Address correspondence and reprint requests to Dr. Ralf Dringen, Centre for Biomolecular Interactions Bremen, University of Bremen, PO. Box 330440, D-28334 Bremen, Germany. E-mail: email@example.com
Formaldehyde is an environmental pollutant that is also generated in substantial amounts in the human body during normal metabolism. This aldehyde is a well-established neurotoxin that affects memory, learning, and behavior. In addition, in several pathological conditions, including Alzheimer's disease, an increase in the expression of formaldehyde-generating enzymes and elevated levels of formaldehyde in brain have been reported. This article gives an overview on the current knowledge on the generation and metabolism of formaldehyde in brain cells as well as on formaldehyde-induced alterations in metabolic processes. Brain cells have the potential to generate and to dispose formaldehyde. In culture, both astrocytes and neurons efficiently oxidize formaldehyde to formate which can be exported or further oxidized. Although moderate concentrations of formaldehyde are not acutely toxic for brain cells, exposure to formaldehyde severely affects their metabolism as demonstrated by the formaldehyde-induced acceleration of glycolytic flux and by the rapid multidrug resistance protein 1-mediated export of glutathione from both astrocytes and neurons. These formaldehyde-induced alterations in the metabolism of brain cells may contribute to the impaired cognitive performance observed after formaldehyde exposure and to the neurodegeneration in diseases that are associated with increased formaldehyde levels in brain.
The neurotoxin formaldehyde is an environmental pollutant that is also generated during normal brain metabolism. The levels of formaldehyde in brain increase with age and in some neurodegenerative disorders. As excess formaldehyde accelerates glycolysis and glutathione export in neural cells, formaldehyde-induced alterations in brain metabolism and oxidative stress may contribute to the pathological progression of neurodegenerative disorders.
Formaldehyde (HCHO) is the simplest aldehyde that is also known as methanal. This compound was first described in 1855 by Alexander Butlerov, while its chemical synthesis by methanol dehydration was first achieved in 1867 by August Wilhelm von Hofmann (Salthammer et al. 2010). In the following decades, the properties of formaldehyde were extensively studied and this compound was one of the earliest to obtain a CAS registry number (50-00-0). Formaldehyde is highly reactive. It can undergo hydration and forms hemiacetals with alcohols or thiohemiacetals with thiols. Formaldehyde also reacts with amines to form Schiff bases and cross-links proteins by forming methylene bridges between amino groups (Metz et al. 2004, 2006). This high reactivity of formaldehyde is the reason for its extensive use in industries (Tang et al. 2009).
Due to its protein cross-linking ability, formaldehyde is frequently used for tissue preservation and fixation (Nazarian et al. 2009). Formalin solution that is used in pathology contains 35% formaldehyde, while for fixation of tissues, tissue sections, or cultured cells, a 4% formaldehyde solution is frequently used (Kiernan 2000). Such a 4% formaldehyde solution contains the aldehyde in a concentration of above 1 M. Thus, the concentrations of formaldehyde that are used for technical processes are several orders of magnitude higher than the concentrations of formaldehyde (0.1–0.4 mM) that are found in body fluids and tissues under normal and pathological conditions (Heck and Casanova 2004; Tong et al. 2013a).
Endogenous and exogenous sources of formaldehyde
Formaldehyde exposure is caused by the generation of this aldehyde within the body and can also be a consequence of contact with elevated levels of environmental formaldehyde (Fig. 1). Some of the endogenous enzymatic reactions that generate formaldehyde as well as exogenous sources of formaldehyde are described below.
Formaldehyde is the oxidation product of methanol. This alcohol can be generated within the body by hydrolysis of protein carboxymethyl esters either non-enzymatically or catalyzed by methylesterases (Lee et al. 2008). In addition, accidental or intentional intake of methanol will further expose the body to this alcohol. In cells, methanol is oxidized to formaldehyde by alcohol dehydrogenase (ADH) 1, by catalase or by a non-enzymatic reaction of methanol with hydroxyl radicals (Harris et al. 2003; MacAllister et al. 2011). In humans and primates, ADH1 appears to be predominately responsible for methanol oxidation, while the majority of methanol oxidation in rats has been reported to be mediated by catalase (Tephly 1991; Skrzydlewska 2003).
