no plagiarism and ACS Reference. 

Mutation Research 475 (2001) 113–121

Review

Role of magnesium in genomic stability

Andrea Hartwig∗
Department of Food Chemistry and Toxicology, University of Karlsruhe,

Profach 6980, D-76128 Karlsruhe, Germany

Received 22 March 2000; received in revised form 29 June 2000; accepted 30 June 2000

Abstract

In cellular systems, magnesium is the second most abundant element and is involved in basically all metabolic pathways. At
physiologically relevant concentrations, magnesium itself is not genotoxic, but is highly required to maintain genomic stability.
Besides its stabilizing effect on DNA and chromatin structure, magnesium is an essential cofactor in almost all enzymatic
systems involved in DNA processing. Most obvious in studies on DNA replication, its function is not only charge-related, but
very specific with respect to the high fidelity of DNA synthesis. Furthermore, as essential cofactor in nucleotide excision repair,
base excision repair and mismatch repair magnesium is required for the removal of DNA damage generated by environmental
mutagens, endogenous processes, and DNA replication. Intracellular magnesium concentrations are highly regulated and
magnesium acts as an intracellular regulator of cell cycle control and apoptosis. As evident from animal experiments and
epidemiological studies, magnesium deficiency may decrease membrane integrity and membrane function and increase the
susceptibility to oxidative stress, cardiovascular heart diseases as well as accelerated aging. The relationship to tumor formation
is more complex; magnesium appears to be protective at early stages but promotes the growth of existing tumors. With respect
to the magnesium status in humans, the daily intake in most industrialized countries does not reach the current recommended
daily dietary allowances (RDA) values, and thus marginal magnesium deficiencies are very common. © 2001 Elsevier Science
B.V. All rights reserved.

Keywords:Genomic stability; Magnesium; Tumorigenesis; Dietary intake; Oxidative stress

1. Introduction

Magnesium is the second most abundant element in
cellular systems. It exerts a large variety of biological
functions, ranging from structural roles by complexing
negatively charged groups, i.e. phosphates in nucleic
acids, catalytic roles in enzyme activation or inhibi-
tion, and regulatory roles by modulating cell prolifer-

∗ Tel.: +49-721-608-2936; fax:+49-721-608-7254.
E-mail address:andrea.hartwig@chemie.uni-karlsruhe.de
(A. Hartwig).

ation, cell cycle progression and differentiation. Even
though less understood as compared to calcium home-
ostasis, the intracellular magnesium content appears
to be regulated by Mg2+ uptake, efflux, and intracel-
lular compartmentization, also in response to external
stimuli. With respect to genomic stability, several as-
pects are of major importance. They include the role of
magnesium in DNA replication and protein synthesis,
its function as cofactor in DNA repair proteins, its role
in maintaining the anti-oxidative status of the cell and
finally its effect on cell cycle regulation and apoptosis.
Magnesium deficiency or the displacement of Mg2+

0027-5107/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0027-5107(01)00074-4

114 A. Hartwig / Mutation Research 475 (2001) 113–121

by other toxic divalent metal ions leads to an increased
genomic instability, as evident by inhibited DNA
repair, oxidative stress, aging, and carcinogenicity.

1.1. Magnesium homeostasis and effects on cellular
functions related to genomic stability

1.1.1. Magnesium homeostasis at the cellular level
While much attention has been given to eluci-

date the homeostasis of Ca2+, H+, K+, or Na+, the
knowledge about the intracellular magnesium distri-
bution is still incomplete, mainly because of technical
limitations measuring Mg2+ inside cells. The total
magnesium content in various cell types ranges from
5 to 30 mM, while the free Mg2+ content has been
determined mostly between 0.4 and 0.6 mM. Magne-
sium is transported across the plasma membrane in
both directions, most likely via Na+/Mg+ exchange.
Whether influx and efflux of Mg2+ are mediated by
the same mechanism operating in opposite directions
or whether there are distinct pathways is still not
known. Both influx and efflux are hormonally con-
trolled and regulated by modifications in intracellular
cAMP levels and in protein kinase C activity. Con-
cerning the intracellular distribution of magnesium,
major intracellular compartments are mitochondria,
the nucleus, and the endoplasmatic reticulum (for
reviews see [1–3]). As discussed in more detail be-
low, bound magnesium can be mobilized from these
compartments to increase the concentration of free
intracellular Mg2+ if required, for example for cell cy-
cle progression or apoptosis. Thus, the distribution of
magnesium within cells is highly regulated, and many
enzymes in different biochemical pathways are acti-
vated or inhibited by changes in free Mg2+, some of
which are relevant for maintaining genomic stability.

