Reviews of Physiology, Biochemistry and Pharmacology 176
Reviews of Physiology, Biochemistry and Pharmacology 176
Reviews of Physiology, Biochemistry and Pharmacology 176
Physiology,
Biochemistry and
Pharmacology
176
Reviews of Physiology, Biochemistry
and Pharmacology
More information about this series at http://www.springer.com/series/112
Bernd Nilius Pieter de Tombe
Thomas Gudermann Reinhard Jahn Roland Lill
Editors
Reviews of Physiology,
Biochemistry and
Pharmacology
176
Editor in Chief
Bernd Nilius
Department of Cellular and
Molecular Medicine
KU Leuven
Leuven, Belgium
Editors
Pieter de Tombe Thomas Gudermann
Heart Science Centre Walther-Straub Institute for Pharmacology
The Magdi Yacoub Institute and Toxicology
Harefield, United Kingdom Ludwig-Maximilians University of Munich
Munich, Germany
This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
v
Rev Physiol Biochem Pharmacol (2019) 176: 1–36
DOI: 10.1007/112_2018_12
© Springer Nature Switzerland AG 2018
Published online: 2 August 2018
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2 Protein Tyrosine Phosphatase (PTP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 Dual-Specificity Phosphatase (DUSP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 ADUSP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 DUSP3/VHR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3 Molecular and Biological Functions of DUSP3/VHR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1 DUSP3 in Tumorigenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2 DUSP3 in Genomic Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 DUSP3 in Blood-Associated Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4 Overview of Current Knowledge on DUSP3/VHR Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Abstract Protein tyrosine kinases (PTK), discovered in the 1970s, have been
considered master regulators of biological processes with high clinical significance
as targets for human diseases. Their actions are countered by protein tyrosine
phosphatases (PTP), enzymes yet underrepresented as drug targets because of the
high homology of their catalytic domains and high charge of their catalytic pocket.
This scenario is still worse for some PTP subclasses, for example, for the atypical
dual-specificity phosphatases (ADUSPs), whose biological functions are not even
completely known. In this sense, the present work focuses on the dual-specificity
phosphatase 3 (DUSP3), also known as VH1-related phosphatase (VHR), an uncom-
mon regulator of mitogen-activated protein kinase (MAPK) phosphorylation.
DUSP3 expression and activities are suggestive of a tumor suppressor or tumor-
promoting enzyme in different types of human cancers. Furthermore, DUSP3 has
other biological functions involving immune response mediation, thrombosis,
hemostasis, angiogenesis, and genomic stability that occur through either MAPK-
1 Introduction
In recent years there has been substantial progress in the development of protein
tyrosine phosphatase (PTP) inhibitors suggesting that these enzymes, for long time
considered undruggable, can provide unique solutions for the treatment of human
diseases (Hendriks et al. 2013; Tonks 2013). Many recent strategies have been used
in the development of drugs that target PTPs as ways of expanding the possibilities
of intervention in the function of these enzymes and in biological processes depen-
dent on them. These strategies include (a) orthosteric inhibitors (reversible compet-
itive, bidentate or uncompetitive, and irreversible), (b) allosteric inhibitors,
(c) oligomerization inhibitors, (d) radioimmunotherapy, and (e) PTP receptor
biological decoy (He et al. 2013; Stanford and Bottini 2017).
The continuous generation of more selective probes of high quality for the
activity of individual PTPs is essential for the successful development of inhibitory
drugs of these enzymes. New chemical inhibitors are enhancing performance in
PTPs already considered clinical targets such as PTP1B and SHP-2 and are bringing
into focus new targets such as STEP, PTPN22, VE-PTP, CD45, CDC25A/B/C, and
LMPTP. Advances of functional studies of RPTPs are revealing new opportunities
for inhibiting PTP domains within the receptor structure using small biological
molecules that act to stabilize the oxidation of catalytic intermediates or even the
formation of receptor complexes. But, there are no magic bullets to attack the PTPs,
and a successful strategy for an enzyme from one of the classes may not be as
effective for a member of another class or even for a different member of the same
class (He et al. 2013; Stanford and Bottini 2017).
This is the case, for example, of the dual-specificity phosphatases (DUSPs),
which have a diversity of structural possibility of substrates, many of which have
not yet been identified both structurally and functionally (Tonks 2013). Among the
members of this subclass I of PTPs (Fig. 1), DUSP6 and PRL-1/2/3 are being
considered druggable targets for some deficiencies of the immune system, including
DUSP3/VHR Is a Potential Drug Target 3
Fig. 1 A brief classification of the protein tyrosine phosphatase (PTP) superfamily within the
human genome showing the numerical distribution in the four classes and highlighting the DUSP3
belonging to the atypical dual phosphatases
cancer, and also for melanomas (Stanford and Bottini 2017). This review has as
differential a specific focus on DUSP3, aiming to enlarge the list of druggable targets
of the atypical dual phosphatases (He et al. 2013; Hendriks et al. 2013). We first
describe all its molecular targets and biological functions described in the literature;
and, secondly, we reexamine all inhibitors already developed and discuss their
mechanism of action, specificity, permeability, bioavailability, and potential as a
drug or as a starting point for drug development.
PTPs are very specific, non-redundant, and catalytically active enzymes. While
protein tyrosine kinases (PTKs) constitute a superfamily of enzymes with the same
evolutionary origin, PTPs have distinct evolutionary origins (Alonso et al. 2004c;
4 L. F. Monteiro et al.
Manning et al. 2002; Vang et al. 2008). PTPs are classified into four subfamilies
according to amino acid sequences that share their catalytic domain and according to
the presence of cysteine (Cys or C) or aspartate (Asp or D) in the catalytic cleft that
acts as a catalytic amino acid. PTPs expression patterns vary; there are enzymes with
wide distribution and some are specific to certain tissues. Most human cells express
30–60% of all the PTPs genes; neural and hematopoietic cells commonly express
more PTPs than other tissues (Vang et al. 2008).