Another endogenous source of formaldehyde are semicarbazide-sensitive amine oxidases (SSAO) which represent a group of copper-containing amine oxidases that are inhibited by semicarbazide and most of them contain topa-quinone at their catalytic centre (Jalkanen and Salmi 2001; Yu et al. 2003). Oxidative deamination of methylamine by SSAO generates formaldehyde together with ammonia and hydrogen peroxide (Yu et al. 2003; O'Sullivan et al. 2004). In mammals, SSAO are either membrane-associated or circulate in a soluble form in the vascular system (Jalkanen and Salmi 2001). Among the SSAO, the vascular adhesion protein (VAP) 1 is one of the most extensively studied members of this group of enzymes (Smith and Vainio 2007; Jalkanen and Salmi 2008).
Formaldehyde is also generated as by-product of reactions catalyzed by lysine-specific demethylase (LSD) 1 and JmjC domain-containing histone demethylases (JHDM) (Cloos et al. 2008; Hou and Yu 2010). These enzymes remove methyl groups from lysine residues in histones, thereby altering the chromatin structure (Cheng and Zhang 2007; Cloos et al. 2008; Hou and Yu 2010; Izzo and Schneider 2010). LSD1 is a flavin-containing enzyme that selectively demethylates the mono- or dimethylated lysine residue in position 4 of histone H3 (Forneris et al. 2009; Hou and Yu 2010). On the other hand, JHDM can remove methyl groups from mono-, di-, or trimethylated lysine residues and require Fe2+ and α-ketoglutarate as cofactors (Cloos et al. 2008; Hou and Yu 2010).
In addition to endogenous sources, the body can also encounter environmental formaldehyde, since a number of commonly used products contain either formaldehyde or formaldehyde-releasing substances (Sasseville 2004; de Groot et al. 2009). Some examples of such products are construction materials, agricultural fertilizers, fumigants, paints, cosmetics, antiperspirants, polish, cleaning agents, and toiletries (Sasseville 2004; de Groot et al. 2009, 2010). In addition, formaldehyde can be produced and released from burning of wood, coal, tobacco, natural gas, and kerosene (de Groot et al. 2009; Laitinen et al. 2010). Moreover, foods like coffee, codfish, meat, poultry, and maple syrup naturally contain formaldehyde (Dhareshwar and Stella 2008; de Groot et al. 2009). Thus, this ubiquitously present compound can enter the human body by inhalation, ingestion, or entry through the skin.
One pertinent question is whether exogenous formaldehyde can pose a big threat to the central nervous system by entering the blood and ultimately reaching the brain after crossing the blood–brain barrier. In healthy individuals, the formaldehyde concentration in the blood is around 0.1 mM (Heck and Casanova 2004) and that in the brain is 0.2–0.4 mM (Tong et al. 2013a). Inhalation of moderate doses of formaldehyde does not severely increase the formaldehyde level in blood (Heck et al. 1985; Franks 2005). This is expected as the formaldehyde-oxidizing enzymes ADH3 and aldehyde dehydrogenase (ALDH) 2 (Fig. 1) are ubiquitously expressed in all tissues (Nishimura and Naito 2006; Alnouti and Klaassen 2008) and will quickly clear a low excess of environmentally derived formaldehyde. However, exposure to high concentrations of exogenous formaldehyde that exceeds the peripheral formaldehyde oxidation capacity will elevate the normal tolerable concentration of formaldehyde in the blood and could lead to neural damage. Indeed, exposure to exogenous formaldehyde has been reported to cause neurotoxicity in humans and animals and the extent of damage depends on the dose of formaldehyde and the duration of the exposure (Kilburn et al. 1985a, b; Songur et al. 2008, 2010). Especially, individuals who carry functional polymorphisms in the genes encoding for formaldehyde-metabolizing enzymes, ADH3 or ALDH2, which are discussed to be associated with reduced formaldehyde-oxidizing capacity (Hedberg et al. 2001; Wang et al. 2002), may be more vulnerable to neural damage by endogenously generated or environmental formaldehyde.