1.2. Effect of magnesium on the fidelity of DNA
replication

One important prerequisite for maintaining the
genomic integrity is the faithful transmission of
genetic information during DNA replication. DNA
templates are copied with remarkably high fidelity,
checking for correct base-pair formation both at the
nucleotide insertion and at subsequent DNA extension
steps. Back in 1976, Loeb and coworkers have shown
that magnesium ions are essential for the fidelity of

DNA replication. Even though Co2+, Mn2+, and Ni2+
could substitute for Mg2+ in DNA polymerases from
different origins, the replacement resulted in a marked
decrease of replication fidelity [4,5]. Only recently,
crystal structures of different DNA polymerases com-
plexed with DNA substrates and incoming nucleotides
revealed the exact role of divalent metals, usually
magnesium ions, in this process. Very similar in all
different polymerases investigated, they position the
nucleotide and promote phosphoryl transfer. Metal
ion A interacts with the 3′-hydroxyl group of the
primer strand and has been proposed to lower its pKa,
thereby facilitating its attack on thea-phosphate of
the incoming dNTP. Metal ion B binds to and facil-
itates the leaving of theb- and g-phosphates. Both
metal ions together are also thought to stabilize the
structure and charge of the pentavalent transition state
occurring during this reaction [6–8].

1.3. Role of magnesium in DNA repair processes

DNA is continuously damaged by environmental
mutagens and by endogenous processes. To keep
mutation frequencies low, procaryotic and eucary-
otic cells have evolved different types of DNA repair
systems. Nucleotide excision repair (NER) is mainly
involved in the removal of DNA damage induced
by environmental mutagens, such as UV radiation,
polycyclic aromatic hydrocarbons, and heterocyclic
aromatic amines. The repair process is mediated by
the coordinated action of more than 20 different pro-
teins, most of them involved in damage recognition
and incision at both sides of the lesion, including
those defect in xeroderma pigmentosum. Magnesium
is an essential cofactor in basically all steps of NER.
Thus, when applying an in vitro incision reaction,
optimal magnesium concentrations were found to be
4.5–7 mM, whereas the incision was inhibited com-
pletely in the absence of Mg2+ as well as at very high
concentrations (18 mM) [9]. Examples for incision
factors with a reported requirement for magnesium
are the DNA-damage recognition protein UV-DDBP
[10], the helicase XPD [11], and the nuclease XPG
[12]. Magnesium-dependent post-incision events are
both polymerization as described above and ligation
[13,14].

Perhaps most important among endogenous pro-
cesses causing DNA damage, reactive oxygen species

A. Hartwig / Mutation Research 475 (2001) 113–121 115

(ROS) are generated as a consequence of oxygen
metabolism, leading to DNA strand breaks, DNA–
protein cross-links and a broad spectrum of oxidative
DNA base modifications [15]. Further DNA damag-
ing reactions are hydrolytic cleavages of N–glycosidic
bonds causing apurinic/apyrimidinic sites, hydrolytic
desaminations as well as methylations of oxygen
and nitrogen atoms in DNA bases [16]. Endogenous
DNA damage is mainly repaired by base excision
repair (BER). According to current models, BER is
started by removal of the modified base by a spe-
cific N–glycosylase, generating an AP site, which
is subsequently incised at its 5′ side by an AP en-
donuclease. After excision of the 5′ terminal dRP,
the single nucleotide gap is filled by DNA poly-
meraseb and sealed by ligase. Alternatively, AP
sites are further processed involving proliferating cell
nuclear antigen and DNA polymerased for the exci-
sion and DNA synthesis reaction (for recent reviews
see [17]). While DNA glycosylases usually do not
require magnesium, the opposite is true for enzymes
involved in later steps of BER. Thus, hydrolytic nu-
cleases usually depend on the participation of metal
ions, mainly magnesium, at active sites, which are
involved either in substrate interactions or directly
in the cleavage of the phosphate–oxygen bond. Dur-
ing the last few years, the exact role of magnesium
ions has been elucidated in some cases. For exam-
ple, in the major human AP site-specific DNA repair
endonuclease HAP 1 (also called APE or Ref-1), a
single magnesium ion binds to a defined Glu residue
in the active center and aids the attack on the P–O3′
bond by polarization of the P–O bond, and perhaps
by correctly orientating the phosphate group rather
than directly participating in the nucleolysis reac-
tion. Even though the enzyme can be activated in
principle by manganese and nickel ions as well, its
activity is greatly reduced (50 and 90%, respectively)
as compared to magnesium [18]. Other examples of
magnesium-dependent endonucleases in BER are the
mammalian apurinic/apyrimidinic endonuclease activ-
ity associated with the 5-hydroxymethyluracil-DNA
glycosylase [19] as well as the flap-endonuclease-1
(FEN-1), a structure-specific endonuclease involved
in DNA replication and DNA repair [20,21]. Finally,
not only the DNA synthesis mediated by polb, but
also its DNA deoxyribophosphodiesterase (dRPase)
activity acts in a Mg2+-dependent manner [22].