Three subfamilies (classes I–III) present cysteine-based catalysis, and together,
they make up almost all of the PTPs, comprising 99 proteins (Fig. 1). Class I is
subdivided into both classical tyrosine phosphatases and dual phosphatases, which
are the more diversified phosphatases that dephosphorylate not only the tyrosine
(Tyr) residue (Alonso et al. 2004c). Class II is also Tyr-specific, and its only
representative is the low molecular weight protein Tyr phosphatase (LMWPTP), a
highly conserved evolutionary enzyme that may have broad implications for human
health (Patterson et al. 2009). Class III differs from the others by the presence of a
rhodanese structure. It is composed of cell cycle regulatory proteins, cell division
cycle 25 (Cdc25) phosphatases, which activate cyclin-dependent kinases (CDKs)
when they remove phosphate pools from some Tyr and Thr residues present in their
regulatory sites (Mustelin 2007). The fourth and last subfamily (Class IV) is formed
by phosphatases that have aspartate as a critical residue in the catalytic cleft: (1) EyA
(eyes absent) proteins play an important role in the organogenesis of vertebrates, and
(2) haloacid dehalogenase (HAD) proteins are able to dephosphorylate both Tyr and
Ser/Thr residues from various substrates, including proteins, sugars, nucleotides, and
phospholipids (Fig. 1) (Bayón and Alonso 2010; Mustelin 2007; Patterson et al.
2009).
The dual-specificity phosphatase (DUSP) group is the largest and most diversified
among the nonclassical PTPs and is composed of 61 proteins (Fig. 1). DUSPs are
able to dephosphorylate both Tyr and Ser/Thr residues due to their catalytic site’s
structure, which is not as deep as and more open than that of classical phosphatases.
The consensus sequence, HC(X)5R, present in the catalytic domain of classical
phosphatases and DUSPs, is highly conserved. At the base of the catalytic cleft is
the Cys residue, which characterizes classes I–III (Alonso et al. 2004c; Farooq and
Zhou 2004; Mandl et al. 2005). DUSP’s catalytic mechanism and that of classical
phosphatases are similar and involve substrate hydrolysis and formation of a stable
phosphoryl intermediate, with an arginine residue near the catalytic slit contributing
directly to the reaction catalysis; a slightly distant aspartate acid protonates the
phosphate group (Fig. 2) (Bayón and Alonso 2010; Denu and Dixon 1995, 1998).
The DUSP subfamily has diverse biological roles as evidenced by its subdivision
into 16 groups. It has been well established that it is involved in mitogen-activated
protein kinase (MAPK) pathway regulation, acting mainly on extracellular-regulated
DUSP3/VHR Is a Potential Drug Target 5
Fig. 2 Details of the DUSP3/VHR crystal structure presenting the four amino acid residues more
relevant to its enzymatic activity. (a) The catalytic cysteine (Cys) 124 (yellow) is shown in close
proximity to histidine (His) 123 and arginine (Arg) 130 residues (red) comprising the catalytic triad.
(b) The regulatory tyrosine (Tyr) 138 (green) is sitting at a central alpha helix, relatively distant
from the catalytic cysteine (yellow) (modified from 1vhr.pdb) (Yuvaniyama et al. 1996)
kinase (ERK)1/2, jun kinase (JNK), and p38 (Bayón and Alonso 2010; Bermudez
et al. 2010; Nunes-Xavier et al. 2011; Patterson et al. 2009; Pulido and Hooft van
Huijsduijnen 2008). In this context, DUSP subfamily members play important roles
in several cell events: (1) in cell cycle regulation, a function usually performed by the
MAPK phosphatases (MKPs, including PAC1, MKP1–5, MKP7, hVH3, hVH5,
PYST2, and MK-STYX) and some other atypical DUSPs, including DUSP3, the
protein of interest in this work (Nunes-Xavier et al. 2011; Patterson et al. 2009;
Pulido and Hooft van Huijsduijnen 2008), (2) in several types of cancer (MKPs1–3,
MKP8, PAC1, DUSP3 and 5, PTEN, and PRLs) (Arnoldussen and Saatcioglu 2009;
Bermudez et al. 2010; Nunes-Xavier et al. 2011; Pulido and Hooft van Huijsduijnen
2008), (3) in immune responses and inflammation (MKP1, 5, and 6, PAC1, and
DUSP3) (Jeffrey et al. 2007; Lang et al. 2006; Salojin and Oravecz 2007), and
(4) hereditary diseases (MTM and Laforin) (Bayón and Alonso 2010; Patterson et al.
2009; Pulido and Hooft van Huijsduijnen 2008).
DUSP subgroups are organized according to shared sequence similarity and by
the presence of specific structures; for example, the MKPs have the CDC25 homol-
ogy 2 (CH2) at their N-terminal, while the myotubularins (MTM) present a
pleckstrin homology (PH) domain at its N-terminus, which explains the activity of
MTM on lipids (Bermudez et al. 2010; Nunes-Xavier et al. 2011; Patterson et al.
2009). The less characterized and even more diversified subgroup are the small
(generally <250 aa) atypical dual phosphatases or ADUSPs (Fig. 1).
6 L. F. Monteiro et al.
2.2 ADUSP
2.3 DUSP3/VHR
DUSP3 (also known as vaccinia H1-related phosphatase or VHR) was the first dual
phosphatase identified in mammals in 1991 (Ishibashi et al. 1992) and also the first
ADUSP crystallized in 1996 (Fig. 2) (Yuvaniyama et al. 1996). It is constitutively
active, widely expressed in several tissues, and can be found in the nucleus or cytosol
as these locations are important for its functions (Alonso et al. 2001; Rahmouni et al.
2006). Among the roles attributed to DUSP3 (discussed in more detail in the
following sections) are cell cycle control, proliferation, and senescence of some
types of cancer and also in the immune responses in which the majority of cases are
mediated by the MAPK substrates (Fig. 3) (Alonso et al. 2003; Bayón and Alonso
2010; Hoyt et al. 2007; Jeffrey et al. 2007; Kondoh and Nishida 2007; Nunes-Xavier
et al. 2011; Patterson et al. 2009; Rahmouni et al. 2006; Salojin and Oravecz 2007).
