Metallothioneins are multipurpose neuroprotectants during brain pathology


  • Milena Penkowa

    1. Section of Neuroprotection, Centre of Inflammation and Metabolism at The Faculty of Health Sciences, University of Copenhagen, Denmark
    Search for more papers by this author

M. Penkowa, Section of Neuroprotection, The Faculty of Health Sciences, University of Copenhagen, Blegdamsvej 3, DK-2200, Copenhagen, Denmark
Fax: +45 3 5327217
Tel: +45 3 5327222


Metallothioneins (MTs) constitute a family of cysteine-rich metalloproteins involved in cytoprotection during pathology. In mammals there are four isoforms (MT-I − IV), of which MT-I and -II (MT-I + II) are the best characterized MT proteins in the brain. Accumulating studies have demonstrated MT-I + II as multipurpose factors important for host defense responses, immunoregulation, cell survival and brain repair. This review will focus on expression and roles of MT-I + II in the disordered brain. Initially, studies of genetically modified mice with MT-I + II deficiency or endogenous MT-I overexpression demonstrated the importance of MT-I + II for coping with brain pathology. In addition, exogenous MT-I or MT-II injected intraperitoneally is able to promote similar effects as those of endogenous MT-I + II, which indicates that MT-I + II have both extra- and intracellular actions. In injured brain, MT-I + II inhibit macrophages, T lymphocytes and their formation of interleukins, tumor necrosis factor-α, matrix metalloproteinases, and reactive oxygen species. In addition, MT-I + II enhance cell cycle progression, mitosis and cell survival, while neuronal apoptosis is inhibited. The precise mechanisms downstream of MT-I + II have not been fully established, but convincing data show that MT-I + II are essential for coping with neuropathology and for brain recovery. As MT-I and/or MT-II compounds are well tolerated, they may provide a potential therapy for a range of brain disorders.


Alzheimer's disease


amyotrophic lateral sclerosis




antioxidant response element


brain-derived neurotrophic factor


experimental autoimmune encephalomyelitis


fibroblast growth factor




glial-derived neurotrophic factor



IL-6KO mice

IL-6 knockout mice (genetic IL-6-deficient mice)


macrophage colony-stimulating factor


matrix metalloproteinase


metal response elements 


multiple sclerosis




MRE-binding transcription factor-1

MT-KO mice

MT-I + II knock-out mice (genetic MT-I + II deficiency)


metallothionein III/growth inhibitory factor


nuclear factor kappa-B (transcription factor)


nerve growth factor




Parkinson's disease


reactive oxygen species


superoxide dismutase


transforming growth factor-β


TGF-β receptor

TgMT mice

mice with transgenic MT-I overexpression


tumor necrosis factor-α


vascular endothelial growth factor

Mammalian metallothioneins (MTs) constitute a superfamily of nonenzymatic polypeptides (61–68 amino acids), which are characterized by low molecular weight (6–7 kDa), distinctive amino acid composition (high cysteine content and no or low histidine) and sequence (unique cysteine distribution as Cys-X-Cys), and a high content of sulfur and metals (metal thiolate clusters) [1–3]. In vivo, the metal-binding involves mainly Zn(II), Cu(I), Cd(II), and Hg(II), while in vitro additional and diverse metals such as Ag(I), Au(I), Bi(III), Co(II), Fe(II), Pb(II), Pt(II), and Tc(IV) may be bound to apothionein (the metal-free form) [4,5]. However, during physiological conditions mammalian MTs mostly contain zinc [6,7].

In mammals, four major subfamilies exist (MT-I, MT-II, MT-III and MT-IV), of which MT-I and -II (MT-I + II) were discovered in 1957 and are the best described MT proteins. The roles of mammalian MT-I + II in the brain have received mounting scientific interest [1,8–10] and are also the focus of this review, which will not address other MT isoforms.

MT-I + II are expressed ubiquitously in mammalian tissues, which rapidly increase their mRNA and proteins in response to pathology or administration [1,10]. In rodents, MT-I + II are regulated and produced coordinately [2], and they are often described together as one functional entity [4,5]. In mammals, MT-I + II consist of 61 and 62 amino acids, respectively, which are devoid of aromatic amino acids, while one-third of the residues are cysteines (in total 20) that form metal thiolate clusters. In the polypeptide chain, cysteines are arranged in series of motifs: Cys-X-Cys, Cys-X-Cys-Cys, C-X-X-C (X is a non-Cys residue), which are absolutely conserved across species [3–6]. The cysteine sulfhydryl groups bind and coordinate 7 moles of divalent metal ions [i.e. Zn(II) or Cd(II)] per mol MT-I + II, while the molar ratio for Cu(I) and Ag(I) is 12.