The third DNA repair system which has attracted
special attention during the last few years is the
mismatch repair pathway (MMR). MMR contributes
greatly to genomic stability by correcting replication
errors; defects in MMR have been associated with
increased susceptibility to hereditary nonpolyposis
colon cancer (HNPCC). In procaryotes, mismatches
are recognized by MutS and initiation as well as
subsequent steps in methyl-directed repair depend on
MutL, which interacts with the sequence-dependent
endonuclease MutH and UvrD helicase [23]. Analysis
of the crystal structure of MutL revealed an ATPase
activity which is highly conserved also in eucaryotic
homologs [24]. This activity has been proposed to
play a role in switching from a mismatch-repair initi-
ation to a processing complex. It appears to be essen-
tial for the entire repair function, since the majority of
mutations in human MLH1 found in HNPCC patients
and almost all of the mutations inEscherichia coli
MutL causing the mutator phenotype are located in
or around the ATP-binding site. Recent studies show
that MutL has an absolute requirement for Mg2+; in
the absence of magnesium, the MutL–ATP associa-
tion is abolished completely [25]. In addition to being
involved in MMR, the eucaryotic homologs MLH
and PMS have been shown to play a role in gene
conversion and chromosome segregation during and
after meiosis [26,27].

Finally, magnesium is also required for double-
strand break repair arising for example after ion-
izing radiation and during meiosis. In this context,
the hPOMp75 protein isolated from HeLa nuclei
has been shown to catalyze DNA annealing and
D-loop formation during homologous recombination
in an ATP-independent but Mg2+-dependent reaction.
Supporting its role in the maintenance of genomic
integrity, this protein has recently been shown to
be identical to the human proto-oncoprotein TLS
commonly mutated in liposarcomas [28].

1.4. Effect of magnesium on chromosome structure
and segregation

Under physiological conditions of pH and intra-
cellular K+ and Mg2+ concentrations, approximately
0.2 mol Mg2+ is bound per mole phosphate in polynu-
cleotides, and in case of DNA, stabilizes the double
helix (reviewed in [29]). Moreover, magnesium retains

116 A. Hartwig / Mutation Research 475 (2001) 113–121

the chromatin structure, and critical levels of Mg2+
are required for the maintenance of the compact state
of heterochromatin [30]. Besides these electrostatical
interactions, magnesium is also involved in chromo-
some function. For example, both free Mg2+ as well as
GTP–Mg play an essential role in the extent of tubulin
polymerization and thus in chromosome segregation
during mitosis [31].

1.5. Role of magnesium in cell cycle progression,
differentiation, and apoptosis

Accumulating evidence supports the involvement
of magnesium in the regulation of cell cycle, prolif-
eration, apoptosis, and differentiation [32]. Back in
1975, Rubin proposed magnesium as a key factor in
the coordinate control of metabolism and growth in
animal cells [33]. Later on, Maguire [34] summarized
evidence for magnesium to be a regulated and reg-
ulatory cation, mainly based on data concerning the
regulation of intracellular magnesium concentrations,
the hormonal regulation of Mg2+ transport, and the
intracellular compartmentation of Mg2+. Even though
at that time no specific physiological event was known
which was regulated by magnesium, he concluded that
Mg2+ may be a chronic regulatory agent as opposed
to Ca2+, which mediates acute “on–off” signals. Nev-
ertheless, these early studies were limited by the lack
of simple and direct methods to study intracellular
free Mg2+ in living cells. In the meantime, the role
of magnesium in the regulation of distinct cellular
processes has been elucidated in more detail. Thus,
levels of free intracellular magnesium increase in cells
undergoing apoptosis. This increase is an early event
in apoptosis, preceding DNA fragmentation and ex-
ternalization of phosphatidylserine, and is likely due
to a mobilization of magnesium from mitochondria.
Therefore, the raise in intracellular free magnesium
appears to serve as a “second messenger” for down-
stream events in apoptosis [35]. On the protein level,
Ca2+- and Mg2+-dependent endonucleases have been
implicated in DNA fragmentation during apoptosis
[36]. With respect to proliferation and differentia-
tion, it has been shown that magnesium restriction
causes differentiation in human leukemic HL-60
cells. Again, a shift in intracellular magnesium was
observed on these conditions: while the total magne-
sium content markedly decreased, no change occurred

in the concentration of free intracellular magnesium.
As possible reason for cell cycle arrest and differ-
entiation, a marked increase in the expression level
of the CDK inhibitor p27kip1, a negative regulator of
cell cycle progression, has been observed [37]. Taken
together, current evidence suggests that besides being
a cofactor of many enzymatic reactions involved in
basically all cellular processes, magnesium is involved
in the regulation of cell cycle control and apoptosis,
mediated mainly by the mobilization of magnesium
from intracellular pools. Nevertheless, much remains
to be done on this field to further clarify this aspect.