Metabolism of formaldehyde
Despite of the multiple endogenous and exogenous sources of formaldehyde, a low physiological level of formaldehyde in body fluids and tissue is maintained by the continuous action of cellular formaldehyde-metabolizing enzymes (Fig. 1). ADH1 is considered to play a negligible role in formaldehyde reduction to methanol because of its very high KM-value for formaldehyde (about 30 mM) (Skrzydlewska 2003). The formaldehyde oxidation product formate is generated by two independent pathways that are mediated by either the mitochondrial ALDH2 or the cytosolic ADH3 (Teng et al. 2001; Friedenson 2011; MacAllister et al. 2011). ADH3, also known as glutathione (GSH)-dependent formaldehyde dehydrogenase, oxidizes formaldehyde to formate in a two-step process (Harris et al. 2003; Staab et al. 2009; Thompson et al. 2010; MacAllister et al. 2011). In the first step, GSH reacts with formaldehyde in an enzyme-independent manner to form S-hydroxymethyl GSH that is subsequently used as ADH3 substrate to generate S-formyl GSH (Harris et al. 2003; Staab et al. 2009; Thompson et al. 2010; MacAllister et al. 2011). The conjugate S-formyl GSH is hydrolyzed by a thiolase to generate formate and GSH (Teng et al. 2001; Harris et al. 2003; MacAllister et al. 2011). Unlike ADH3, the reaction catalyzed by ALDH2 is a single-step GSH-independent process (Teng et al. 2001; MacAllister et al. 2011). Since ADH3 has a very low KM-value for S-hydroxymethyl GSH (less than 10 μM) compared to that of ALDH2 for formaldehyde (0.2–0.5 mM) (Casanova-Schmitz et al. 1984; Heck et al. 1990), ADH3 is likely to be especially important for the oxidation of low concentrations of formaldehyde.
The formate generated by formaldehyde oxidation can undergo further oxidization to carbon dioxide in a metabolic pathway involving tetrahydrofolate (THF), wherein formate is first converted to 10-formyl THF (Fig. 1) in an ATP-dependent reaction (Skrzydlewska 2003; Krupenko 2009; Krupenko et al. 2010). This reaction is catalyzed either by the cytosolic methylene tetrahydrofolate dehydrogenase (MTHFD) 1 or by its mitochondrial isoform MTHFD1L (Tibbetts and Appling 2010). 10-formyl THF is subsequently oxidized by the cytosolic 10-formyl THF dehydrogenase, also known as ALDH1L1, or its mitochondrial isoform ALDH1L2 to carbon dioxide. Both enzymes use NADP+ as a co-factor and regenerate THF (Skrzydlewska 2003; Krupenko 2009; Krupenko et al. 2010). Although formate oxidation takes place predominantly by the THF-dependent pathway, catalase-mediated oxidation of formate has also been reported (Cook et al. 2001; Skrzydlewska 2003).
Formaldehyde metabolism (Fig. 1) is best studied for the liver (Skrzydlewska 2003; Tibbetts and Appling 2010), but it is very likely that other organs including the brain will also use the enzymatic pathways that are well known for formaldehyde metabolism in liver. In brain, at least all the enzymes required for complete formaldehyde oxidation are expressed (Table 1).
Table 1. Formaldehyde-producing and formaldehyde-metabolizing enzymes in the brain
Differences in the rate of formaldehyde metabolism have been described between species for the formaldehyde metabolism. For example, formate is metabolized at a slower rate in the liver of monkeys and humans compared to rats, partly because rats have a higher hepatic THF content (Tephly 1991; Skrzydlewska 2003). Also, species-specific differences in the kinetic parameters of the enzymes involved in formaldehyde metabolism may contribute to the different rates of formaldehyde oxidation observed and subsequently may determine the consequences of an exposure to formaldehyde and/or it metabolites.
Generation and oxidation of formaldehyde in brain cells
Several reports have demonstrated that the enzymes required to produce or metabolize formaldehyde are expressed in the brain on the mRNA or protein level (Table 1). Of these enzymes, only the expression of ADH1 in the brain has been controversially discussed, since this dehydrogenase was not detected in brain by some investigators (Julia et al. 1987; Galter et al. 2003). Despite the presence of ADH1 mRNA in cultured neural cells, methanol generation was not found for formaldehyde-exposed cultured brain cells (Tulpule and Dringen 2012; Tulpule et al. 2013), suggesting that oxidation to formate is the preferred pathway of formaldehyde metabolism in brain cells. Cultured astrocytes and neurons contain the mRNAs for SSAO and LSD1 as well as for the enzymes involved in formaldehyde metabolism (Tulpule and Dringen 2012; Tulpule et al. 2013). These studies indicate that formaldehyde may be produced locally in the brain and that among the different types of brain cells at least astrocytes and neurons have the potential to generate and oxidize formaldehyde.
