenomics, which analyze global genetic response to toxic substances and elucidate the etiology has been developed. And budding yeast, Saccharomyces cerevisiae is one of the most studied model eukaryote and more than 70% of open reading frames (ORFs) are verified. The information of molecular function, biological process, cellular localization, related metabolic pathway, and others of each gene is collected, integrated, and updated. In this study, we employed DNA microarray analysis to explore and compare the toxicity of these two C1 compounds in S. cerevisiae.

Materials and Methods

Pre-cultures of S. cerevisiae S288C (Mat alpha SUC2 mal mel gal2 CUP1) were grown in YPD (2% polypeptone, 1% yeast extract, and 2% glucose) at 25 °C for 2–3 days. This strain was chosen because the DNA microarrays were produced using S288C DNA as the PCR template.

The pre-culture was diluted and grown overnight to an optical density (OD 660 nm) of 1.0. In the dose–response experiments, methanol or formaldehyde was added to the YPD cultures which were then grown for another 2 h along with control cultures without addition of these C1 compounds, respectively. Cells were harvested by centrifugation and stored at −80 °C until RNA was extracted.

Microarray experiments were carried out as described [4]. Poly (A)+ RNA was purified from three independent cultures of both control and treated cells. mRNA from C1 compound-treated cells was fluorescently labeled with Cy5-dUTP while mRNA from untreated cells was labeled with Cy3-dUTP. The two fluorescently labeled cDNAs were mixed and then hybridized to yeast microarrays (ver.2.0, DNA Chip Research, Inc., Yokohama, Japan) for 16–20 h at 65 °C. About 6,000 genes could be analyzed on these microarrays under these conditions [5]. Detected signals for each ORF were normalized by intensity-dependent (LOWESS) methods (http://www.silicongenetics.com/cgi/Sig.cgi/index.smf). Genes designated induced or repressed in treated cells (1) passed a one sample t test (P value cutoff 0.05) and (2) exhibited greater or less than twofold expression, respectively, relative to the control. Induced and repressed genes were characterized according to accepted gene ontology categories (MIPS, Munich Information Center for Protein Sequences, http://mips.gsf.de/ and Saccharomyces Genome Database, http://www.yeastgenome.org/). The data obtained in this experiment are available under accession number GSE9231 in the Gene Expression Omnibus Database (http://www.ncbi.nlm.nih.gov/geo/).

Clustering analyses were applied using GeneSpring (Ver.7.3.1). The settings for calculations were as follows: similarity was measured by Pearson’s correlation, the separation ratio was 1.0, and the minimum distance was 0.001.

Results and Discussion

To elucidate the toxicity of methanol and formaldehyde, appropriate concentrations for microarray experiments had to be determined. From initial dose–response experiments, final concentrations of 1.23 M methanol and 1.8 mM formaldehyde were chosen, as these doses were found to significantly inhibit growth without killing cells. Generally, in lower concentrations of chemicals which have no effect on growth, we are not able to detect differences in gene expressions, and in higher concentrations which cause cellular death, we are not able to extract enough amount and quality of mRNA. In the condition above, it was sufficient to induce large changes in the gene expression profiles.

Methanol Toxicity

Methanol was found to repress 314 and to induce 375 ORFs. Table 1 shows the number of induced genes by category of biological function. Among 375 induced ORFs, 115 were categorized in “metabolism” and 60 of them within the subcategory of “C-compound and carbohydrate metabolism”. Genes within the subcategory of “energy” (13.8%) comprised the greatest fraction. Functional categories for the repressed genes were also analyzed. “Protein synthesis (15.8%)” was found to comprise the largest category of repressed genes.
Table 1

Categories of ORFs induced by methanol and formaldehyde treatment.