2. Genotoxicity and anti-genotoxicity

From studies conducted either in bacteria or in
mammalian cells in culture, there is no evidence for
genotoxic effects of magnesium salts at physiolo-
gically relevant doses. Thus, at concentrations up to
20 mM magnesium sulfate, no increased mutation
frequency was observed after treatment of diverse
strains ofSalmonella typhimuriumand no enhance-
ment of chromosomal aberrations was seen in Chinese
hamster lung cells [38]. Similarly, no enhanced fre-
quency of sister-chromatid exchanges was observed in
Chinese hamster ovary (CHO) cells [39]. Only at
8 mg/ml (40 mM) MgCl2, an increase in chromatid
gaps and chromatid breaks in Chinese hamster
lung fibroblasts was detected, most likely due to
ionic imbalance [40]. In contrast, magnesium ex-
erts pronounced protective effects in combination
with certain transition metals. For example, raising
the extracellular concentration of MgCl2 inhibited
nickel-induced DNA strand breaks, DNA–protein
cross-links, sister-chromatid exchanges, chromosomal
aberrations, and cell transformation [39]. Further-
more, the addition of magnesium prevented nickel-
and cobalt-induced inhibition of nucleotide excision
repair [41,42]. The reason for these anti-genotoxic
effects may include not only altered metal transport at
the cellular level, but also competitions at critical bind-
ing sites. Thus, excess magnesium given to nuclear
extracts of nickel-treated human HeLa cells reversed
the nickel-induced disturbance of DNA–protein inter-
actions involved in DNA-damage recognition during
nucleotide excision repair, indicating the displacement
of Mg2+ by Ni2+ and vice versa. Nevertheless, these

A. Hartwig / Mutation Research 475 (2001) 113–121 117

protective effects are metal-specific, since in the same
test system no reversal was seen in case of nuclear
extracts derived from cadmium-treated cells [43].

3. Carcinogenicity, anti-carcinogenicity, and aging
as related to nutrition in animals and humans

The relationship between magnesium and cancer
is rather complex, and several aspects have to be
considered separately, including (i) the potential im-
pact of magnesium deficiency on tumor incidences,
(ii) disturbed magnesium homeostasis frequently
observed in tumor cells, and (iii) the effect of either
magnesium deficiency or supplementation on the
progression of existing tumors.

3.1. Animal studies

In a long-term feeding study in mice, magnesium
chloride at dose levels up to 2% in the diet was
shown to be not carcinogenic [44]. On the other hand,
there is convincing evidence from animal studies that
magnesium-deficient diets increase the incidences of
thymic tumors and leukemias [45,46]. Furthermore,
magnesium-deficient diets increased chromosomal
aberrations in rats in maternal and fetal tissues [47].
However, as emphasized by Durlach et al. [48],
carcinogenic effects of magnesium deficiency were
restricted to rats and not observed in other species.

In combination with other carcinogens, magnesium
has been shown to increase or decrease their tumori-
genic potency, depending on the compound investi-
gated and on the experimental conditions. This issue
has been extensively reviewed [47,49] and the main
findings can be summarized as follows. With respect
to toxic metal compounds, magnesium prevented the
formation of pulmonary tumors after i.p. injections
of nickel or lead in mice [50]. Furthermore, when
magnesium carbonate was administered together with
the strong carcinogen Ni3S2 i.m. in male Fischer rats,
it strongly inhibited the formation of local tumors as
well as kidney tumors after simultaneous injection
into the renal cortex; however, no protection towards
Ni3S2 was seen after dietary treatment with magne-
sium carbonate. On the other hand, dietary magne-
sium did protect from 3-methyl cholantrene-induced
fibrosarcomas [51]. Enhanced tumor incidences in

the respiratory tract were observed when magnesium
oxide was used as a carrier dust for intra-tracheal
instillation of benzo(a)pyrene [52].

Nevertheless, it is important to emphasize that the
predominantly protecting effects of magnesium exist
only at the very early stage of tumorigenesis. Thus,
magnesium deficiency antagonizes tumor implantation
and inhibits growth of induced or spontaneous tumors
in the rat (e.g. [53]), while magnesium intake stimu-
lates tumor development. The reason appears to be a
disturbance in magnesium homeostasis in carcinogen-
esis, leading to magnesium accumulation in tumors
with magnesium depletion in non-neoplastic tissues.
These alterations in magnesium distribution may play
an important role on the neoplastic development at the
membrane, cytosol, or chromatin level [48].

One other aspect repeatedly discussed in literature
is the relationship between magnesium deficiency and
oxidative stress. Thus, magnesium-deficient animals
show an increased susceptibility to an in vivo oxida-
tive stress, and their tissues are more susceptible to in
vitro peroxidation. Consequences are increased ox-
idative damage of cellular lipids, proteins, and nucleic
acids, which may lead to altered membrane functions,
perturbations of intracellular calcium metabolism,
cardiovascular diseases, accelerated aging, and car-
cinogenesis (for reviews see [54]). This is also sup-
ported experimentally in isolated bovine endothelial
cells: when growing in magnesium-deficient medium,
they exerted higher levels of lipid peroxidation and
intracellular oxidative injury after exposure to oxy-
gen radicals produced by dihydroxyfumarate and
Fe3+–ADP as compared to control cells [55].

Finally, experimental magnesium deficiency leads
to alterations in the immune response. Thus, for
example, rats on magnesium-deficient diet showed
accelerated thymus involution accompanied by
enhanced apoptosis [56].