The metal thiolate clusters (Scys-M-Scys) exist in two separate globular domains, the α- and β-domains, which are linked by a small, lysine-rich region, although the domains have few contacts [1,3]. The α-domain in the C-terminus (amino acid residues 33–61 in rat MT-II) contains 11 cysteines and is able to bind four divalent or six monovalent metals, while the N-terminal β-domain (amino acid residues 1–29 in rat MT-II) includes nine cysteines capable of binding three divalent or six monovalent metals [1,4,6] (Fig. 1). These residues are either bridging cysteines, which can bind two divalent ions or they are terminal cysteines that bind only one divalent metal [3,4,11].

Figure 1.

Schematic drawing of the mammalian MT-II protein showing the two metal-thiolate clusters (C-terminal a-domain and N-terminal b-domain) including the 20 cysteine residues (blue squares) and their sulfur atoms (S), which bind to divalent or monovalent cations (in this case Zn). The domains are linked by a short peptide containing amino acid residues 30–32 in mammalian MT-II (LINK). In the b-domain, three divalent or six monovalent metal ions are coordinated, while in the a-domain four divalent or six monovalent cations can be bound. Both bridging and terminal cysteines are present in mammalian MT-2. The bridging cysteines bind to two separate, divalent cations, while the terminal cysteines chelate one divalent metal. If monovalent metals are bound, all cysteines can chelate two cations.

When metal ions bind to apothionein, the polypeptide chain will rapidly fold resulting in the formation of the two native, three-dimensional metal thiolate clusters residing in each domain [3,5]. In the α-domain, the only known MT secondary structure can be found (a short α-helix present in case the protein is fully loaded with divalent (not monovalent) metals) [5,6].

The antigenic part (epitope) of the MT-I + II proteins is formed by a lysine-rich region, residues 20–25, together with the seven N-terminal residues 1–7, which after protein folding are seen in close proximity in the three-dimensional structure [1,4,6].

The most studied human MT genes are found on chromosome 16, which features very high levels of segmentally duplicated sequence among the human autosomes and abundant genetic polymorphisms, which are also existing in the MT-I + II genes [1,5].

In the chromosome 16 q13 region, MT genes are tightly linked, and as a minimum they consist of 11 MT-I genes (MT-I-A, -B, -E, -F, -G, -H, -I, -J, -K, -L, and -X) encoding functional or nonfunctional RNA, and one gene for the other MT isoforms (the MT-2 A gene, MT-3 gene, and MT-4 gene) [1,3,4]. However, a gene called MT-like 5 (MTL-5) has been described in the q13 region of chromosome 11, and it encodes a testis-specific MT-like protein named tesmin [3,4,6,7,12].

Compared with the human genes, the mouse MT genes are less complex, as they only have one functional gene for each major MT isoform (one gene encoding MT-1, MT-2, MT-3 and MT-4) and these are all located on chromosome 8. As in humans, the mouse genome also contains an MTL-5 gene, which is located on chromosome 19B [1–3,7,12].

However, this review will focus only on mammalian MT-I + II isoforms, while all the other MT isoforms and related MT-like structures (genes or their products) will not receive further attention. The major topics reviewed here are the in vivo roles of mammalian MT-I + II in immunoregulation, neuroprotection and cerebral regeneration, a field receiving growing scientific interest.

Cerebral MT-I + II expression

Brain MT-I + II mRNA and proteins are present in low amounts in physiological conditions and are expressed during embryonic development and in neonatals, and with increasing postnatal age MT-I + II immunoreactivity increases and becomes continually more widespread in the CNS [8,9]. In the brain, astrocytes are the main source of MT-I + II, although other cell types, such as choroid plexus epithelia, endothelium and meningeal cells, may also show MT-I + II [1,10]. In neurons the data on MT-I + II expression have been inconsistent, and MT-I + II positive neurons have only been intermittently described [9,13], although MT-I + II were demonstrated to exert direct protective effects upon neurons, as shown in primary neuronal cultures [14,15]. However, it is in general agreed that the levels of MT-I + II are several-fold higher in astrocytes relative to neurons.

Thus far, all the brain disorders studied in animals and humans have shown that MT-I + II mRNA and proteins are acutely and highly increased in reactive astroglia as part of the acute inflammation and host defense response [16–20]. To some extent, MT-I + II are also increased in the vascular endothelium, choroid plexus, ependyma, activated microglia/macrophages, and meninges, while neuronal and oligodendroglial MT-I + II immunoreactivity have not been consistently reported [21–23].

MT-I + II mRNA increases are seen within 24 h after an insult to the brain followed by many fold increases in their protein levels as seen typically after 1–3 days postinjury [20–24].