DUSP3 is an ~21 kDa enzyme with 185 residues and no known signal sequence
in addition to its basic core in which the HC(X)5R consensus sequence common to
the Class I PTPs is present. DUSP3’s catalytic triad consisting of His-123 (red),
DUSP3/VHR Is a Potential Drug Target 7
Cyc D1
PKC
Cell Adhesion and Migraon
NF-κB
Phosphatase Acvators
PLCγ
MAPK
STAT5 EGFR
VRK3
Paxillin FAK DUSP3/VHR PTKs
NBS1
NPM1
HRNPC NUCL
ATM cMyc
CHK1/2 ARF
p53 p21
Chromosome Segregaon,
Fig. 3 DUSP3 substrates, validated both in vivo and in vitro (first layer of neighbors), identified
in vitro, and predicted in silico (first and second layer of neighbors), all work in conjunction to
mediate this phosphatase’s diverse biological functions shown in the scheme (magenta periphery).
Under specific conditions, some protein kinases (Tyr or Ser/Thr) can phosphorylate DUSP3 by
increasing its phosphatase activity (in blue on the right) on its phospho-substrates
Cys-124, (yellow), and Arg-130 (red) are highlighted in the Fig. 2a. Its three-
dimensional (3D) structure has a catalytic slit of 6 Å depth, which is shallower
than the classical PTPs (9 Å) but is more open, which explains substrate differences
between these groups and also supports the possibility of other substrates mediating
DUSPs’ activities. This is the case for the ADUSP group, which is generally more
diverse in substrates than PTP classical group. Furthermore, it is noteworthy to
mention that if other ADUSPs share structural similarities and biological function-
alities with DUSP3, the task of dephosphorylating some substrates of DUSP3 could
also be played by these enzymes (Barford et al. 1994; Bayón and Alonso 2010;
Ishibashi et al. 1992).
Unlike most MKPs, DUSP3 is not regulated in response to MAPK activation
(Kang and Kim 2006), and although it does not have the CH2 structure at its
N-terminus, DUSP3 behaves as a true MKP capable of dephosphorylating
ERK1/2, JNK, and, to a lesser extent, p38 (Cerignoli et al. 2006; Hoyt et al. 2007;
Kondoh and Nishida 2007; Todd et al. 1999), which suggest the existence of an
CH2-independent catalytic mechanism. Also, while MKPs have their catalytic
activity increased by binding of their substrates, DUSP3 has an increase in activity
caused by VRK3, Zap70, and Tyk2. These enzymes phosphorylate DUSP3 at its
8 L. F. Monteiro et al.
Tyr-138 residue (shown in green on the Fig. 2) in cells that express them, and it has
been well established that mutations in this residue changes normal DUSP3 function
(Alonso et al. 2003; Hoyt et al. 2007; Kang and Kim 2006). The existence of this
regulatory site in DUSP3 is favorable with respect to the investigation of its potential
as an on-off therapeutic intervention (Bayón and Alonso 2010; Nunes-Xavier et al.
2011).
In addition to the MAPK substrates, the receptor tyrosine kinase ErbB, of
epidermal growth factor (EGF)-2, has been shown to have specific Tyr dephosphor-
ylation by DUSP3 in non-small cell lung cancers (Mustelin 2007). In immune cells
under stimulation by cytokines and growth factors, DUSP3 dephosphorylates the
signal transducer and activator of transcription 5 (STAT5), whose activity and
nuclear translocation are regulated by phosphorylation/dephosphorylation events
on two specific Tyr residues (Bayón and Alonso 2010).
The three classical MAPKs have been shown to be involved in the DNA damage
response and repair pathways activated by different types of genotoxic stress that
culminate in the regulation of ATM (a serine/threonine kinase) or phosphorylation of
the histone, H2AX isoform (Arnoldussen and Saatcioglu 2009; Jeffrey et al. 2007;
Lang et al. 2006; Salojin and Oravecz 2007). Therefore, novel roles for DUSP3 in
the maintenance of genomic stability have emerged, uncovering new biological
functions for this dual phosphatase. More recently, Forti (Forti 2015) and collabo-
rators (Panico and Forti 2013) have suggested several nuclear proteins, involved in
different aspects of the DNA damage response and repair, as putative protein
interactors that could be dephosphorylated by DUSP3, including nibrin (NBS1),
nucleophosmin (NPM), nucleolin (NUCL), heterogeneous nuclear ribonucleopro-
tein (hnRNP) C1/C2, and ATM/ATR (Alonso et al. 2004a, b; Jeong et al. 2006).
New relationships between protein targets and functions will be discussed in the next
sections proposing DUSP3 phosphatase as a good candidate for inhibition in clinical
settings (Fig. 3).
invasion of cancer cells (Henkens et al. 2008; Rahmouni et al. 2006). For instance,
Hela cells lacking DUSP3 arrest at the G1/S and G2/M transitions of cell cycle and
initiate senescence (Fig. 3) (Rahmouni et al. 2006; Stein et al. 1991).
MAPK activation has a growth-promoting role during the early G1 phase of the
cell cycle (Rahmouni et al. 2006). On the other hand, constitutively elevated ERK1/2
and JNK activity arrests cells in G2/M phase and G1/S-phase of the cell cycle,
respectively (Chau and Shibuya 1999; Rahmouni et al. 2006). Rahmouni et al. show
that the biphasic DUSP3 expression during cell cycle allows the regulation of the
MAPK activation (Rahmouni et al. 2006). They also demonstrated that low expres-
sion of DUSP3 in early G1 phase permits greater ERK1/2 and JNK activities, and the
continuously rising levels of DUSP3 during the S, G2, and M phase then prevent
MAPKs’ growth-arresting effects during these phases. This result shows the impor-
tance of DUSP3 in the G1 phase, in which there is well-known overexpression of
cyclin D1 (Baldin et al. 1993). This cyclin plays an important role in cell cycle
progression and is involved in many types of cancers (Sicinski et al. 1995; Tetsu and
McCormick 1999) such as breast cancer, in which cyclin D1 expression is
deregulated by several mechanisms (Ahnstrom et al. 2005; Sicinski and Weinberg
1997). One of these DUSP3-associated mechanisms occurs via overexpression of
breast cancer gene BRCA1-IRIS (a splice variant locus of the gene brca1) (ElShamy
and Livingston 2004), which induces cyclin D1 overexpression and increases cell
proliferation. BRCA1-IRIS can transcriptionally induce cyclin D1 expression
through binding and activation of the c-jun/AP1 transcription factor, which induces
its transcription in a DUSP3-independent manner (Nakuci et al. 2006). Moreover,
BRCA1-IRIS can also activate cyclin D1 expression in a breast cancer cell line in a
non-transcriptional fashion, in a DUSP3-dependent manner (Fig. 3) (Hao and
ElShamy 2007).