3.2. Humans

The average dietary intake of magnesium is ap-
proximately 300 mg per day; main sources are green
vegetables, grains, meat, and — depending on its
hardness — drinking water. Magnesium balance is
mediated by intestinal absorption and renal excretion.
Thus, the amount of magnesium absorbed in the small
intestine from a low magnesium diet may be as high as

118 A. Hartwig / Mutation Research 475 (2001) 113–121

75%, from a high magnesium diet as low as 25% [57].
In most industrialized countries, magnesium intake
is marginal in the entire population. Recommended
daily dietary allowances (RDA) values for the US
have been increased recently to 320 mg Mg per day
for adult women and to 420 mg Mg per day for adult
men [58]. In Germany, Switzerland, and Austria, the
reference values are 300 for women and 350 for men
of 25 years and older [59]. However, the actual mag-
nesium intake has been found to be 67–77% of RDA
in West Germany, England, and some regions of North
America. Even lower magnesium intakes of barely
50% RDA were observed in Newfoundland and Japan
[60]. These low magnesium intakes were confirmed in
a recent study conducted in Germany, where magne-
sium intake and balance were determined in a control
and in a magnesium-supplemented group. Addition-
ally, the authors showed that a full balance required
at least 375 mg Mg per day for women and men,
while positive balances were observed between 950
and 1020 mg Mg per day [61]. Whether the current
RDA values are sufficient in stress situations and to
prevent cardiovascular diseases was also questioned
by Seelig [62], emphasizing inter-individual genetic
differences and interactions with other dietary con-
stituents in magnesium-dependent reactions. Reasons

Fig. 1. Effect of magnesium on cellular processes related to genomic stability.

for magnesium deficiency can be manifold. They in-
clude inadequate dietary intake, malabsorption in the
gastrointestinal tract, and renal losses, for example
due to a drug therapy with diuretics. Alterations in
magnesium metabolism are being increasingly im-
plicated as significant factor in the pathogenesis of
ischemic heart disease (IDH); hypomagnesia has been
shown to cause an increased incidence of arrythmias,
coronary vasospasm, and hypertension, and may pre-
dispose to the pathogenesis of artherosclerosis [63].
On the other hand, oral magnesium supplementa-
tion of magnesium-deficient persons can lower blood
pressure, improve brain performance, concentration,
and stress tolerance [60]. As emphasized by Durlach
et al. [64], magnesium deficiency may be particu-
larly pronounced in elderly persons, since magnesium
absorption decreases with age, and in some cases,
urinary magnesium leakage may be increased, lead-
ing to a reduction of magnesium-exchangeable pools.
Due to altered membrane functions and increased
susceptibility to oxidative stress described above,
magnesium-deficiency may constitute an important
factor in accelerated aging processes. With respect to
carcinogenicity, early epidemiological studies pointed
towards higher incidences of neoplasms and lympho-
proliferative disorders in region with low magnesium

A. Hartwig / Mutation Research 475 (2001) 113–121 119

Fig. 2. Proposed effects of magnesium status on genomic stability in experimental animals and humans: (1) results from animal studies;
(2) protective effects are restricted to very early stages of tumor development; (3) requires simultaneous administration via the same route;
dietary supplementation usually ineffective. For references and further details see text.

contents in soil and drinking water (reviewed in [65]).
This is supported by a recent case-control study con-
ducted in Taiwan, where high levels of magnesium
in drinking water exerted protective effects against
gastric cancer [66].

4. Conclusions and perpectives

As evident from the multiple functions of magne-
sium in basically all cellular processes, magnesium is
highly required to maintain genomic stability (sum-
marized in Fig. 1). Being not genotoxic at physiolog-
ically relevant concentrations itself, it keeps mutation
frequencies low by facilitating high replication fidelity
and by supporting basically all DNA repair processes
as well as chromosome segregation during mitosis. In
some cases, it exerts protective effects in combination
with other genotoxic and/or carcinogenic agents such
as certain toxic metal ions, most likely by competition
at critical binding sites. Thus, it is not surprising that
intracellular magnesium is compartmentalized, and
changes in free Mg2+ concentrations serve as signals
for cell cycle regulation and apoptosis. Animal ex-
periments and epidemiological studies show inverse
correlations between magnesium status and cardio-
vascular diseases, and magnesium deficiency appears

to be one susceptibility factor for accelerated aging.
Nevertheless, the relationship between magnesium
status and tumorigenesis is more complex: while mag-
nesium deficiency tends to increase tumor incidences
in animals and humans, magnesium promotes the
growth of pre-existing tumors due to profound changes
of magnesium homeostasis in tumor cells; thus, pro-
tecting effects are restricted to early steps in tumor
development (summarized in Fig. 2). With respect to
magnesium status in humans, magnesium intake in
many countries is considerably lower compared to cur-
rent RDA values and marginal magnesium deficiency
is very common. Thus, it is highly recommended to in-
crease actual magnesium intake to maintain genomic
stability and to protect from age-related diseases.

References

[1] A. Romani, A. Scarpa, Regulation of cell magnesium, Arch.
Biochem. Biophys. 298 (1992) 1–12.