Increased MT-I + II expression is seen in various types of CNS pathology models such as in traumatic, excitotoxic, and ischemic/hypoxic injury, multiple sclerosis including its animal model experimental autoimmune encephalomyelitis (EAE), amyotrophic lateral sclerosis (ALS), Alzheimer's disease (AD), Parkinson's disease (PD), Pick's disease, pellagra encephalopathy, immobilization stress, and peripheral nerve injury [1,19,21,25–29].

Cellular MT-I + II distribution

MT-I + II have been considered as strictly intracellular proteins [4,29], but in recent years, mounting data indicated that MT-I + II are distributed both intra- and extracellularly [20,30–32].

Inside cells, MT-I + II are distributed in cytoplasm and subcellular organelles like lysosomes and mitochondria. Depending on the cell cycle phase, differentiation or in case of toxicity, MT-I + II are rapidly translocated to the nucleus, as seen during early S-phase and in oxidative stress [4,33–35]. Due to their small size, MT-I + II can diffuse through nuclear pore complexes, although the nuclear trafficking is relying on specific cytosolic partner proteins and the appearance of nuclear binding proteins, which in the presence of ROS enhance the nuclear localization of MT-I + II [29,36,37]. Also, perinuclear localization of MT mRNA may contribute to the nuclear import of MT-I + II proteins, as well as some structural alterations in the proteins per se (such as lack of post-translational acetylation of lysine and cysteine) are anticipated to regulate the nuclear trafficking [29,36].

Once in the nucleus, MT-I + II are selectively and actively retained by nuclear factors, which are likely to make use of saturable and energy-dependent binding mechanisms, in that elimination of the ATP pool hampers the nuclear translocation and/or retention of MT-I + II [37]. However, the precise intracellular MT-I + II trafficking system has yet to be clarified.

In addition, cells have been demonstrated to actively secrete MT-I + II in vitro, although there is no known signal peptide for cellular export [35,38].

In vivo, MT-I overexpressing transgenic mice display significant MT-I + II immunoreactivity in the brain extracellular space [20]. In the brain, the astrocytes, not the neurons, are the major source of MT-I + II, even though these proteins primarily protect the neurons [9,31,39]. Hence, it is considered that astroglia may secrete MT-I + II to the extracellular space in order for them to protect the surrounding neurons [30]. This is supported by studies of primary cell cultures, which showed that extracellular MT-I + II exert direct effects upon neurons, as MT-administration enhanced the survival, differentiation and postinjury recovery of cortical, hippocampal, and dopaminergic neurons [14,15].

The experimental data from in vitro and in vivo studies that are reviewed here have shown consistently that intra- and extracellular MT-I + II promote analogous functions [14,28,29,31,35,40–41], which specifies that MT-I + II have roles both in and outside cells.

Regulation of MT-I + II

MT-I + II are regulated in a coordinate manner [2] and are rapidly increased by various pathological conditions [1,20]. However, physiological and lifestyle-related parameters like nutritional condition and physical activity have also been reported to regulate MT-I + II mRNA and proteins [7,11].

Administration of essential or toxic metals like Zn, Cu, Cd, Hg increase MT-I + II biosynthesis by inducing their transcription, for which several cis-acting DNA elements, metal response elements (MREs) in the promoter region are binding sites for trans-acting transcription factors [3,7,43,44]. The MT-I + II gene transcription is initiated when metals occupy the MRE-binding transcription factor-1 (MTF-1), which is a multiple Zn finger protein and the only known mediator of the metal responsiveness of MT-I + II [3,44].

Reactive oxygen species (ROS) and oxidative stress also increase expression of MT-I + II, which are highly efficient free radical scavengers in the brain [1,45]. ROS increase the MT-I + II transcriptional response, as shown by exposure to free radicals like superoxide anions and hydroxyl radicals, which rapidly increase MT-I mRNA levels in a dose-dependent manner [43,46]. The mechanism involves an antioxidant response element (ARE) in the promoter region, ARE-binding transcription factors, as well as the MTF-1, transcription factors of the basic zipper type (Fos and Fra-1), NF-E2-related factor 2, and the upstream stimulatory factor family (USF, a basic helix–loop–helix–leucine zipper protein), although it is likely that other and yet unidentified proteins are involved [7,46]. Thereby, metals and ROS activate MT-I + II gene transcription by different signaling pathways, response elements and transcription factors.

In addition, MT-I + II are also increased by glucocorticoid hormones like corticosterone and dexamethasone, which signal through glucocorticoid response elements (GREs) present in the gene regulatory region, and also catecholamines (norepinephrine, isoproterenol) activate MT-1 + II gene transcription [1,7,47].

During CNS inflammation, major MT-I + II regulatory factors are proinflammatory cytokines and especially interleukin (IL)-6 [1]. Accordingly, IL-6, IL-3, tumor necrosis factor (TNF)-α, macrophage-colony stimulating factor (M-CSF), and interferons increase brain MT-I + II expression in a cytokine-specific manner as demonstrated by using transgenic mice with cytokine overexpression [48–51] or cytokine deficiency [29,52–55].