In the latter scenario, overexpression of the components of the axis BRCA1-
IRIS/EGFR/ErbB2 decreases the expression of DUSP3 in normal (human
mammary epithelial) and breast cancer cell lines (MCF-7 and SKBR3). This
DUSP3 downregulation generates JNK activation and subsequently activation of
c-Jun/AP1 and cyclin D1 overexpression that can transform normal mammary
epithelial cells, whereas overexpressing DUSP3 in breast cancer cells can block
the cyclin D1 overexpression. The downregulation of BRCA1-IRIS reduces the
expression of cyclin D1 by elevating the expression levels of DUSP3 mRNA and
protein, reducing the protein expression of EGFR and ErbB2 (Hao and ElShamy
2007). These results suggest that BRCA1-IRIS overexpression, and consequently
DUSP3 downregulation, play important roles in the development of more aggressive
endocrine-resistant breast cancer phenotypes that depend on the expression and
activation of growth factor or cell membrane receptors. In this regard, DUSP3
restoration might be sufficient to overcome either the cellular over-proliferation or
transformation dependent on Cyclin D1 overexpression (Hao and ElShamy 2007).
Following the same rationale, DUSP3 expression is lower in non-small cell lung
cancer (NSCLC) tissues in comparison with normal lung tissues. Contrary to other
studies, DUSP3 activity in NSCLC cells/tissues had very limited activity against
MAPKs but regulated ErBb and EGFR signaling that is involved in epithelial cell
10 L. F. Monteiro et al.
growth regulation (Wang et al. 2011). These tyrosine kinase receptors are related to
different diseases, including many types of cancer (Hynes and MacDonald 2009;
Hynes and Stern 1994; Nicholson et al. 2001). Wang et al. (2011) demonstrated the
capability of DUSP3 to dephosphorylate EGFR and ErB2 in cell lines. EGF stim-
ulation of DUSP3-expressing NSCLC cells led to a reduced Tyr-992 phosphoryla-
tion on EGFR mainly at the early times of treatment. DUSP3 downregulation
strongly increases phosphorylation at the EGFR Tyr-992 residue, with or without
EGFR stimulation. Therefore, DUSP3 overexpression reduces phosphorylation of
EGFR Tyr-992 and makes the cell less responsive to EGF. DUSP3 also inhibited the
downstream signaling events such as Tyr-783 phosphorylation in phospholipase C
(PLC)γ and protein kinase C (PKC) activation. On the other hand, Tyr-416 phos-
phorylation of Src activation is not affected when H1299 cells are stimulated with
EGF, therefore showing specificity of DUSP3 actions (Fig. 3).
In vitro DUSP3 overexpression could suppress cell proliferation in 2D and 3D
H1290 cell cultures. Xenograft DUSP3 overexpression also has been shown to
reduce tumor size. In patients with non-small cell lung cancer (NSCLC), DUSP3
mRNA and transcribed protein expression is significantly lower in tumor tissues than
in adjacent normal lung tissues. These data suggest that DUSP3 expression is also
particularly downregulated in these cancer cells and its overexpression can suppress
cancer cell growth. A decrease in DUSP3 expression might form part of the
pathogenic-initiating mechanisms of NSCLC. Wagner et al. (2013) tried to explain
this possible mechanism involved in DUSP3 downregulation and the onset of lung
tumorigenesis in NSCLC. They found that KDM2A, an H3 lysine 36 (H3K36)
demethylase (Shi et al. 2007), promotes lung tumorigenesis by epigenetically
enhancing ERK1/2 signaling through DUSP3 downregulation; this finding is in
contrast to those from Wang et al. (2011), who related the low DUSP3 expression
to EGFR pathway overactivation. These differences may probably be due to the
heterogeneous NSCLC molecular etiology (Herbst et al. 2008) and can also be
reinforced by the Lewis lung carcinoma experimental metastasis model, where
DUSP3/ mice developed a metastatic growth dependent on DUSP3 absence
through a mechanism of intense recruitment of macrophages to the lung metastasis
(Vandereyken et al. 2017a).
As many other genes are aberrantly expressed in tumors, dusp3 gene has been
shown overexpressed in some types of cancers. Several studies relate that
overexpression of DUSP3 can be associated with the onset or development of a
carcinogenic phenotype. However this cannot be considered a cause of such pheno-
type especially because other expression levels of other ADUSPs have not been
assessed. Even so, it is important to mention that in cervical cancer cell lines such as
HeLa, SiHa, CaSki, C33, and HT3 and other epithelial cells from the high-grade
squamous intraepithelial lesions (SIL), invasive squamous cell carcinomas, primary
cervical adenocarcinomas, and adenocarcinoma in situ of the uterine cervix, DUSP3
is upregulated when compared with primary keratinocytes and normal tissues. In
these cell lines, DUSP3 is located in the cytoplasm and nucleus, while in normal
keratinocytes, it is mainly cytoplasmic; conversely, in cervix cancer tissues, DUSP3
has a mainly nuclear localization (Henkens et al. 2008). In HeLa cells the
DUSP3/VHR Is a Potential Drug Target 11
distribution of DUSP3 is cell cycle phase dependent, mostly located in the nucleus of
interphasic cells; during the metaphase DUSP3 is concentrated around the chroma-
tin, and in telophase it is between the daughter chromatids (Rahmouni et al. 2006).
The increased amount of DUSP3 in cell lines and tissues of cervix cancer is not due
to increased expression of DUSP3 mRNA or stabilization of its mRNA, since its
levels in both cancer cell lines and normal keratinocytes are the same, but due to an
increase in DUSP3 protein half-life (posttranslational stabilization). Interestingly,
DUSP3’s half-life also varies in different phases of the cell cycle being lower in the
G1 phase (Henkens et al. 2008; Rahmouni et al. 2006). This dynamic expression and
differential distribution of DUSP3 in cervix cancers allow this phosphatase to
regulate ERK1/2 and JNK activity during the S and G2/M phases of the cell cycle;
excessive activity of these kinases can trigger cell cycle arrest in G1/S or
G2/M. Consequently, cervix cancer cells overexpressing DUSP3 can drive G1
progression despite an inhibitory signal or unfavorable environmental conditions
(Henkens et al. 2008).