[2] R.D. Grubbs, Hormonal regulation of magnesium homeostasis
in cultured mammalian cells, in: H. Sigel, A. Sigel (Eds.),
Metal Ions in Biological Systems, Marcel Dekker, New York,
1990, pp. 177–192.

[3] T. Guenther, Membrane transport of magnesium, in: H. Sigel,
A. Sigel (Eds.), Metal Ions in Biological Systems, Marcel
Dekker, New York, 1990, pp. 215–225.

[4] M.A. Sirover, L.A. Loeb, Metal activation of DNA synthesis,
Biochem. Biophys. Res. Commun. 70 (1976) 812–817.

120 A. Hartwig / Mutation Research 475 (2001) 113–121

[5] M.A. Sirover, L.A. Loeb, On the fidelity of DNA replication,
J. Biol. Chem. 252 (1977) 3605–3610.

[6] S. Doublié, S. Tabor, A.M. Long, C.C. Richardson, T.
Ellenberger, Crystal structure of a bacteriophage T7 DNA
replication complex at 2.2 Å resolution, Nature 391 (1998)
251–258.

[7] J.R. Kiefer, C. Mao, J.C. Braman, L.S. Beese, Visualization
DNA replication in a catalytically activeBacillus DNA poly-
merase crystal, Nature 391 (1998) 304–307.

[8] T.A. Steitz, DNA polymerases: structural diversity and
common mechanisms, J. Biol. Chem. 274 (1999)
17395–17398.

[9] P. Calsou, B. Salles, Properties of damage-dependent DNA
incision by nucleotide excision repair in human cell-free
extracts, Nucleic Acids Res. 22 (1994) 4937–4942.

[10] S. Keeney, G.J. Chang, S. Linn, Characterization of a
human DNA damage binding protein implicated in xeroderma
pigmentosum E, J. Biol. Chem. 268 (1993) 21293–21300.

[11] P. Sung, V. Bailly, C. Weber, L.H. Thompson, L. Prakash,
S. Prakash, Human xeroderma pigmentosum group D gene
encodes a DNA helicase, Nature 365 (1993) 852–855.

[12] A. O’Donovan, R.D. Wood, Identical defects in DNA repair
in xeroderma pigmentosum group G and rodent ERCC group
5, Nature 363 (1993) 185–188.

[13] H. Teraoka, K. Tsukada, Eucaryotic DNA ligase. Purification
and properties of the enzyme from bovine thymus, and
immunochemical studies of the enzyme from animal tissues,
J. Biol. Chem. 257 (1982) 4758–4763.

[14] S.W. Yang, J.Y.H. Chan, Analysis of the formation of
AMP-DNA intermediate and the successive reaction by
human DNA ligases I and II, J. Biol. Chem. 267 (1992)
8117–8122.

[15] B. Halliwell, O.I. Aruoma, DNA damage by oxygen-derived
species. Its mechanism and measurement in mammalian cells,
FEBS Lett. 281 (1991) 9–19.

[16] T. Lindahl, Instability and decay of the primary structure of
DNA, Nature 362 (1993) 709–715.

[17] V.A. Bohr, G.L. Dianov, Oxidative DNA damage processing
in nuclear and mitochondrial DNA, Biochimie 81 (1999)
155–160.

[18] G. Barzilay, C.D. Mol, C.N. Robson, L.J. Walker, R.P.
Cunningham, J.A. Tainer, I.D. Hickson, Identification of
critical active-site residues in the multifunctional human DNA
repair enzyme HAP1, Nat. Struct. Biol. 2 (1995) 561–568.

[19] S. Cannon-Carlson, H. Gokhale, G.W. Teebor, Purification and
characterization of 5-hydroxymethyluracil-DNA glycosylase
from calf thymus, J. Biol. Chem. 264 (1989) 13306–13312.

[20] B. Shen, J.P. Nolan, L.A. Sklar, M. Park, Essential amino
acids for substrate binding and catalysis of human flap
endonuclease 1, J. Biol. Chem. 271 (1996) 9173–9176.

[21] G. Frank, J. Qiu, M. Somsouk, Y. Wenig, L. Somsouk,
J.P. Nolan, B. Shen, Partial functional deficiency of E160D
flap endonuclease-1 mutant in vitro and in vivo is due to
defective cleavage of DNA substrates, J. Biol. Chem. 273
(1998) 33064–33072.

[22] Y. Matsumoto, K. Kim, Excision of deoxyribose phosphate
residues by DNA polymeraseb during DNA repair, Science
269 (1995) 699–702.

[23] P. Modrich, R. Lahue, Mismatch repair in replication fidelity,
genetic recombination, and cancer biology, Annu. Rev.
Biochem. 65 (1996) 101–133.

[24] C. Ban, W. Yang, Crystal structure and ATPase activity of
MutL: implications for DNA repair and mutagenesis, Cell 95
(1998) 541–552.

[25] C. Ban, M. Junop, W. Yang, Transformation of MutL by ATP
binding and hydrolysis: a switch in DNA mismatch repair,
Cell 97 (1999) 85–97.