Gene categories

Methanol treatment

Formaldehyde treatment

Number of induced genes

Percentage in each category

Number of induced genes

Percentage in each category

Metabolism (1519)

115

7.6

175

11.5

Energy (369)

51

13.8

37

10.0

Cell cycle and DNA processing (1010)

21

2.1

67

6.6

Transcription (1078)

23

2.1

58

5.4

Protein synthesis (480)

2

0.4

9

1.9

Protein fate (1155)

31

2.7

134

11.6

Protein with binding function or cofactor requirement (1049)

31

3.0

102

9.7

Regulation of metabolism and protein function (249)

5

2.0

20

8.0

Cellular transport, transport facilitation, and transport routes (1042)

68

6.5

105

10.1

Cellular communication/signal transduction mechanism (234)

6

2.6

16

6.8

Cell rescue, defense and virulence (554)

50

9.0

123

22.2

Interaction with the cellular environment (463)

19

4.1

50

10.8

Transposable elements, viral, and plasmid proteins (120)

0

0.0

0

0.0

Cell fate (273)

5

1.8

20

7.3

Development (69)

3

4.4

6

8.7

Biogenesis of cellular components (863)

30

3.5

71

8.2

Cell type differentiation (452)

16

3.5

40

8.9

Unclassified proteins (1394)

98

7.0

133

9.5

Numbers in parentheses indicate total ORFs in each category

Within the category of “energy”, 4 aldehyde dehydrogenases ALD4 (major mitochondrial; 4.92-fold), ALD3 (3.83-fold), ALD2 (3.68-fold), and ALD6 (major cytoplasmic; 2.11-fold), were strongly up-regulated by methanol treatment. Products of these genes may be involved in oxidizing the intermediate formaldehyde to formic acid. As shown below, these ALDs were also strongly induced by formaldehyde treatment. Other alcohol dehydrogenases, ADH5 (1.69-fold), ADH7 (1.65-fold), and ADH2 (1.59-fold) were not induced more than twofold. Adh5p reduces acetaldehyde to ethanol during fermentation of glucose, while Adh2p oxidizes ethanol to acetaldehyde [6, 7]. ADH7 is a non-essential gene that encodes a member of the cinnamyl alcohol dehydrogenase family [8]. These data suggest that yeast slowly metabolizes methanol to formaldehyde to avoid abrupt accumulation of the more toxic latter compound. The presumed methanol detoxification pathway via formaldehyde is shown in Fig. 1a. The glutathione-dependent branch is not considered induced.
https://static-content.springer.com/image/art%3A10.1007%2Fs12010-009-8684-y/MediaObjects/12010_2009_8684_Fig1_HTML.gif
Fig. 1

Expected methanol and formaldehyde detoxification pathways in yeast. Induction and repression values of respective ORFs upon cell exposure to (a) methanol and (b) formaldehyde are shown in parentheses. t test P values less than 0.05 are indicated with a single asterisk (*); t test P value less than 0.01 are indicated with two asterisks

Two putative aryl-alcohol dehydrogenases, AAD10 (2.97-fold) and AAD6 (2.08-fold) were also up-regulated in the “energy” category. However, there is a very high nucleotide sequence similarity among all seven AADs, AAD3, AAD4, AAD6, AAD10, AAD14, AAD15, and AAD16. While AAD4 and AAD6 were reported to be induced by oxidative stress in a Yap1p-dependent manner [9], they were not shown to contribute to alkyl-alcohol degradation.

Glucokinase, GLK1 (2.26-fold), and hexokinase, HXK1 (2.29-fold), were found to be up-regulated, suggesting that glycolysis was slowed or halted [10, 11]. Herrero and coworkers [12] reported that transcription of these genes was de-repressed when cells were grown on a non-fermentable carbon source. During methanol treatment, it is possible that cells are oxidizing the methanol as a carbon source.

In the category of “metabolism”, just three ORFs, MHT1 (2.12-fold), MET13 (2.09-fold), and MET17 (2.04-fold), within the subcategory of “metabolism of methionine and metabolism of cysteine” were up-regulated. This is the one of major differences found between methanol and formaldehyde toxicity. Far more ORFs involved in methionine and cysteine metabolism were up-regulated during formaldehyde treatment, as shown below. Seventeen ORFs, SUT1, HSP12, OPI3, POT1, CAT2, POX1, GUT2, CRC1, TIP1, GUT1, YDR018C, GPT2, ATG15, FAA2, YML131W, PCS60, and AGP2, within the subcategory of “lipid, fatty acid, and isoprenoid metabolism” were also found to be induced. This may suggest that cells are making more membrane components to overcome methanol toxicity.