As in epithelial cervix cancer cell lines, DUSP3 is overexpressed in prostate
cancer compared with normal prostate (Arnoldussen et al. 2008). In normal prostate
epithelial cells, androgen withdrawal leads to decreased cell proliferation, increased
apoptosis, and atrophy (Mercader et al. 2007). Conversely, tumor prostate
epithelial cells can overcome this deficiency and move to an androgen-independent
state (Feldman and Feldman 2001). When androgen-responsive prostate cancer
cell lines are treated with R1881 (a synthetic androgen), various MKPs are
overexpressed. Furthermore, apoptosis-inducing treatments with TPA (12-O-
tetradecanoylphorbol-13-acetate) or thapsigargin (TG) plus R1881 increase the
expression of these phosphatases. Among these overexpressed phosphatases,
DUSP3 mRNA and protein are significantly augmented. Furthermore, deregulation
of the MAPK pathway is also involved in prostate carcinogenesis (Bakin et al.
2003). For instance, in LNCaP cells (androgen-responsive prostate cancer cell line),
treatment with TPA/TG plus R1881 decreases JNK phosphorylation significantly
(Engedal et al. 2002) but did not affect phosphorylated ERK1/2 levels (Lorenzo and
Saatcioglu 2008), which seem to be caused by cellular stress, while their activation
has also been implicated in apoptosis (Wada and Penninger 2004).
These data suggest that overexpression of DUSP3 specifically inactivates JNK,
but not ERK1/2, what further suggests that DUSP3 can reverse apoptosis induced by
JNK in LNCaP cells when these cells are stimulated with androgens. In vivo,
androgen-sensitive grafts expressing DUSP3 have increased resistance to
castration-induced apoptosis, and tumor regression is inversely correlated to
DUSP3 expression (Shi 2007). This possible mechanism suggests a role for
DUSP3 in prostate cancer progression and also that DUSP3 knockdown in prostate
cancer may activate JNK, leading to apoptosis (Fig. 3). However and again, these
mechanisms were not explored considering the knockdown of other MKPs or
ADUSPs that could overlap their actions on the dephosphorylation of MAPKs.
DUSP3 is also overexpressed in dysplastic nevi (DNs, benign melanocytic
tumors, or lesions with a more hyper-proliferative phenotype); it is expressed in
the epidermis and DN nevus cells, but it is higher in the DN epidermis compared
12 L. F. Monteiro et al.
The roles of DUSP3 in cell cycle checkpoints and its relation to ERK1/2 have been
extensively studied. Although DUSP3 is not upregulated in response to MAPK
activation, it controls the cell cycle transition at G1/S and G2/M in an ERK1/2-
and JNK-dependent manner. DUSP3 knockdown in HeLa cells provokes cell cycle
arrest and senescence in human cancer cell (Rahmouni et al. 2006), associated with
an accelerated death of mitotically arrested cells (Tambe et al. 2016). However, an
unexpected action of the pair DUSP3-ERK1/2 that has been studied very recently is
the effect on genomic stability, which can be investigated by several aspects.
At the outset, inhibition of ERK1/2 has not resulted in defects in chromosomal
events, spindle assembly checkpoint signaling (Foley and Kapoor 2013), or mitotic
exit (Roberts et al. 2002; Shinohara et al. 2006), but instead, its overactivation
induced multipolar spindles and aneuploidy in cells (Eves et al. 2006). Because of
the strong relationship between ERK1/2 and DUSP3, the role of this phosphatase on
the formation of multipolar spindles in cancer cells was also investigated (Tambe
et al. 2016). In early mitotic mammalian cells, both DUSP3 and ERK1/2
(non-phosphorylated) are in the spindle apparatus (Rahmouni et al. 2006; Willard
DUSP3/VHR Is a Potential Drug Target 13
and Crouch 2001), and the transient inhibition of DUSP3 leads to the formation of
multipolar spindles in human mitotic cancer cells, which is reversed by the silencing
or chemical inhibition of ERK1/2 (Tambe et al. 2016). Additionally, the authors
showed that the ectopic overexpression of DUSP3 reduced ERK1/2 activity and
reversed the DUSP3 siRNA-induced multipolar spindles in HeLa cells (Tambe et al.
2016). Together, these data suggest that the relationship between ERK1/2 and
DUSP3 may have other roles in addition to cell cycle control and proliferation, as
demonstrated with respect to genomic stability (Fig. 3) (Tambe et al. 2016). But
these authors did not explore other relationships between MKPs or ADUSPs and
ERK1/2 regulating the spindle apparatus, suggesting a possible dependence on
substrates phosphorylated by MAPK and not directly to DUSP3 itself. In addition,
it has been shown that other DUSPs are also related to correct cell division. For
example, DUSP5 and DUSP7 are related with genomic stability because they are
connected to other proteins related to spindle formation and nuclear envelop
breakdown, respectively (Matta et al. 2007; Pfender et al. 2015).
Besides that, DUSP3 is extensively present in the nucleus of various cell lines
(Forti 2015; Henkens et al. 2008), especially after a genotoxic stress (Forti 2015).