Acute formaldehyde exposure in concentrations of up to 1 mM for up to 3 h does not cause severe toxicity in cultured astrocytes or neurons (Song et al. 2010; Tulpule and Dringen 2011, 2012; Tulpule et al. 2013). A rapid metabolism of cellular formaldehyde may contribute to the resistance of cultured brain cells to formaldehyde toxicity, since formaldehyde has been reported to be more cytotoxic than its metabolites, methanol and formate (Oyama et al. 2002; Lee et al. 2008). Both, cultured astrocytes and neurons clear exogenously applied formaldehyde with a similar rate of around 0.2 μmol/(h × mg) (Tulpule and Dringen 2012; Tulpule et al. 2013) which is about 20% of the formaldehyde oxidation rate reported for liver cells (Dicker and Cederbaum 1984). The KM-value for formaldehyde clearance by cultured astrocytes is around 0.19 mM, suggesting that both the cytosolic ADH3 and mitochondrial ALDH2 could contribute to formaldehyde oxidation (Tulpule and Dringen 2012).
Although cultured astrocytes and neurons have comparable rates of formaldehyde clearance, the metabolic fate of the disposed formaldehyde differs between these two types of neural cells. Although astrocytes convert the majority (> 90%) of formaldehyde to formate that is subsequently exported from the cells (Tulpule and Dringen 2012), only about 25% of the formaldehyde cleared by cultured neurons is detected as extracellular formate (Tulpule et al. 2013). The underlying reason for this difference might be a poor export of formate from cultured neurons and/or a higher capacity of these cells to further oxidize formate to carbon dioxide (Fig. 1). Although the putative formate exporters, GABA-gated channels (Mason et al. 1990) and monocarboxylate transporter (MCT) 1 (Moschen et al. 2012) are expressed in both astrocytes and neurons (Debernardi et al. 2003; Olsen and Sieghart 2009; Lee et al. 2011; Velez-Fort et al. 2011), the expression level of MCT1 in neurons has been reported to be very low (Debernardi et al. 2003). However, if poor export of formate would be the only reason behind the lower extracellular accumulation of this metabolite in cultured neurons, these cells should accumulate large amounts of formaldehyde-derived formate, which is not the case (Tulpule et al. 2013). Thus, the lower extracellular accumulation of formaldehyde-derived formate in cultured neurons compared to cultured astrocytes is likely to be predominantly caused by oxidation of formaldehyde-derived cellular formate to carbon dioxide. The enzymes involved in the oxidation of 10-formyl THF require NADP+ as electron acceptor (Krupenko 2009; Krupenko et al. 2010), and the availability of NADP+ in cytosol and mitochondria depends on the pathways involved in NADPH consumption and NADPH regeneration. As such pathways differ between astrocytes and neurons (Dringen et al. 2007), the NADP+ availability could also contribute to the differences observed in formate release from astrocytes and neurons that were exposed to formaldehyde (Tulpule and Dringen 2012; Tulpule et al. 2013).
Alterations of the metabolism of brain cells upon exposure to formaldehyde
A large number of adverse consequences have been reported for an exposure of brain cells to formaldehyde in vivo and in vitro (Table 2). Recently, it was demonstrated that formaldehyde in the concentration range between 0.1 mM and 1 mM strongly affects basal metabolic properties of cultured astrocytes and neurons, that is, formaldehyde stimulates glycolytic flux and the export of the antioxidative tripeptide GSH from brain cells.
Table 2. Consequences of a formaldehyde exposure of rodent brain cells in vivo and in vitro
The articles by Lu et al. (2008), Usanmaz et al. (2002), and Tong et al. (2011) describe data that have been obtained on mice, whereas all other studies were performed on rats or rat brain cells.