Among the 375 genes induced by methanol treatment, Table 2 lists the most highly induced genes. Fifty were categorized within “cell rescue, defense, and virulence”. As shown in Table 3, 12 genes were induced within the subcategory of “oxidative stress response” (21.8%) while ten were induced within the subcategory of “oxygen and radical detoxification” (38.5%). These data indicate that oxidative stress is a major consequence of exposure to methanol.
Table 2

Most highly induced ORFs by methanol.

Systematic name

Standard name

Expression ratio

Description

YGL158W

RCK1

21.50

Ser/thr protein kinase

YBR072W

HSP26

13.32

Heat shock protein

YDL218W

 

11.15

Weak similarity to hypothetical protein YNR061c

YOL052C-A

DDR2

10.74

Heat shock protein DDRA2

YKL217W

JEN1

10.62

Lactate and pyruvate permease

YNL195C

 

9.91

Conserved hypothetical protein

YKL187C

 

9.37

Mitochondrial membrane protein

YDR034W-B

 

9.33

Protein of unknown function

YPL223C

GRE1

8.89

Induced by osmotic stress

YOL084W

PHM7

7.49

Protein of unknown function, expression is regulated by phosphate levels

YMR175W

SIP18

7.27

Osmotic stress protein

YDR070C

FMP16

7.26

Found in mitochondrial Proteome

YEL070W

DSF1

7.18

Expression is glucose repressed by Mig1 and Mig2

YOR383C

FIT3

7.16

Mannoprotein that is incorporated into the cell wall via a glycosylphosphatidylinositol (GPI) anchor

YBR116C

 

6.99

Questionable protein

YHR139C

SPS100

6.89

Sporulation-specific wall maturation protein

YHR096C

HXT5

6.72

Hxt family protein with intrinsic hexose transport activity

YGR256W

GND2

6.48

Phosphogluconate dehydrogenase

YGR236C

SPG1

6.43

Protein required for survival at high temperature during stationary phase

YPR192W

AQY1

6.41

Water channel that mediates the transport of water across cell membranes

Table 3

Subcategories of ORFs in “cell rescue” category induced by methanol and formaldehyde treatment.

Subcategories of “cell rescue”

Methanol treatment

Formaldehyde treatment

Number of induced genes

Percentage of induced genes

Number of induced genes

Percentage of induced genes

Cell rescue, defense and virulence (554)

50

9.0

123

22.2

Stress response (450)

41

9.1

107

23.8

Oxidative stress response (55)

12

21.8

30

54.6

Osmotic and salt stress response (59)

6

10.2

12

20.3

pH stress response (8)

1

12.5

1

12.5

Heat shock response (20)

4

20.0

6

30.0

Cold shock response (3)

0

0.0

0

0.0

Unfolded protein response (69)

5

7.3

11

15.9

DNA damage response (77)

1

1.3

7

9.1

Nutrient starvation response (15)

0

0.0

8

53.3

Electromagnetic waves stress response (2)

0

0.0

0

0.0

Disease, virulence, and defense (35)

3

8.6

7

20.0

Resistance proteins (33)

3

9.1

5

15.2

Virulence, disease factors (2)

0

0.0

1

50.0

Detoxification (117)

17

14.5

36

30.8

Detoxification involving cytochrome P450 (4)

1

25.0

0

0.0

Detoxification by modification (8)

3

37.5

6

75.0

Detoxification by export (3)

0

0.0

0

0.0

Oxygen and radical detoxification (26)

10

38.5

18

69.2

Degradation/modification of foreign compounds (1)

0

0.0

0

0.0

Degradation/modification of foreign polynucleotides (1)

0

0.0

0

0.0

Numbers in parentheses indicate total ORFs in each subcategory

A number of studies have examined ethanol tolerance in yeast. It is conceivable that methanol elicits some of the same responses as this C2 alkanol. Ethanol-tolerant mutants of sake strains of yeast have been reported to be resistant to killer toxin and to a cell wall lytic enzyme [13]. Both phenotypes are suggestive of an altered cell wall structure. It is has also been reported that the cell membrane is the primary target of ethanol toxicity and that tolerance can be partly explained in terms of membrane lipid composition [1418].