DUSP3 co-localizes with the three MAPK isoforms (ERK1/2, JNK, p38) in the
nucleus upon gamma radiation-induced damage; it may also indicate that this
phosphatase may have other substrates and roles in genomic stability, indirectly or
completely not related to these kinases. Indeed, a study of our group, through a
computational approach and experimental validation analysis, identified novel
DUSP3 substrates specifically involved in genomic stability or integrity (Forti
2015). After DUSP3 has been found co-localized with phosphorylated H2AX
(S139), a classical marker of DNA strand breaks (that has also been observed for
the DUSP4) (Lawan et al. 2011), a deep bioinformatics analysis of human nuclear
proteins containing the Thr-X-Tyr motif (commonly present in the activation loop of
MAPKs), 121 putative DUSP3 substrates related to DNA damage and response and
repair were identified. Many of these data were experimentally validated in HeLa
cells exposed to ionizing radiation (gamma) sections; DUSP3 was found to
independently co-localize with pATM (S1981), pATR (S428), pBRCA1 (S1423),
BRCA2, centromere protein (CENP)-F, cyclin A, nibrin (NBS1), apurinic/
apyrimidinic endonuclease (APE1), the double-strand break repair protein
(MRE11), DNA repair protein (RAD50), checkpoint kinase (pCHK) 2 (T68), and
p53 (S15) proteins (Forti 2015). These data are consistent with the hypothesis that
DUSP3 could participate in DNA damage repair by interacting and/or
dephosphorylating these proteins with different timing following stress exposure.
According to these data, DUSP3 downregulation (by siRNA silencing or chemical
inhibition, as discussed later) provokes different effects in tumor cell lines under
genotoxic stress, including senescence, decreased cell proliferation/survival associ-
ated with an increased DNA damage, and an impaired and/or delayed DNA strand
break repair (Torres et al. 2017). However, another caveat of this study is that these
T-X-Y-containing substrates could also be dephosphorylated by other MKPs or
ADUSPs acting on MAPKs, and not only necessarily by DUSP3. With this in
mind, our group invested in another strategy, now using proteomics and
14 L. F. Monteiro et al.
Several experimental strategies have been employed by our group to access these
questions, and preliminary results point to a differential phosphorylation of specific
(not all) tyrosines on these proteins, which are dependent on DUSP3 expression
and/or activity (Fig. 3). Also, we are confirming that DUSP3-deficient cells showed
deficient repair of specific DNA lesions promoted by UV radiation, with direct
implications of the NER pathway (Forti et al. unpublished results). Once DUSP3
physically interacts with these three proteins (Panico and Forti 2013) and dephos-
phorylates them, thus increasing/decreasing their activity/function toward different
biological processes, we have now strong evidences that DUSP3 is a player in
genomic stability maintenance, through still unknown mechanisms but very likely
through these new protein partners or substrates (NUCL, hnRNP C1/C2, and NPM),
paving new ways for the drug development and clinical trials targeting DUSP3
independently of its actions on MAPK.
The first report about DUSP3’s role and expression in circulatory system cells were
in T lymphocytes (Ishibashi et al. 1992). In thymocytes, DUSP3 mRNA levels are
very low compared to brain and heart tissues, in which DUSP3 expression levels are
already low, and especially when compared to expression of other phosphatases such
as DUSP1, which is considered the most abundant phosphatase for many tissues
(Tanzola and Kersh 2006). However, T cells at resting state constitutively express
DUSP3 (Alonso et al. 2001; Ishibashi et al. 1992), and the activation of T cells does
not induce the expression of this enzyme that primarily reduces ERK1/2 phosphor-
ylation, but this inhibitory effect is more marked to JNK phosphorylation (Alonso
et al. 2001). On the other hand, p38 kinase is also activated in T cells in response to
receptor activation and UV radiation, but the co-expressions of active and inactive
DUSP3 had no significant impact on the regulation of the p38 kinase pathway
(Alonso et al. 2001).
Resting T cells express the hematopoietic protein tyrosine phosphatase (HePTP)
and DUSP3; after TCR stimulation additional MAPK phosphatases are synthesized.
However, compared to VHX and MKP6, two other small dual-specific phosphatase
expressed in T cells, DUSP3 was the most efficient to reduce the TCR-induced
activation of an NFAT-AP-1-driven reporter when the three proteins were expressed
at equal amounts; therefore the phosphatase is more potent to inhibit the TCR
signaling to IL-2 (Alonso et al. 2001). When ectopically overexpressed into Jurkat
cells by transfection, DUSP3 inhibits ERK1/2 activation in response to T cell antigen
receptor (TCR) activation by IL-2 (Alonso et al. 2001), and this effect is dependent
on direct DUSP3 phosphorylation by zeta-chain-associated protein (ZAP)-70 on the
Tyr138 residue (Alonso et al. 2003). The ZAP-70 tyrosine kinase is a key component
of the TCR signaling pathway in which the DUSP3-Y138F mutant was shown to
increase TCR-induced ERK1/2 activation and the induction of the IL-2 gene expres-
sion. Basal activity of ZAP-70 has been shown to be sufficient to stimulate T cell
16 L. F. Monteiro et al.
functions, but when there is TCR stimulation by an antigen, this becomes even more
evident and causes DUSP3 translocation from the cytosol to the T cell pole, thereby
facilitating the phosphorylation of DUSP3 on the Tyr138 residue (Alonso et al.
2003).
DUSP3 also selectively dephosphorylates IFN-β-induced signal transducers and
activators of transcription 5 (STAT5), which leads to subsequent inhibition of
transcription factor activities. In this context, STAT5’s Src homology 2 domain
(SH2) was required for effective dephosphorylation by DUSP3 since the recruitment
of the phosphatase activity on STAT5 requires DUSP3 phosphorylation on the
Tyr138 residue. Tyrosine kinase 2 (TYK2), which mediates STAT5 phosphoryla-
tion, has also been shown to be responsible for DUSP3 phosphorylation on the
Tyr138 residue very likely due to the high homology between this kinase and SYK
and also ZAP-70 (Fig. 2) (Hoyt et al. 2007).
In 2010, a data collection extracted from the RefDIC (Reference Genomics
Database of Immune Cell) database about the expression pattern of genes encoding
PTPs in mice was published. This work showed that DUSP3 expression was
detected mainly in macrophages but also in immature dendritic cells, mast cells,
and neutrophils although these data were all related to the DUSP3 basal expression
in cells that did not receive any type of physical or chemical treatment or differen-
tiation stimulus (Arimura and Yagi 2010). At the level of proteins, monocytes and
macrophages demonstrated higher DUSP3 level expression when compared to
neutrophils and B and T cells (Singh et al. 2015), in addition to platelets, which
also presented DUSP3 levels significantly higher than B and T lymphocytes
(Musumeci et al. 2014). These data bring to light the importance of research about
the role of DUSP3 in circulatory system cells since these cells have different
expression patterns from each other in addition to different physiological functions
(Fig. 3).