Decrease in the number of neuron
Gurel et al. (2005); Aslan et al. (2006); Sarsilmaz et al. (2007)
Astrocytes are more glycolytic than neurons (Bolaños et al. 2010), a feature which has been attributed to expression of the glycolysis-promoting enzyme PFKFB3 in astrocytes (Herrero-Mendez et al. 2009), an inhibited pyruvate dehydrogenase complex (Halim et al. 2010) and a low rate of NADH shuttling into mitochondria in astrocytes (Berkich et al. 2007; Neves et al. 2012). Despite the differences in basal rates of glucose consumption and lactate release in cultured astrocytes and neurons, application of formaldehyde significantly increases these rates in both types of brain cells (Tulpule and Dringen 2012; Tulpule et al. 2013). However, the extent of stimulation of glycolytic flux in formaldehyde-exposed cells compared to the basal condition differs between the culture types investigated. For example, at a formaldehyde concentration of 0.5 mM, the lactate release and glucose consumption rates were doubled in cultured neurons (Tulpule et al. 2013), while this concentration of formaldehyde did not affect glycolysis in cultured astrocytes (Tulpule and Dringen 2012). Astrocytes had to be exposed to 1 mM formaldehyde to elevate glycolysis by 50% (Tulpule and Dringen 2012).
The accelerated glycolysis in formaldehyde-exposed neural cells is likely to be caused by the formaldehyde-derived formate which is known to inhibit mitochondrial cytochrome c oxidase (Nicholls 1975; Wallace et al. 1997). This view is supported by the observation that incubation of astrocytes with formaldehyde for 90 min is required for the accelerated lactate release to persist even after removal of formaldehyde (Tulpule and Dringen 2012). This long delay most likely reflects the slow mitochondrial accumulation of formaldehyde-derived formate to concentrations that are sufficient to inactivate respiration, as most of the formate is efficiently exported from astrocytes. Moreover, the persistent lactate release of astrocytes exposed to formaldehyde was not further enhanced by application of azide, an inhibitor of mitochondrial cytochrome c oxidase (Tulpule and Dringen 2012). Thus, formaldehyde-derived formate is likely to stimulate glycolytic flux as a consequence of an inhibited respiration, as also other inhibitors of respiratory chain complexes stimulate glycolytic lactate production in cultured astrocytes and neurons (Pauwels et al. 1985; Scheiber and Dringen 2011).
Formaldehyde-accelerated glutathione export
GSH is an important antioxidant (Lushchak 2012; Schmidt and Dringen 2012; Lu 2013) that is also involved in the formaldehyde oxidation catalyzed by ADH3 (Fig. 1). Under basal conditions, cultured astrocytes and neurons as well as cells of the oligodendroglial cell line OLN-93 export GSH, although with variable rates (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). Formaldehyde treatment stimulated GSH export from all three types of cultured neural cells without severely altering the ratio of GSH to glutathione disulfide (GSSG) (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). This accelerated GSH export from formaldehyde-treated neural cells is mediated by multidrug resistance protein (Mrp) 1 (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). Mrp1 is a member of ATP-binding cassette transporters and transports, besides GSH, a wide array of substrates including GSSG and GSH conjugates (Keppler 2011; Yin and Zhang 2011). The potential of formaldehyde to accelerate GSH export differs between different brain cell culture types. For example, exposure to 0.5 mM formaldehyde increased the respective GSH export rates of cultured astrocytes, neurons and OLN-93 cells by 10-, 5- and 20-fold, respectively (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). However, half-maximal cellular GSH depletions were observed at similar incubation parameters for all types of neural cells after incubation for 1 h with 0.3 mM formaldehyde (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). Formaldehyde exposure does not impair the capacity of neural cells to synthesize GSH. At least formaldehyde-treated neurons restored their cellular GSH levels after application of amino acid precursors for GSH synthesis (Tulpule et al. 2013).
The molecular mechanism involved in the formaldehyde-accelerated Mrp1-mediated GSH export from neural cells is not resolved so far. Since the stimulation of GSH export is observed within minutes after formaldehyde application (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013), de novo synthesis of Mrp1 is unlikely to explain the stimulated GSH efflux. Furthermore, the finding that removal of formaldehyde instantly decelerates the stimulated GSH export (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013) indicates that the mechanism responsible for formaldehyde-accelerated GSH export is quickly reversible. Assuming that cellular GSH is the transported Mrp1 substrate (Fig. 2a), formaldehyde could stimulate GSH export by a reversible, covalent activation of this transporter. Alternatively, a formaldehyde-induced recruitment of intracellular Mrp1 molecules into the cell membrane could explain the accelerated GSH export. Such a reversible translocation of Mrp1 from the Golgi to the cell surface has been reported for cultured astrocytes treated with bilirubin (Gennuso et al. 2004).