The localization of the products of the methanol-induced genes was also analyzed according to MIPS. A relatively large percentage was localized to the “cell wall”, “plasma membrane”, and “integral membrane/endomembrane” (Table 4). Products of genes induced by formaldehyde were also found to be localized to the cell wall and membrane. However, a total of 622 genes were found to be induced by formaldehyde while 375 were induced by methanol. We define the “impact” of methanol and formaldehyde on a given cell component as the ratio of the number of component-associated genes induced to the total number of induced genes (Table 4). By this measure, methanol has a substantial impact on the cell wall and on membranes.
Table 4

Cellular localization of products of ORFs induced by methanol and formaldehyde.

Localization

Methanol treatment

Formaldehyde treatment

Numbera

%b

Impactc

Number

%

Impact

Extracellular (54)

12

22.2

3.2

12

22.2

1.9

Bud (150)

5

3.3

1.3

7

4.7

1.1

Cell wall (44)

8

18.2

2.1

12

27.3

1.9

Cell periphery (216)

18

8.3

4.8

25

11.6

4.0

Plasma membrane (186)

27

14.5

7.2

18

9.7

2.9

Integral membrane/endomembranes (172)

27

15.7

7.2

32

18.6

5.1

Cytoplasm (2845)

166

5.8

44.3

339

11.9

54.5

Cytoskeleton (204)

4

2.0

1.1

15

7.4

2.4

ER (552)

28

5.1

7.5

87

15.8

14.0

Golge (158)

1

0.6

0.3

10

6.3

1.6

Transport vesicles (141)

1

0.7

0.3

10

7.1

1.6

Nucleus (2139)

81

3.8

21.6

210

9.8

33.8

Mitochondria (1047)

76

7.3

20.3

116

11.1

18.6

Peroxisome (52)

6

11.5

1.6

3

5.8

0.5

Endosome (58)

0

0.0

0.0

10

17.2

1.6

Vacuole (284)

17

6.0

4.5

43

15.1

6.9

Microsomes (5)

0

0.0

0.0

1

20.0

0.2

Lipid particles (27)

3

11.1

0.8

5

18.5

0.8

Punctate composite (140)

11

7.9

2.9

15

10.7

2.4

Ambiguous (237)

12

5.1

3.2

39

16.5

6.3

Number in parentheses indicate total ORFs in each localization or cell component

aNumber of induced genes whose products are found associated with the indicated localization

b(Number of induced gene products associated with each localization)/(total number of gene products associated with the localization) × 100

cImpact is defined as the (number of component-associated genes that induced)/(the total number of induced genes) × 100

Some discrepancies were noted between this study and earlier reports on ethanol tolerance. Yazawa et al. [19] reported that URA7 and GAL6 (LAP3) null mutants exhibited greater tolerance to ethanol than the wild-type parent strain. In our study, URA7 was also strongly repressed (0.20-fold), but LAP3 was mildly induced (1.48-fold). Some studies have reported that ergosterol plays an important role in ethanol tolerance [2022]. In the present study, no ORFs involved in ergosterol biosynthesis were found to be up-regulated more than twofold by methanol, from ERG10 (0.82-fold; cytosolic acetyl-CoA C-acetyltransferase) to ERG4 (1.11-fold; sterol C-24 reductase). Fujita et al. [23] compared yeast response to various straight-chain alcohols using cDNA microarrays and concluded that gene expression profiles differed as a function of the log Pow values of the alcohols. They also reported genes essential for alcohol tolerance based on screening a library of deletion mutants [24]. Based on a comparison of their results with the present study, we conclude that toxicity induced by methanol differs from that induced by ethanol.