In another study, DUSP3 activity was identified as being relevant in the inflam-
matory response to S. aureus, both in humans and mice, through the NF-kB
signaling as a negative feedback component. The cytokine inflammatory response
was improved after DUSP3 knockdown in macrophages as observed by an increase
in pro-inflammatory cytokine production via NF-kB. This cytokine increase may
lead to a hyper-responsiveness of the host’s immune system, leading to the so-called
immune paralysis (Yan et al. 2014). In contrast to its pro-inflammatory properties, it
has been recently shown that DUSP3 deficiency in mice is capable of promoting
tolerance to LPS-induced endotoxin shock and polymicrobial septic shock, which is
mainly dependent on macrophages. This protection is associated with an expressive
increase of anti-inflammatory M2-like macrophages, decreased TNF production, and
impaired ERK1/2 activation. In in vivo experiments, it was observed that after
LPS-induced endotoxic shock, the DUSP3/ mice recovered their normal
temperature, while the DUSP3+/+ mice remained hypothermic. Finally, it was also
demonstrated that resistance to septic shock was transferable through monocytes to
wild-type (WT) mice and was associated with M2-like macrophage dominance
(Singh et al. 2015). Resistance to sepsis in females, but not in males or ovariecto-
mized females (OVX), was associated with decreased ERK1/2, phosphoinositide
DUSP3/VHR Is a Potential Drug Target 17
kinase (PI3K), and protein kinase B (AKT) activation (Vandereyken et al. 2017b).
DUSP3 expression in peritoneal macrophages was abolished in the mice that
received DUSP3/ bone marrow cells, and after LPS challenge 70% of the female
mice that received DUSP3/ bone marrow cells survived until the end of the
experiment, compared to 9% of the female mice that received DUSP3+/+ bone
marrow cells and male mice that received DUSP3/ and DUSP3+/+ cells that
died within 4 days. These data demonstrate that in the absence of DUSP3, female
sex hormones are involved in the observed resistance of DUSP3/ mice to
LPS-induced lethality, thus suggesting that DUSP3 inhibition combined with estro-
gen administration may lead to protection against septic shock (Vandereyken et al.
2017b).
Another recent study has also elucidated the importance of DUSP3 in platelet
aggregation mechanisms, despite previous researches implicating other non-DUSP
phosphatases in platelet signaling such as CD148, PTP1B, SHP1, and SHP2
(Musumeci et al. 2014; Tautz et al. 2015). This phosphatase seems to be implicated
in platelet signaling through collagen receptor glycoprotein VI (GPVI) and C-type
lectin-like type II (CLECII) receptors to promote reduction of SYK and phospholi-
pase C γ2 (PLCγ2) tyrosine phosphorylation, without affecting the overall tyrosine
phosphorylation (such as ERK1/2 and JNK). In DUSP3 absence, thrombus forma-
tion was significantly impaired without affecting bleeding, suggesting that this
enzyme plays a key role in arterial thrombosis but is not necessary for primary
hemostasis. Ex vivo experiments further demonstrated that DUSP3 deficiency
resulted in defective platelet aggregation, granule secretion, and αIIbβ3 integrin
activation in response to GPVI and CLEC-2 receptor stimulation without affecting
GPCR-mediated platelet activation (such as the purinergic [P2Y12], thromboxane
[TXA2], and thrombin receptors) (Musumeci et al. 2014; Tautz et al. 2015). The
pharmacotherapy currently employed for thrombosis has been mainly based on
G-protein-coupled receptor (GPCR) inhibition and the formation of molecules that
can stimulate these receptors. Although currently available and commonly used
platelet precursors increase patient survival, leading to decreased mortality and
morbidity, these also produce other adverse effects such as increased risk of gastro-
intestinal toxicity, neutropenia, thrombocytopenia, and bleeding in combination with
an increased incidence of arterial thrombosis cases. Because of these and other
reasons, DUSP3 constitutes a potential target for the development of new, safer
therapies in platelet aggregation (Tautz et al. 2015) and, here, through MAPK-
independent mechanisms.
(continued)
19
Table 1 (continued)
20
Ref# Compound name Structure IC50 (DUSP3) Activity toward other phosphatasesa
3 Stevastelin A 2.7 μM (Hamaguchi et al. –
1997)
(continued)
Table 1 (continued)
22
Ref# Compound name Structure IC50 (DUSP3) Activity toward other phosphatasesa
11 GATPT 2.92 μM (Shi et al. 2007) N.A.
2006)
(continued)
Table 1 (continued)
24
Ref# Compound name Structure IC50 (DUSP3) Activity toward other phosphatasesa
18 SA4 78 nM (Wu et al. 2009) MKP1, CD45
a
Within one order of magnitude of DUSP3 IC50 reported value
L. F. Monteiro et al.
DUSP3/VHR Is a Potential Drug Target 25
open to studies, and the use of metal ions in a clinical setting is most likely not
feasible due to the nonspecific nature of the redox reactions.
Other groups have searched for natural or synthetic organic compounds with
DUSP3 inhibitory activity. In 1995, Hamaguchi et al. (Hamaguchi et al. 1995)
isolated an organic acid, RK-682, from a Streptomyces strain via monitoring
DUSP3 inhibitory activity (Table 1, compound 1). The authors identified cellular
phosphotyrosine level enhancement, which was different than the pattern caused by
orthovanadate. The group also evaluated its effect on the human B cell leukemia
(Ball-1) cell cycle, observing an arrest at G1/S in accordance with DUSP3’s loss of
function. They calculated the IC50 to be 2.0 μM, and RK-682 has been since then
considered a classic DUSPase inhibitor with no inhibitory potential against PPases
but also displaying promiscuous binding to other PTPases (Carneiro et al. 2015).
The lactone moiety of this compound is thought to interact with the active site of the
phosphatase, thereby inhibiting its activity (Hirai et al. 2011).