Mrp1 efficiently exports GSH conjugates (Keppler 2011; Yin and Zhang 2011). As the formaldehyde metabolism in neural cells involves the generation of the GSH conjugates S-hydroxymethyl GSH and S-formyl GSH (Fig. 1), these conjugates could also serve as substrates of Mrp1 (Fig. 2b). Since both conjugates are known to be labile (Ahmed and Ahmed 1978; Uotila 1981), they are likely to disintegrate into GSH and formaldehyde or formate immediately after being exported.
Direct experimental evidence that discriminates between the potential two mechanisms (Fig. 2) that may be involved in the formaldehyde-induced accelerated GSH export via Mrp1 is missing so far. However, determination of the kinetic parameters for the GSH export from astrocytes revealed that the KM-values of the basal as well as the formaldehyde-accelerated GSH export from astrocytes are identical (about 100 nmol/mg or 25 mM), but that the Vmax-value for the stimulated GSH export is eightfold higher than that for the basal GSH export (Tulpule et al. 2012). These data suggest that at least for formaldehyde-treated astrocytes GSH rather than a GSH conjugate is exported via Mrp1, since the KM-values of Mrp1 for its substrate GSH are normally higher than 5 mM, while that for GSH conjugates are below 1 mM (Burg et al. 2002; Cole and Deeley 2006; Deeley and Cole 2006).
Application of formaldehyde does not deprive the cells completely of their GSH and about 5% residual GSH still remains within neural cells (Tulpule and Dringen 2011; Tulpule et al. 2012, 2013). In cultured astrocytes, this low cellular GSH content represents a residual GSH concentration of about 0.4 mM (Dringen and Hamprecht 1998) which will be sufficient to drive ADH3-catalyzed GSH-dependent formaldehyde oxidation, since the KM-value of ADH3 for S-hydroxymethyl GSH is less than 10 μM (Casanova-Schmitz et al. 1984; Heck et al. 1990) and this reaction involves recycling of GSH (Fig. 1). Thus, the stimulated GSH export is unlikely to compromise GSH-dependent formaldehyde oxidation.
Evidence for the role of formaldehyde in pathology
In healthy individuals, the formaldehyde concentration in the blood has been reported to be around 0.1 mM (Heck and Casanova 2004) while that in the brain is about 0.2 mM (hippocampus) and 0.4 mM (cortex) (Tong et al. 2013a). These levels of formaldehyde represent the normal physiological balance between formaldehyde-generating and formaldehyde-disposing processes. However, an increased activity of formaldehyde-generating enzymes or an acute exposure to high amounts of exogenous formaldehyde without a concurrent elevation in the capacity to clear formaldehyde will raise formaldehyde level in the body and will lead to formaldehyde stress (He et al. 2010). Indeed, an increased expression/activity of the formaldehyde-generating enzymes VAP1/SSAO, LSD1 and JHDM has been reported for various diseases (Table 3). While a broad spectrum of pathological conditions are associated with elevated levels of VAP1/SSAO, an increase in the expression of the histone demethylases has especially been observed in different types of cancer (Table 3). The elevated expression of formaldehyde-generating enzymes is accompanied by increased formaldehyde levels in diabetic rats (Tong et al. 2013a), in cancer tissue (Tong et al. 2010) and in some human cancer cell lines (Kato et al. 2001; Tong et al. 2010).
Table 3. Elevation in expression or activity of formaldehyde-generating enzymes in human diseases
Ferrer et al. (2002); del Mar Hernandez et al. (2005); Unzeta et al. (2007)
Increased expression of formaldehyde-generating enzymes (Table 3) as well as elevated formaldehyde levels have also been reported in brains of patients suffering from neurodegenerative diseases like Alzheimer's disease (AD) or multiple sclerosis (MS) (Khokhlov et al. 1989 cited in Miao and He 2012; Tong et al. 2011, 2013a). Some hypotheses have been postulated that link the increase in formaldehyde level to neuropathology. For example, some human subjects who suffered from methanol poisoning developed symptoms of MS which has been discussed to be an effect of methanol oxidation to formaldehyde and the subsequent modification of proteins resulting in an immune reaction (Schwyzer and Henzi 1983; Henzi 1984). Along that line it was discussed that formaldehyde methylates proteins like tau (in AD) or myelin basic protein (in MS) which in turn elicits an immune response by the body that is characteristic for these diseases (Monte 2010; Lu et al. 2013). Also, inhibition of SSAO in a murine model of MS has been shown to reduce the incidence and severity of this disease (Wang et al. 2006) which could, at least partly, be the consequence of a lowered formaldehyde generation. Moreover, formaldehyde exposure has been implicated to be a risk factor for the development of amyotrophic lateral sclerosis (Weisskopf et al. 2009), a disease that is characterized by degeneration of motor neurons (Kiernan et al. 2011).