[26] M.S. Williamson, J.C. Garne, S. Fogel, Meiotic gene
conversion mutants inSaccharomyces cerevisiae. I. Isolation
and characterization of pms1-1 and pms1-2, Genetics 110
(1985) 609–646.

[27] N. Hunter, R.H. Borts, Mlh1 is unique among mismatch
repair proteins in its ability to promote crossing-over during
meiosis, Genes Dev. 11 (1997) 1433–1442.

[28] H. Baechtold, M. Kuroda, J. Sok, D. Ron, B.S. Lopez, A.
Akhmedov, Human 75 kDa DNA pairing protein is identical to
the pro-oncoprotein TLS/FUS and is able to promote D-loop
formation, J. Biol. Chem. 274 (1999) 34337–34342.

[29] G.L. Eichhorn, The effect of metal ions on the structure and
function of nucleic acids, in: G.L. Eichhorn, L.G. Marzilli
(Eds.), Metal Ions in Genetic Information Transfer, Elsevier,
New York, 1981, pp. 1–46.

[30] A. Weith, Mg2+-dependent compactness of heterochromatic
chromosome segments, Exp. Cell Res. 146 (1983) 199–203.

[31] S. Grover, E. Hamel, The magnesium–GTP interaction in
microtubule assembly, Eur. J. Biochem. 222 (1994) 163–172.

[32] F.I. Wolf, A. Cittadini, Magnesium in cell proliferation and
differentiation, Frontiers Biosci. 4 (1999) 1–11.

[33] H. Rubin, Central role for magnesium in coordinate control
of metabolism and growth in animal cell, Proc. Natl. Acad.
Sci. U.S.A. 72 (1975) 3551–3555.

[34] M.E. Maguire, Magnesium: a regulated and regulatory cation,
in: H. Sigel, A. Sigel (Eds.), Metal Ions in Biological Systems,
Marcel Dekker, New York, 1990, pp. 135–153.

[35] M.M. Chien, K.E. Zahradka, M.K. Newell, J.H. Freed,
Fas-induced B cell apoptosis requires an increase in free
cytosolic magnesium as an early event, J. Biol. Chem. 274
(1999) 7059–7066.

[36] C. Giannakis, I.J. Forbes, P.D. Zalewski, Calcium/magnesium
dependent nuclease: tissue distribution, relationship to
inter-nucleosomal DNA fragmentation and inhibition by zinc,
Biochem. Biophys. Res. Commun. 181 (1991) 915–920.

[37] V. Covacci, N. Bruzzese, A. Sgambato, A. Di Francesco,
M.A. Russo, F.I. Wolf, A. Cittadini, Magnesium restriction
induces granulocytic differentiation and expression of P27Kip1

in human leukemic HL-60 cells, J. Cell. Biochem. 70 (1998)
313–322.

[38] Y. Oguma, F. Yokota, K. Inoue, K. Shimamura, Mutagenicity
studies of magnesium sulfate — reverse mutation test with
bacteria and chromosomal aberration test with mammalian
cells in culture, J. Toxicol. Sci. 23 (Suppl. 1) (1998) 81–90.

[39] K. Conway, X.-W. Wang, L. Xu, M. Costa, Effect of magne-
sium on nickel-induced genotoxicity and cell transformation,
Carcinogenesis 8 (1987) 1115–1121.

A. Hartwig / Mutation Research 475 (2001) 113–121 121

[40] J. Ashby, M. Ishidate Jr., Clastogenicity in vitro of the Na,
K, Ca and Mg salts of saccharin, and of magnesium chloride,
consideration of significance, Mutat. Res. 163 (1986) 63–73.

[41] A. Hartwig, L.H.F. Mullenders, R. Schlepegrell, U. Kasten,
D. Beyersmann, Nickel(II) interferes with the incision step in
nucleotide excision repair, Cancer Res. 54 (1994) 4045–4051.

[42] U. Kasten, L.H.F. Mullenders, A. Hartwig, Cobalt(II) inhibits
the incision and the polymerization step of nucleotide excision
repair in human fibroblasts, Mutat. Res. 383 (1997) 81–89.

[43] M. Hartmann, A. Hartwig, Disturbance of DNA damage
recognition after UV-irradiation by nickel(II) and cadmium(II)
in mammalian cells, Carcinogenesis 19 (1998) 617–621.

[44] Y. Kurata, S. Tamano, M.A. Shibata, A. Hagiwara, S.
Fukushima, N. Ito, Lack of carcinogenicity of magnesium
chloride in a long-term feeding study in B6C3F1 mice, Food
Chem. Toxicol. 27 (1989) 559–563.

[45] P. Bois, Tumour of the thymus in magnesium-deficient rats,
Nature 204 (1964) 1316.

[46] K.S. Kasprzak, M.P. Waalkes, The role of calcium, magne-
sium, and zinc in carcinogenesis, in: L.A. Poirier, P.M.
Newbarne, M.W. Pariza (Eds.), Essential Nutrients in
Carcinogenesis, Plenum Press, New York, 1986, pp. 497–515.