Although the activation of MT-I + II gene transcription is by far the best described regulatory mechanism, repression of MT gene activity has also been reported [4,7]. Hence, during Zn deficiency, MTF-1 may form a complex with a Zn-responsive inhibitor, named MT transcription inhibitor, which prevents MTF-1 from interacting with the MREs, and thereby MT-I + II gene transcription could be negatively controlled due to the levels of trace metals [4,7,43]. In human cells, MT-IIA gene activation is inhibited by Zn finger protein PZ120, which interacts with the MT-IIA transcription start site and inhibits gene expression [56]. Also, transcription factors Fos and Fra-1 can inhibit MT-I + II biosynthesis by interaction with ARE [7].

However, MT-I + II biosynthesis is also affected by post-transcriptional events, since their protein levels do not necessarily reflect the levels of mRNA expression [4,57]. Hence, Cu treatment of adult rats reduced renal MT-I + II mRNA levels while at the same time, the renal MT-I + II protein expression was significantly increased [58], which suggests that post-transcriptional regulation occurs and this may likely affect either the translation and/or the protein degradation. In fact, MT-I + II are to some degree regulated by means of intracellular protein degradation, which takes place in both lysosomal and nonlysosomal compartments [4,59]. In general, intracellular MT-I + II proteins occur as either metal-containing proteins (MTs) or as apothioneins, and their depletion and/or restitution may depend on the bound metals, subcellular localization, and the tissue examined. Hence, turnover rates of cytosolic apothioneins versus lysosomal metal-bound MT-I + II proteins are quite different, in that lysosomal MT-I + II proteolysis occurs more readily than in the cytosol, although bound metals stabilize MT-I + II proteins and prevent their lysosomal proteolysis [59,60]. In the cytoplasm, the 26S proteasome complex degrades apothionein, which due to the lack of metals has a shorter half-life than MTs [4,11]. The type of metal complex may also in itself affect the MT-I + II degradation, as the half-life of Cd-containing proteins is close to 3 days, while Zn-binding reduces half-life to 18–20 h. Also, animal age and the chemical pretreatment may determine the half-life of MT-I + II, as well as MT-I in some cases has reduced half-life relative to MT-2 [61]. Thus, it is evident that other factors than the metals per se can regulate MT-I + II turnover [4,11,61].

The CNS roles of endogenous MT-I + II

Recently, rising interest in MT-I + II neuroprotective functions and therapeutic potential has been evident. During the genomic era it became possible to modify the MT-I + II genes in cultured cells and in animal embryos leading to the generation of MT-I + II knockout (MT-KO) mice [62] and transgenic MT-I overexpressing (TgMT) mice [63]. These genotypes have provided important answers concerning the roles of MT-I + II in the disordered CNS, although at first the data from MT-KO mice were rather disappointing, since these mice developed normally, appeared viable and fertile without any phenotypic changes [62]. Consequently, MT-I + II were considered as dispensable factors and/or proteins that may have abundant compensatory backup systems.

Years later, this concept was substantially contradicted, as it was demonstrated that neuronal survival and brain tissue repair are compromised when MT-I + II are absent. It became clear that during brain disorders, MT-KO mice show significantly enhanced brain tissue destruction, neuronal cell death, and clinical symptoms, when compared with those of wild-type controls [49,64–66]. Accordingly, even if the MT-I + II proteins may be dispensable during healthy, physiological conditions, they are unquestionably essential for coping with brain damage [1,30,39]. The major histopathological changes seen in brains of MT-KO mice are enhanced inflammatory responses including increased recruitment of macrophages, lymphocytes and their CD34 + hematogenous progenitor cells and enhanced secretion of proinflammatory factors like IL-1, IL-3, IL-6, IL-12, TNF-α, lymphotoxin-α (LTα), macrophage activator factor (Mac-1), intercellular adhesion molecule (ICAM-1) and acute phase response gene EB22 [31,49,65–69].

These studies also gave insight into the MT-I + II in vivo antioxidant functions in the brain, as MT-I + II deficiency resulted in amplified ROS formation and oxidative stress including highly increased lipid peroxidation, protein nitrosylation and DNA oxidation, when compared with those of WT controls [19,21,67,70]. In addition, MT-KO mice display significantly increased neurodegeneration and apoptotic cell death relative to WT controls as shown during traumatic brain injury, kainic acid-induced epileptic seizures, 6-aminonicotinamide (6-AN)-induced pellagra encephalopathy, ischemia, cytokine-induced meningoencephalitis, peripheral nerve injury, PD, EAE and ALS [23,27,40,62,65,67–71]. During these brain disorders, the MT-KO mice also developed worse clinical symptoms and showed significantly poorer neurological outcome relative to WT controls (Fig. 2).