Formaldehyde-induced alterations in neural metabolism as potential contributors to neurodegeneration
Figure 3 summarizes the current knowledge on formaldehyde metabolism and on formaldehyde-induced alterations in the glucose and GSH metabolism of neural cells. The potential of cultured brain cells to efficiently metabolize formaldehyde suggests that also the cells in brain deal quite well with the moderate amounts of formaldehyde that are generated under physiological conditions. Similar to liver cells, brain cells are likely to use both cytosolic and mitochondrial pathways for formaldehyde oxidation to formate and further to carbon dioxide (Figs 1 and 3).
Cultured brain cells efficiently produce and export glycolytically generated lactate and also release GSH into the medium, although the basal rates of glycolysis and GSH export differ between different types of neural cells (Tulpule and Dringen 2011, 2012; Tulpule et al. 2012, 2013). These pathways are not affected by low concentrations of formaldehyde, but as soon as formaldehyde levels are increased in pathological conditions, an accelerated generation of formate is likely to stimulate glycolytic flux by inhibition of the mitochondrial respiration (Fig. 3). In addition, an excess of formaldehyde deprives brain cells of GSH by stimulating Mrp1-mediated GSH export (Fig. 3). Although, caution should be exercised while extrapolating in vitro data to the situation in the brain, a speculation on potential consequences of elevated formaldehyde levels in brain on the cellular metabolism is tempting, especially since the formaldehyde concentrations that have been shown to alter metabolic properties of cultured brain cells (0.1–1 mM) are in the concentration range reported for the normal brain (0.2–0.4 mM). Thus, mild elevations in brain formaldehyde concentrations could already strongly affect energy and GSH metabolism of this organ.
The potential pathological implications of metabolic changes exerted by excess of formaldehyde in the brain are shown in Fig. 4. Astrocytes and neurons in brain are likely to efficiently metabolize an excess of formaldehyde, as also reported for brain homogenates (Iborra et al. 1992). Subsequently, the formate generated from formaldehyde is either released from brain cells or inactivates mitochondrial cytochrome c oxidase. An inhibition of the mitochondrial respiratory chain will stimulate glycolytic flux in the brain cells to, at least transiently, meet their energy demand. However, prolonged exposure to formaldehyde is likely to result in energy crisis that in turn will disrupt the functions of brain cells. This may also be the underlying mechanism of the neurotoxicity of formate in hippocampal brain slices (Kapur et al. 2007). Besides this impairment of energy metabolism, formaldehyde-induced accumulation of both formate and lactate in the brain would cause cerebral acidosis (Skrzydlewska 2003; Rose 2010) which would subsequently induce astrocytic swelling, impairment of neuronal signal transmission and neurological deficits (Staub et al. 1993; Li et al. 2011; Zhao et al. 2011).
Exposure to high levels of formaldehyde will cause GSH depletion in brain cells together with GSH accumulation in the extracellular space. As GSH is involved in important cellular functions in the brain like protection against reactive oxygen species and detoxification of xenobiotics (Lushchak 2012; Schmidt and Dringen 2012; Lu 2013), GSH depletion may contribute to the severe oxidative stress reported for brain after prolonged exposure to formaldehyde (Zararsiz et al. 2006, 2007, 2011; Songur et al. 2008). A loss in cellular GSH would under normal conditions be compensated by increased GSH synthesis. However, lactacidosis caused by the formaldehyde-induced production of lactate (Skrzydlewska 2003; Rose 2010) impairs GSH synthesis (Lewerenz et al. 2010) and cellular GSH levels are likely to remain low. Thus, chronic exposure to formaldehyde may render brain cells incapable of fully restoring their cellular GSH levels.