Formaldehyde Toxicity

Formaldehyde was found to induce 622 ORFs more than twofold and to repress 610 ORFs more than twofold, relative to the control. As shown in Table 1, 22.2% of the up-regulated genes belong to the category of “cell rescue, defense, and virulence” while 11.6% belong to the category of “protein fate”. More than one-third of the genes (34.6%) in the category of “protein synthesis” were strongly repressed. However, repression of genes categorized in “protein synthesis” was often observed [5, 2527] in other chemical treatments and changes in growth environment. Sirisattha et al. [28] have pointed out that down-regulated genes might reflect the decreased growth rate. Also, Brauer et al. [29] demonstrated that differences in growth rates resulted in differences in gene expressions. And they consider the transcriptional response was closely correlated with the growth rate, not with cause of growth retardation. The repression of protein synthesis is a common observation in yeast which shows lower growth rate [4] and we cannot ascribe the repression to characteristic of formaldehyde toxicity.

Five methionine biosynthesis genes, MET17, MET3, MET16, MET10, and MET6, and two methionine permease genes, MUP3, and MUP1, were included among the 20 most highly induced genes (Table 5). ADH1, encoding alcohol dehydrogenase that is responsible for converting acetaldehyde to ethanol, was strongly repressed 0.13-fold. It has been reported that overexpression of ADH1 in a wild-type strain rendered it hyper-resistant to formaldehyde [30]. Four other genes involved in aldehyde metabolism, AAD16, AAD6, AAD14, and AAD3, were also highly induced. Most of the genes needed for methionine biosynthesis are also needed for glutathione biosynthesis, and almost all of the latter genes were up-regulated. Twenty ORFs in the category of “metabolism of methionine” (35 entries) were up-regulated. On the other hand, only three ORFs in this same category were found to be up-regulated by treatment with methanol.
Table 5

Most highly induced ORFs by formaldehyde treatment.

Systematic name

Standard name

Expression ratio

Description

YDL218W

 

331.79

Weak similarity to hypothetical protein YNR061c

YHR139C

SPS100

118.03

Sporulation-specific wall maturation protein

YKL071W

 

104.47

Protein of unknown function, cytoplasmic

YBR008C

FLR1

90.54

Putative H+ antiporter regulated by yAP-1

YLR303W

MET17

73.41

O-acetylhomoserine sulfhydrylase

YFL057C

AAD16

72.23

Aryl-alcohol dehydrogenase

YFL056C

AAD6

49.33

Putative aryl-alcohol dehydrogenase, stress response

YHR096C

HXT5

36.73

Hxt family protein with hexose transport activity

YNL331C

AAD14

35.02

Putative aryl-alcohol dehydrogenase

YCR107W

AAD3

34.22

Aryl-alcohol dehydrogenase

YJR010W

MET3

27.72

Sulfate adenylyltransferase

YPR167C

MET16

26.12

3’-phosphoadenylylsulfate reductase

YLL060C

GTT2

23.75

Glutathione S-transferase

YCR102C

 

23.61

Similarity to Zinc-type alcohol dehydrogenase

YFR030W

MET10

23.11

Sulfite reductase flavin-binding subunit

YHL036W

MUP3

22.55

Low-affinity methionine permease

YLR460C

 

21.46

Protein of unknown function

YGR055W

MUP1

21.07

High affinity methionine permease

YGR197C

SNG1

21.06

Probable transporter, resistance to nitrosoguanidine

YER091C

MET6

20.61

5-methyltetrahydropteroyltriglutamate—homocysteine methyltransferase

Thus, glutathione appears to play a very important role in formaldehyde detoxification. The glutathione-dependent formaldehyde dehydrogenase gene SFA1 [31] was found to be strongly up-regulated, 8.30-fold (Fig. 1b), while the gene encoding S-formylglutathione hydrolase, YJL068C [32], whose product regenerates glutathione, was also up-regulated (3.70-fold). Expression of ADH6 was induced 2.20-fold by formaldehyde treatment (Fig. 1b). However, it is not clear whether cells were able to alleviate the toxicity of formaldehyde efficiently by reducing it to methanol because Adh6p is a medium-chain alcohol dehydrogenase [33].

Within the category of “cell rescue”, more than half of the genes belonging to the subcategories of “oxidative stress response”, and “oxygen and radical detoxification” were induced (Table 3). As noted above, one of the major radical scavengers, glutathione, is synthesized via methionine and cysteine biosynthesis.