[47] L.T. Bell, M. Branstrator, C. Roux, L.S. Hurley, Chromosomal
abnormalities in maternal and fetal tissues of magnesium- or
zinc-deficient rats, Teratology 12 (1975) 221–226.

[48] J. Durlach, M. Bara, A. Guiet-Bara, Magnesium and its
relationship to oncology, in: H. Sigel, A. Sigel (Eds.), Metal
Ions in Biological Systems, Marcel Dekker, New York, 1990,
pp. 549–578.

[49] K.S. Kasprzak, Effects of calcium, magnesium, zinc, and
iron on nickel carcinogenesis: inhibition versus enhancement,
in: N.D. Hadjiliadis (Ed.), Cytotoxic, Mutagenic and
Carcinogenic Potential of Heavy Metals Related to Human
Environment, Kluwer Academic Publishers, Dordrecht, 1997,
pp. 93–106.

[50] L.A. Poirier, J.C. Theiss, L.J. Arnold, M.B. Shimkin,
Inhibition by magnesium and calcium acetates of lead
subacetate- and nickel acetate-induced lung tumors in strain
A mice, Cancer Res. 44 (1984) 1520–1522.

[51] T. Patiroglu, G. Sahin, O. Kontas, K. Üzüm, R.
Raraymen, Protective effect of magnesium supplementation
on experimental 3-methyl cholantrene-induced fibrosarcoma
and changes in tissue magnesium distribution during carcino-
genesis in rats, Biol. Trace Element Res. 56 (1997) 179–185.

[52] F. Stenback, A. Sellakumar, P. Shubik, Magnesium oxide as
carrier dust in benzo[a]pyrene-induced lung carcinogenesis
in Syrian hamsters, J. Natl. Cancer Inst. 54 (1975) 861–867.

[53] B.J. Mills, R.D. Lindemann, C.A. Lang, Magnesium
deficiency inhibits biosynthesis of blood glutathione and
tumor growth in the rat, Proc. Soc. Exp. Biol. Med. 181
(1986) 326–332.

[54] Y. Rayssiguier, J. Durlach, E. Gueux, E. Rock, A. Mazur,
Magnesium and ageing. I. Experimental data: importance of
oxidative damage, Magnesium Res. 6 (1993) 369–378.

[55] B.F. Dickens, W.B. Weglicki, Y.-S. Li, I.T. Mak, Magnesium
deficiency in vitro enhances free radical-induced intracellular
oxidation and cytotoxicity in endothelial cells, FEBS Lett.
311 (1992) 187–191.

[56] C. Malpuech-Brugere, W. Wojciech, E. Gueux, J. Kuryszko,
E. Rock, Y. Rayssiguier, A. Mazur, Accelerated thymus
involution in magensium-deficient rats is related to enhanced
apoptosis and sensitivity to oxidative stress, Br. J. Nutr. 81
(1999) 405–411.

[57] R.J. Elin, The assessment of magnesium status in humans, in:
H. Sigel, A. Sigel (Eds.), Metal Ions in Biological Systems,
Marcel Dekker, New York, 1990, pp. 579–596.

[58] Food and Nutrition Board, Dietary Reference Intakes for
Calcium, Phosphorous, Magnesium, Vitamin D, and Fluoride,
National Academy Press, Washington, DC, 1999.

[59] Deutsche Gesellschaft für Ernährung (DGE), Östereichische
Gesellschaft für Ernährung (ÖGE), Schweizerische Gesell-
schaft für Ernährung (SGE), Schweizerische Vereinigung für
Ernährung (SVE), Referenzwerte für die Nährstoffzufuhr,
Umschau Braus GmbH, Frankfurt am Main, 2000.

[60] J.R. Marier, Dietary magnesium and drinking water, in: H.
Sigel, A. Sigel (Eds.), Metal Ions in Biological Systems,
Marcel Dekker, New York, 1990, pp. 85–104.

[61] M. Wangemann, A. Selzer, C. Leitzmann, S. Golf, V. Graef,
N. Katz, Empfehlungen zur Magnesium-Zufuhr, Magnesium
Bull. 17 (1995) 79–85.

[62] M.S. Seelig, Nutritional status and requirements of
magnesium, Magnesium Bull. 8 (1986) 170–185.

[63] N. Brautbar, A.T. Roy, P. Hom, D.B.N. Lee, Hypomagnesia
and hypermagnesia, in: H. Sigel, A. Sigel (Eds.), Metal Ions
in Biological Systems, Marcel Dekker, New York, 1990,
pp. 285–320.

[64] J. Durlach, P. Bac, V. Durlach, Y. Rayssiguier, M. Bara,
A. Guiet-Bara, Magnesium status and ageing: an update,
Magnesium Res. 11 (1997) 25–42.

[65] J.M. Blondell, The anticarcinogenic effect of magnesium,
Med. Hypotheses 6 (1980) 863–871.

[66] C.-Y. Yang, M.-F. Cheng, S.-S. Tsai, Y.-L. Hsieh, Calcium,
magnesium, and nitrate in drinking water and gastric cancer,
Jpn. J. Cancer Res. 89 (1998) 124–130.