The formaldehyde-induced accumulation of extracellular GSH in brain can also be detrimental, since GSH has been suggested to act as a neurotransmitter and neuromodulator at glutamate receptors (Janáky et al. 2007) which play important roles in memory and learning (Davis et al. 2013; Mukherjee and Manahan-Vaughan 2013). Also, accelerated extracellular GSH hydrolysis by the astrocytic ectoenzyme γ-GT (Dringen et al. 1997) caused by the increased extracellular GSH concentration would generate the neurotransmitter glutamate (Fernandez-Fernandez et al. 2012; Schmidt and Dringen 2012). Thus, excessive accumulation of extracellular GSH as well as GSH-derived glutamate may cause excitotoxicity which has at least been demonstrated in vitro (Regan and Guo 1999a, b).
To address the molecular mechanisms that are involved in the development of adverse neural effects of an elevated concentration of formaldehyde, it has to be discriminated between direct and indirect consequences of formaldehyde exposure. Acute exposure of neural cells to formaldehyde and/or the rapid generation of formaldehyde-derived metabolites will directly affect basal metabolic parameters (Fig. 4, light gray squares), which may subsequently lead to indirect, delayed consequences (Fig. 4, dark gray squares). Little is known so far on the mechanisms that link acute direct consequences of a formaldehyde exposure, such as accelerated glycolysis or GSH export, to the known adverse effects of formaldehyde on neural cells (Table 2). Activation of signaling cascades as well as alterations in protein expression are likely to be involved in the development of the delayed indirect effects of an exposure to excess of formaldehyde. For example, formaldehyde-exposed neuronal PC12 cells show endoplasmic reticulum stress, decreased levels of the antioxidant proteins thioredoxin and paraoxonase 1 (Tang et al. 2011; Luo et al. 2012) and a decreased expression of the anti-apoptotic protein Bcl-2, while the expression of pro-apoptotic Bax protein increases (Tang et al. 2012). Also, the expression of the rate-limiting enzyme in dopamine synthesis, tyrosine hydroxylase, is lowered in PC12 cells after exposure to formaldehyde (Lee et al. 2008). Further studies are now required to investigate the signaling pathways that link the acute formaldehyde-induced metabolic alterations to the known brain pathology of an excess of formaldehyde (Table 2)
Conditions such as aging and diseases like MS and AD which are associated with increased levels of formaldehyde in brain (Khokhlov et al. 1989 cited in Miao and He 2012; Tong et al. 2011, 2013a, b) show impaired mitochondrial function (Sullivan and Brown 2005; Mahad et al. 2008; Boumezbeur et al. 2010; Leuner et al. 2012) together with an increase in brain lactate content (Parnetti et al. 2000; Ross et al. 2010; Paling et al. 2011). Moreover, ageing, MS and AD have been connected with oxidative stress in the brain (Haider et al. 2011; van Horssen et al. 2011; Belkacemi and Ramassamy 2012; Sohal and Orr 2012; Steele and Robinson 2012). These reports strengthen the view that formaldehyde may, at least to some extent, have a role in the initiation and/or progression of pathological symptoms of neurodegenerative conditions (Yu 2001; Monte 2010). An adequate supply of lactate to neurons has been shown to foster memory formation (Suzuki et al. 2011), while GSH depletion in the brain has been demonstrated to result in behavioral changes (Steullet et al. 2010). Thus, the formaldehyde-induced alterations in glucose and GSH metabolism may contribute to the deficits in behavior, cognition and learning observed in formaldehyde-exposed animals (Pitten et al. 2000; Malek et al. 2003; Lu et al. 2008; Tong et al. 2011, 2013a, b)
Conclusions and future perspectives
In conclusion, elevation of brain formaldehyde levels is likely to alter brain cell metabolism which may affect the function of this vital organ. Although some studies have correlated that neurodegenerative conditions are associated with increased levels of formaldehyde in the brain and others have connected such diseases with impaired energy metabolism and oxidative stress, a direct causal link between formaldehyde, impaired metabolism and oxidative stress remains to be demonstrated. Interestingly, resveratrol which is known to be neuroprotective for AD (Richard et al. 2011; Li et al. 2012) is a formaldehyde scavenger (Tyihák and Király-Véghely 2008), suggesting that the beneficial effects of resveratrol could also include removal of excess formaldehyde. Further studies that will combine the quantification of formaldehyde levels in post-mortem brains with metabolite profiles and analysis of oxidative stress markers are now required to provide further experimental evidence for a direct contribution of formaldehyde in the pathology of neurodegenerative disorders.
Conflict of interest
The authors have no conflict of interest to declare.