Within the “protein fate” category, 32 ORFs (36.0%) in the “protein processing (proteolytic)” subcategory and 63 ORFs (24.6%) in the “protein/peptide degradation” subcategory were induced (Table 6). These data suggest that formaldehyde treatment interferes with protein synthesis by blocking ribosome assembly and that denatured proteins were degraded by proteolysis. In contrast, only 32 ORFs (2.8%) in the “protein fate” category were induced and just 15.8% of the genes in the “protein synthesis” category were repressed by methanol treatment.
Table 6

Subcategories of methanol and formaldehyde-induced genes in “protein fate” category.

Subcategories of “protein fate”

Methanol treatment

Formaldehyde treatment

No. of induced genes

Percentage of induced genes

No. of induced genes

Percentage of induced genes

Protein fate (folding, modification, destination) (1155)

32

2.8

134

11.6

Protein folding and stabilization (93)

2

2.2

17

18.3

Protein targeting, sorting, and translocation (281)

4

1.4

20

7.1

Protein modification (616)

15

2.4

76

12.3

 With fatty acids (33)

1

3.0

1

3.0

 With sugar residues (69)

0

0.0

1

1.5

 By phosphorylation, de-phosphorylation, auto-phosphorylation (186)

6

3.2

14

7.5

 By acetylation, de-acetylation (69)

1

1.4

2

2.9

 By ubiquitination, de-ubiquitination (79)

1

1.3

15

19.0

 By ubiquitin-related proteins (21)

0

0.0

3

14.3

 Post-translational modification of amino acids (26)

0

0.0

1

3.9

 Protein processing (proteolytic) (89)

5

5.6

32

36.0

Assembly of protein complexes (199)

5

2.5

19

9.6

Protein/peptide degradation (256)

11

4.3

63

24.6

 Cytoplasmic and nuclear (188)

7

3.7

51

27.1

 Lysosomal and vacuolar (26)

2

7.7

6

23.1

Thirteen ORFs in the subcategory of “DNA recombination and DNA repair” within the category of “cell cycle and DNA processing” were induced and seven ORFs were mildly induced (1.5–2.0 fold) by formaldehyde treatment, whereas only five ORFs in this subcategory were induced by methanol treatment (Table 7).
Table 7

DNA repair genes induced by methanol and formaldehyde treatment.

Systematic name

Standard name

Fold induction

Description

Methanol treatment

YFL014W

HSP12

5.87

Heat shock protein

YGR144W

THI4

3.09

Protein involved in thiamine biosynthesis and DNA repair

YOR346W

REV1

2.48

DNA repair protein

YML058W-A

HUG1

2.41

Hydroxyurea and UV and gamma radiation induced

YDR030C

RAD28

2.08

Protein involved in transcription-coupled repair

Formaldehyde treatment

YFL014W

HSP12

8.57

Heat shock protein

YIL143C

SSL2

4.18

DNA helicase

YER142C

MAG1

4.03

3-methyladenine DNA glycosylase

YGL163C

RAD54

4.00

DNA-dependent ATPase of the Snf2p family

YMR173W

DDR48

3.74

Heat shock protein

YGL058W

RAD6

3.12

E2 ubiquitin-conjugating enzyme

YKL025C

PAN3

3.09

Component of the Pab1p-dependent poly(A) ribonuclease

YAR007C

RFA1

2.92

DNA replication factor A, 69 kDa subunit

YNL312W

RFA2

2.74

DNA replication factor A, 36 kDa subunit

YML032C

RAD52

2.72

Recombination and DNA repair protein

YDR227W

SIR4

2.65

Silencing regulatory and DNA repair protein

YKL213C

DOA1

2.21

Protein involved in ubiquitin-dependent proteolysis

YEL037C

RAD23

2.13

Nucleotide excision repair protein

Methanol vs. Formaldehyde Toxicity

One hundred and thirty-six ORFs were repressed by both methanol and formaldehyde and 39 of them were in the category of protein synthesis (8.1%). One hundred and twenty-six ORFs were induced by both C1 compounds, including putative aryl-alcohol dehydrogenases, AAD6 and AAD10, aldehyde dehydrogenases, ALD2, ALD3, and ALD4, glutaredoxin, GRX1, glutathione S-transferase, GTT1, heat shock proteins, HSP12, HSP26, HSP30, and HSP31, Hxt family proteins with intrinsic hexose transport activity, HXT5 and HXT9, superoxide dismutases, SOD1 and SOD2, among others. However, among the more than 20 ORFs associated with the yeast proteasome that were up-regulated by formaldehyde only UBC5 was up-regulated by methanol.