Figure 2.

Schematic drawing of the main anti-inflammatory, antioxidant and anti-apoptotic actions of MT-I + II leading to neuroregeneration, angiogenesis and repair. MT-I + II modulate an array of vital cellular functions that involve cytoprotection, angiogenesis, DNA repair and the maintenance of tissue homeostasis. During pathology, MT-I + II inhibit inflammation and cytokines and protect against oxidative stress, degeneration, and apoptosis.

In contrast to brain disordered MT-KO mice, the TgMT mice showed significantly less neuropathological damage, while their tissue repair and neurological outcome were improved relative to WT control mice [13,20,28,31,40,55,72,73].

Thus, TgMT mice subjected to diverse brain disorders display reduced inflammatory responses of macrophages and lymphocytes including significantly decreased levels of proinflammatory cytokines, matrix metalloproteinases (MMPs), and ROS. Also, the amounts of delayed brain tissue damage consisting of oxidative stress, neurodegeneration and apoptotic cell death were radically reduced in TgMT mice relative to wild-type controls [13,20,28,31,55,72]. To this end, comparisons of the MT-I + II containing cells in the brain with the cell populations suffering from oxidative stress and apoptotic death showed clearly that damaged and/or dying cells are devoid of MT-I + II expression, which are confined to surviving cells, and this likely reflects the cytoprotection conferred by MT-I + II [19,72].

In addition, MT-I overexpression after brain injury stimulates the astroglial responses including the expression of anti-inflammatory cytokines, growth factors, neurotrophins and their receptors, such as IL-10, fibroblast growth factor (FGF), FGF-receptor (FGF-R), transforming growth factor (TGF)-β, TGF-β-receptor (TGF-β-R), vascular endothelial growth factor (VEGF), nerve growth factor (NGF), neurotrophin (NT)-3–5, brain-derived neurotrophic factor (BDNF), glial-derived neurotrophic factor (GDNF) [13,20,28,31,72,73]. Concomitantly, MT-I overexpression improves brain tissue repair including neuronal regrowth and vascular remodeling by angiogenesis, as well as the TgMT mice show improved clinical outcome, when compared with those of the wild-type control mice [13,15,28,31,72].

To study brain restoration and tissue repair, a suitable model is the traumatic, focal brain injury to the cortex, which results in a cortical necrotic cavity without viable cells that gradually will be replaced with glial scar tissue, vascular network and extracellular matrix [10,24,54,74].

These processes are significantly enhanced by MT-I + II, which are essential for the CNS wound repair to occur [30,31,72,64]. Hence, in the injured MT-KO mice the lesion cavity persists after 3 months, by which severe inflammation is ongoing; while in wild-type controls, the necrotic cavity is usually replaced with a scar after 30 days [65,66], while in MT-Tg mice this scar tissue is established before day 20 postinjury [31,72].

Moreover, following brain pathology MT-I + II are essential for the recruitment of neuroglial precursor cells [19,22,75] and their migration towards the site of injury. Hence, the increased tissue repair promoted by MT-I + II after injury is likely mediated in part by regeneration, where newly formed cells repopulate the tissue and in part by regrowth and sprouting of surviving cells.

The roles of exogenous MT-I + II

Shortly after the first data emerged from genetically modified MT-KO and TgMT mice subjected to neuropathology, new studies were conducted focusing on the potential therapeutic use of MT-I + II proteins. For this, adult rodents were injected intraperitoneally with exogenous MT-I and/or MT-II (MT-I/II) proteins during healthy conditions and neuropathological disorders like brain injury, pellagra encephalopathy and EAE [9,22,28,30,31,75]. The used MT-I/II proteins contained metals, which were mainly Zn (the Zn content was approximately 7%) and small amounts of Cd (the Cd content was < 0.5%). Therefore, these metals were included in the control treatment regimen.

The intraperitoneal administration of exogenous MT-I/II modulates immunoregulation and improves neuroprotection and CNS recovery in vivo during brain pathology, reflecting that MT-I + II have extracellular roles. This was far from anticipated at the time of the first publication (2000), as MT-I + II had been considered as strictly intracellular proteins [4].

At first, exogenous MT-I and/or MT-II proteins were injected intraperitoneal in rats with EAE that were evaluated clinically and histopathologically. In a dose- and time-dependent manner, the MT-I/II treatment reduced the severity of neurological symptoms and the mortality relative to placebo control groups.

MT-I/II treatment in EAE reduced significantly the activation and recruitment of macrophages and T lymphocytes including levels of IL-1β, IL-6, IL-12, TNF-α and ROS, which was seen in brain, spleen, and bone marrow [9,42]. The EAE lesions (plaques) with demyelination, apoptotic cell death, axonal degeneration and transection were radically reduced by MT-I/II administration relative to control treatment [22]. Concomitantly, MT-I/II-treated animals displayed improved remyelination, regeneration and clinical recovery from EAE relative to the placebo groups. This therapeutic effect was due to MT-I + II-activation of oligodendroglial progenitors/stem cells and enhanced expression of growth and trophic factors (FGF, TGF-β, NT-3–5 and NGF), which were significantly enhanced by MT-I + II in EAE and even more during the recovery phases, when compared with those of the placebo controls [22,76].

In later studies, exogenous MT-I/II proteins were administered during experimental models of traumatic brain injury (freeze lesion with dry ice) and pellagra encephalopathy (induced by administration of 6-AN). The acute (primary) injuries (the trauma- or 6-AN-induced necrosis) were comparable in the treatment groups, but in the following days, some pronounced differences in the responses to pathology appeared. Thus, animals receiving MT-I/II-treatment showed significantly less oxidative stress, neurodegeneration and apoptotic cell death (delayed damage) in the days/weeks following the primary injuries [28–31,75]. In these studies, the MT-I/II treatment also enhanced repair responses including expression of growth/trophic factors, astrogliosis, angiogenesis, neuronal regrowth [75], and particularly after the traumatic brain injury, it was evident that MT-I/II enhance reorganization of the necrotic lesion cavity [31]. The metal bound state of MT was preferred because the metalloform is likely to be the more physiological relevant form of the protein, and also because it is significantly less susceptible to degradation than apothionein. However, none of the effects of the MT-I/II treatment were seen after administration of the metals per se, but the latter may still be important as MT-I/+II adopt their tertiary structure upon chelation of metal ions [1,3,4]. However, the molecular mechanisms by which the MT-I/II treatment promoted neuroprotection and repair remain to be fully clarified (Fig. 3).

Figure 3.

Summary of the major biological functions of MT-I + II.

The MT-I + II molecular mechanisms

To clarify the specific functions of the MT-I + II proteins, many different approaches and techniques have been applied throughout thousands of studies. Although they described the MT-I + II structure, chemical characteristics, regulation, expression, distribution, degradation and the consequences of reducing or increasing MT-I + II in cells; they have not yet clarified the precise signaling and mechanisms by which MT-I + II exert immunoregulatory and neuroprotective actions.

However, many possibilities are likely, since MT-I + II are indeed multipurpose proteins involved in a broad range of functions, which include, but are not restricted to metal ion homeostasis, scavenging of ROS, redox status, immune defense responses, protein–protein and protein–nucleotide interactions, regulation of Zn fingers and Zn-containing transcription factors, mitochondrial respiration, thermogenesis, body energy metabolism, angiogenesis, cell cycle progression, and cell survival and differentiation [1,6,29,30,33,39,77,78]. Some of these MT-I + II actions may have therapeutic relevance in a range of acute and chronic neurological disorders, in which inflammation and oxidative stress are central in the pathophysiology [79–84]. Accordingly, MT-I + II may signal through diverse molecular pathways.

The immunomodulatory actions of MT-I + II reduce proinflammatory mediators including cytokines, MMPs, and adhesion molecules [20,32,72].

The reduction of brain IL-1, IL-6, IL-12 and TNF-α could be a central mechanism in the MT-I + II anti-inflammatory effects, since these cytokines are major immune activators that increase leukocyte activation, transendothelial migration, and chemoattraction, thereby leading to neuroinflammatory infiltrates [79–81]. Hence, genetic deficiency or overexpression of these cytokines or their receptors will diminish or enhance the brain inflammatory leukocytes [74,81,84,85]. Thus, IL-6 knockout mice (genetic IL-6-deficient mice) (IL-6KO) mice are resistent to EAE sensibilization, while IL-6 overexpressors show spontaneous chronic neuroinflammation and degeneration [73,86], which reflects that IL-6 is activating hematopoiesis, acute phase responses, and inflammation. IL-1 and IL-12 are also central pro-inflammatory cytokines that are crucial in the development of Th1 cells and initiation of autoimmune attacks, demyelination and neurodegenerative diseases and neuronal cell death by apoptosis and necrosis [49,80,82,84,85]. As the pro-inflammatory cytokines mediate significant neurotoxicity and chronic pathology, the MT-I + II-inhibition of their mRNA and protein biosynthesis [66] will likely contribute to improved neuroprotection.

It was recently shown that MT-I + II share certain structural and functional similarities with beta- and delta-chemokines CCL-17 and CX3CL-1 in vitro[87], whereby MT-I + II may regulate leukocyte chemotaxis, although this has yet to be confirmed in vivo. Other cell culture studies showed that MT-I + II may inhibit monocytic activation and invasion including secretion of cytokines [88–90]. Moreover, MT-I + II inhibit macrophage-induced T cell proliferation and the activation of cytotoxic T cells and antigen-specific B cells [91–94].