A total of 21 genes were induced >4-fold by both treatments. Among them, nine are associated with lipids, membranes, and the cell wall. HSP12 and HSP26 are known to be involved in ethanol tolerance in a gal6 null mutant and their expression was reported to be up-regulated by ethanol stress [34, 35]. In the present study, both genes were also up-regulated by methanol and formaldehyde treatment.

Gasch et al. [36] found two clusters of genes which were up-regulated or down-regulated, respectively and named them the “environmental stress response (ESR) genes”. These ESR genes commonly shift their expression responses regardless of stress conditions. And many of the ESR-induced and ESR-repressed genes were correlated with growth rate [29]. However, among 375-induced ORFs in methanol treatment, and among 622-induced ORFs in formaldehyde treatment, only about 100 genes were growth rate correlated and ESR-induced genes, respectively (data not shown). These findings suggest that most of the up-regulated ORFs in C1 compound treatments were thought to be mainly compound specific, but not dependent on growth rate.

Cluster analysis was used to characterize the toxicity of these C1 compounds (Fig. 2) relative to previous DNA microarray data generated in our laboratory for heavy metals, detergents, food additives, herbicides, and physiologically active substances. The gene expression profile for formaldehyde was not grouped in a same cluster that included methanol but was grouped in a cluster that included two fungicides, tetrachloroisophthalonitrile, and thiuram, and a mycotoxin, patulin. Patulin and Thiuram also activated protein degradation and DNA repair genes [4, 27, 37]. These chemicals and the mycotoxin are known to attack thiol and amino groups and to promote protein denaturation [3840].
https://static-content.springer.com/image/art%3A10.1007%2Fs12010-009-8684-y/MediaObjects/12010_2009_8684_Fig2_HTML.gif
Fig. 2

Cluster analysis of the expression profiles after methanol and formaldehyde treatment. Hierarchical cluster analysis was performed using GeneSpring as described in the text

Conclusions

Formaldehyde is much more toxic than methanol, based on its inhibition of yeast growth at a much lower concentration than methanol. The concentration of methanol and formaldehyde used in the present study was 1.23 M and 1.8 mM, respectively. As shown in Fig. 1, methanol strongly induced ALDs, in contrast to modest or no induction of ADHs, presumably to mediate rapid elimination of the formaldehyde derived from the metabolism of methanol. This is a remarkable and very rational toxicity-evading mechanism to minimize concentrations of a highly toxic metabolic intermediate.

As shown in Table 7, formaldehyde significantly stimulated genes involved in DNA repair. It is well known that formaldehyde is mutagenic to experimental animals [41], bacteria, and to lower eukaryotes [42]. Exposure to formaldehyde leads to the formation of DNA-protein crosslinks as the major form of DNA damage [43, 44]. While methanol and formaldehyde induced 126 and repressed 136 of the same ORFs, respectively, the toxicities of two chemicals were found to differ somewhat. In this study, a haploid strain S. cerevisiae S288C was used. The toxicity due to formaldehyde may be exacerbated in a haploid compared to a diploid strain. And this may have contributed to the differences in toxicity observed between the methanol and formaldehyde.

Cluster analysis highlighted these differences. We interpret the microarray results as follows. Methanol caused oxidative stress and also seems to affect membrane integrity. Formaldehyde denatures proteins and DNA via oxidization. To minimize the toxicity of formaldehyde, whether added directly or generated via oxidation of added methanol, yeast synthesizes glutathione, degrades denatured proteins, and repairs damaged DNA. Formaldehyde is a much worse toxicant than methanol, as evidenced by the induction of aldehyde-oxidizing genes upon exposure to methanol as a novel mechanism for minimizing concentrations of methanol-generated formaldehyde.

Acknowledgements

We thank Mr. Tsuyoshi Awa for arranging for this study and are grateful to Dr. Michihiko Saito for technical advice.

Copyright information

© Humana Press 2009