To this end, MT-I + II may also reduce inflammation by interfering directly with cell–cell interactions as MT-I + II were demonstrated to bind specifically to the membranes of macrophages, T and B cells, which thereby are inactivated [91–95].

These MT-I + II anti-inflammatory effects can also be seen in humans, as patients with autoimmune and allergic diseases show depletion of systemical MT-I + II and occurrence of anti-MT-I + II autoIgGs against MT-I + II, an alteration that is most pronounced during clinical exacerbations [96,97]. However, as in animals, the human MT-I + II levels can be fully replenished by various agents, among which steroids like glucocorticoid and cortisone can be used; and interestingly, steroids in general increase MT-I + II levels right before the patients show significant clinical improvements [97,98]. Also in MS patients, the MT-I + II expression levels are highest during the recovery and remission of disease [76].

In fact, the molecular mechanism of steroid-mediated immuno suppression could be a steroid-caused MT-I + II augmentation, given that glucocorticoid-treated patients show significant MT-II increases in their peripheral leukocytes shortly before the therapeutic effect of steroid commenced [98]. In support of this, dexamethasone-induced MT-II can be used as an indicator of glucocorticoid sensitivity [99]. This correlation between steroids and MT-I + II also exists in the brain, where MT-I + II mRNA and proteins are enhanced significantly by glucocorticoids [100].

In case MT-I + II are central mechanisms of steroid therapeutic effects, then MT-I + II might be used as a more specific anti-inflammatory agent likely having less side-effects than steroids.

As proinflammatory cytokine profiles are associated with development of human type-2 diabetes, which also affects the brain, we recently examined MT-I + II in such patients. Interestingly, systemical MT-I + II expression and function are depleted in type-2 diabetics relative to healthy subjects [101]. Hence, both constitutive and stress-related MT-I + II were deficient in the patients versus the healthy control subjects, which suggests that an absence of MT-I + II may have a key role in the pathogenesis of type-2 diabetes [101]. Indeed, in a following study of experimental diabetes, it was shown that diabetic MT-I + II depletion can be fully restored by medication, and such MT-I + II replenishment is associated with disease remission [102].

In addition, the MT-I + II inhibition of MMPs, which are Zn-dependent endopeptidases produced by inflammatory cells, may also contribute to amelioration of a number of human autoimmune diseases, where MMPs are involved in pathophysiological events like diapedesis of infiltrating cells, tissue degradation and blood–brain barrier breakdown [103,104].

Furthermore, MT-I + II stimulate astroglial responses including expression of anti-inflammatory signals, growth/trophic factors [72,66,68]. Although astrogliotic scarring traditionally has been considered as inhibitors of neuroregeneration, mounting and convincing data have now shown that reactive astrocytes provide essential neuroprotection and recovery. Hence, astrocytes endow neurons with antioxidants, energy substrates, anti-inflammatory and trophic/growth factors; and they improve neurogenesis and neurological outcome [70,105,106].

Hence, ablation of astroglia during brain pathology leads to massive increases in neurodegeneration, demyelination, infiltration by leucocytes, and cell death [106]. Thus, astroglial responses activated by MT-I + II may contribute to increased neuron survival, regeneration and CNS recovery. Also, MT-I + II increase expression of IL-10, FGF, TGF-β, VEGF, NGF, NT-3–5, BDNF, GDNF and their receptors; and this could in itself mediate neuroprotection as well as contribute to the MT-I + II-mediated repair, angiogenesis and vascular remodeling [19,20,22,30,32,39]. Together, these actions of MT-I + II can contribute to overall improvements in CNS cell survival and recovery [1,27,28,31,72,64–66,76].

Taken as a whole, these effects upon cerebral inflammation suggest that MT-I + II could be causing a general shift in the balance between pro- and anti-inflammatory molecules.

The actual mechanisms through which MT-I + II inhibit neurodegeneration and cell death remain to be fully described, although a range of anti-apoptotic effects have been shown in animals and humans. First, the anti-inflammatory effects as well as the antioxidant properties of MT-I + II could each contribute to decreased neurodegeneration and cell loss [79–83], although it is unlikely that these MT-I + II effects are the only responsible mechanisms.

Hence, when the MT-I + II anti-inflammatory actions are counterbalanced, as done by using double transgenic mice overexpressing both MT-I and pro-inflammatory cytokine IL-6, it was evident that MT-I + II still reduce neurodegeneration and cell death significantly [32,72,73]. However, the MT-I + II antioxidant effects likely contribute to neuroprotection, as the cerebral ROS formation and oxidative stress are inversely related to the MT-I + II levels but not to the expression of other antioxidants such as Cu/Zn-super oxide dismutase (Cu/Zn-SOD), Mn-SOD, and catalase [32,40,65,72]. During mitochondria-specific oxidative stress, MT-I + II are indispensable and have key roles in the mitochondrial protection, which did not relate to other antioxidants like glutathione peroxidase, catalase, Mn-SOD, and Cu/Zn-SOD [45].

MT-I + II may also prevent neuronal damage by having critical roles in metal ion homeostasis. Particularly the Zn regulation by MT-I + II may have major importance, since Zn is central for a broad range of functions. However, tight control of the Zn levels is necessary as an overload or deficiency of this metal leads to severe neurotoxicity [1,107]. Also, MT-I + II transfer Zn directly to mitochondrial factors, Zn-finger proteins and transcription factors, which also are essential for several signaling pathways and cell fate [4,34,78]. Along with Zn, various metal ions with a neurotoxic potential are bound and released by MT-I + II, which can thereby influence a range of cellular metabolites and pathways in the brain. A disrupted metal ion homeostasis causes oxidative stress, degeneration and neuronal cell death, and accordingly, dysregulation of metals has been associated with many pathologies including stroke, epilepsy, PD, AD, and traumatic brain injury [6,10,21,44,107]. Hence, the MT-I + II regulation of metal ion availability and levels in the CNS is most likely to contribute to the MT-I + II protective functions. Besides having roles in metal ion regulation, MT-I + II proteins also obtain their tertiary structure and enhanced molecular stability from their chelation of metals [4,6,11].

To this end, it is important that the different MT-I/II treatments injected into animals were all fully loaded Zn7–MT complexes, as the metal ensures protein stability, folding and longer half-life [3,4].

However, MT-I + II interact and modulate many intracellular messengers that are directly or indirectly regulating the apoptotic cascade, and therefore MT-I + II may affect additional pathways during their responses to damage and promotion of tissue repair. The nucleotides ATP and GTP [5,34,108] bind to MT-I + II proteins, whereby both structural and functional changes are seen in the proteins [108]. Also, the MT-I + II and ATP levels inside cells are interrelated, which in itself could affect cell loss or survival, since ATP depletion is part of the apoptotic cascade [40]. The MT-I + II and ATP connection may also be implicated in other actions, such as the MT-I + II-caused stabilization and rejuvenation of the ageing mitochondrial genome [40] and MT-I + II-regulation of energy balance and metabolism [77,78]. To this end, MT-I + II can donate Zn directly to mitochondrial aconitase (m-aconitase) by means of direct protein–protein interaction [109].

In addition, MT-I + II regulate the levels, activity and cellular localization of the transcription factor NFκB [10,70,95], which is involved in cell fate during neuropathology. Besides, MT-I + II induce a range of common proto-oncogenes (like bcl-2 and c-myc) whilst pro-apoptotic proteins (like p53 and caspase-3) are inhibited [11,20,32,33,41,74].

The roles of MT-I + II in cell fate and the MT-I + II connection to other factors involved in cell cycle regulation have led to many studies of MT-I + II roles in cancer. It is not surprising that MT-I + II may prevent tumor cell death by protecting against pro-apoptotic treatment regimes [11,33]. However, when the cancer is located in ectodermal tissues (such as colon, bladder and skin), a positive correlation exists between increased MT-I + II levels and an improved prognosis [11].

Final comments

This review summarizes the current knowledge and advances in the understanding of MT-I + II roles in immunomodulation and neuroprotection. The findings indicate that MT-I + II inhibit efficiently proinflammatory cytokines, ROS, MMPs and pro-apoptotic signals, which all may cause a broad range of brain disorders. As shown by many independent groups, the MT-I + II levels are inversely related to the degree of brain damage observed after traumatic injury, EAE, epileptic seizures, ischemia, and neurodegenerative diseases like PD, ALS and pellagra [9,15,20,22,27,30–32,40,69–71]. Consequently, MT-I + II might provide new drug targets against neurological disorders, especially those containing autoimmunity, neurodegeneration and neuron loss. As MT-I + II compounds are in general well tolerated, they may be used in the future as therapeutic and/or preventive medications.


These studies were supported by IMK Almene Fond, Vera og Carl Michaelsens Legat, The Lundbeck Foundation, The Danish Medical Research Council, The Danish Medical Association Research Fund, Toyota Fonden, Frænkels Mindefond, Scleroseforeningen, Kathrine og Vigo Skovgaards Fond, Fonden til Lægevidenskabens Fremme, Dir. Leo Nielsens Legat, Th. Maigaard's Eftf. Fru Lily Benthine Lunds Fond. Thanks are given to Adam Bohr and Kristian Kolind for excellent procedural assistance.