Maintaining the long hydrophobic side chain of RK-682, which was found to be
critical for DUSP3 binding, the group later developed enamine derivatives to
improve permeability and enable in vivo action. One of the derivatives containing
m-methylbenzylamine substitution on the enamine group showed improved perme-
ability and high selectivity toward DUSP3, displaying neglectable IC50 values
toward other selected phosphatases. This compound, called RK-682 enamine 2c
(Table 1, compound 2), at a concentration of 30 μM, caused an increase of pERK
and pJNK and led to an arrest at the G1/S transition in mouse fibroblast cells
(NIH3T3 cells). According to 3D models, the enhanced selectivity of this drug is
due to the interaction of the m-methylbenzylamine group with the margins of the
active pocket (Hirai et al. 2011).
The same group identified and synthesized metabolites from a Penicillium strain.
This class of compounds has been called stevastelins and is composed of valine,
threonine, serine, and a 3,5-dihydroxy-2,4-dimethylstearic acid moiety (Table 1,
compounds 3 and 4). In 1996, they observed that some of these compounds’
derivatives had inhibitory potential against DUSP3 and that the long-chain aliphatic
hydrocarbon group was necessary for inhibition (Hamaguchi et al. 1997). While
stevastelin B had a strong immunosuppressive effect in situ for Jurkat cells, it did not
display in vitro inhibition of DUSP3. Conversely, stevastelin A, a sulfonylated
derivative of stevastelin B, had a strong inhibitory effect on DUSP3 but not CD45
(a PTPase) or PPases. The authors hypothesized that stevastelin B might be
sulfonylated or phosphorylated after incorporation into the cell, turning it into a
biologically active form; in contrast, stevastelin A is not diffusible through the
membrane and is therefore not active in vivo. The effect of stevastelins A and B
on the cell cycle of tsFTZ10 cell lines (mammary carcinoma cdc2 mutant) was also
evaluated, and upon release of a G2 phase arrest, the cells treated with stevastelin B
had an inhibited cell cycle transition (much like those treated with vanadate), while
stevastelin A had no visible effects. However, it would be interesting to investigate
whether DUSP3’s loss of function causes an inhibition of the G2/M transition or
whether stevastelin B causes an arrest at G1/S, which would indicate selective
DUSP3 inhibition (Hamaguchi et al. 1997).
26 L. F. Monteiro et al.
Our group has also assayed the in vivo effects of low doses of GATPT (250 nM)
in HeLa (cervix carcinoma) and MeWo (metastatic melanoma) cells and observed
higher accumulation of DNA damage and higher pERK1/2 and pH2AX levels
following gamma radiation exposure. Also, a downregulation of homologous
recombination (HR) and nonhomologous end joining (NHEJ) pathways of DNA
repair, following pretreatment with GATPT and subsequent gamma radiation of
HeLa cells, was assessed by the use of EJ5-GFP and DR-GFP constructs (Bennardo
et al. 2008). The accumulation of DNA damage, as assessed by the alkaline comet
assay, was also seen in the MRC-5 fibroblast cell line and in the ATM-deficient cell
lines (AT5-BIVA) (Torres et al. 2017), showing the impact of DUSP3 inactivation
on genomic instability.
Park et al. (2008) assayed 194 compounds that scored high on docking simula-
tions against DUSP3 and found 17 molecules that displayed at least 50% inhibition
at 50 μM concentration. The two most potent inhibitors shared a common
[2-(2,5-dioxoimidazolidin-4-ylidenemethyl)-pyrrol-1-yl]-benzoic acid scaffold and
showed IC50 values around 4 μM (Table 1, compound 12 and 13). The authors
explored the features of the lowest-energy conformation of these two compounds,
which are similar to the cases in which the benzoate group points toward the catalytic
residue and is stabilized by four hydrogen bonds involving Arg125 and Arg130.
However, they differ in that the imidazolidine-2,4-dione group is accommodated in a
small binding pocket for the first molecule but is exposed to bulk solvent for the
second one, most likely due to bulky chlorophenyl group substitution.
In 2009, Park et al. (Park et al. 2009) assessed NSC-87877 inhibitory activity
(Table 1, compound 14), a known SHP-2 inhibitor, against DUSP3. They derived
an IC50 value of 7.83 0.98 μM and deduced that this quinolinesulfonic acid acts as
a competitive inhibitor because of binding in the catalytic site. They also performed
in vitro experiments with active, phosphorylated ERK in the presence of DUSP3 and
increasing concentrations of NSC-87877, and they observed dose-dependent inhi-
bition of DUSP3 activity toward ERK. In vivo experiments with EGF-treated HEK
293 cells and transfected with a FLAG-DUSP3 expression plasmid also demon-
strated a rise in phospho-ERK levels as a result of increasing concentrations of the
inhibitor. This compound, however, shows inhibitory potential for several other
PTPs, therefore being considered very nonspecific.
In the same year, a group of researchers at the Burnham Institute (La Jolla, USA)
and Liège University (Liège, Belgium) developed multidentate small molecule
inhibitors with potent and selective activity over DUSP3 (Wu et al. 2009). Starting
from screening of thousands of drug-like molecules, they picked the most active hit,
2-((Z)-4-oxo-5-((E)-3-phenylallylidene)-2-thioxothiazolidin-3-yl)ethanesulfonic
acid. They subsequently searched for analogs of this compound, keeping the oxo-
thioxothiazolidinyl-ethanesulfonic acid moiety as the hydrophilic target for the
catalytic cleft and searching for additional hydrophobic regions to stabilize the
docking into the active site. Five of these structures displayed inhibitory activity
on the nanomolar scale and were at least ten times more potent for DUSP3 than other
PTPs (Table 1, compound 15–19). They also assayed the in vivo performance of
these compounds, finding SA3 to be the most active, as a result of displaying strong
28 L. F. Monteiro et al.
antiproliferative effect over CaSki and HeLa cells and reduction in cell growth and
thymidine incorporation without inducing apoptosis. However, the compounds did
not significantly hamper normal keratinocyte cell growth, indicating they would be
good candidates for cancer treatment. SA3 was also co-crystallized with DUSP3,
showing clear electron density in the active site for the oxo-thioxothiazolidinyl-
ethanesulfonic acid moiety, displaying multiple hydrogen bond interactions within
the P-loop and a salt bridge with the guanidinium group of Arg130 (Wu et al. 2009).
Our group has also assessed in vivo effects of SA3 (Torres et al. 2017); by using
concentrations of 20 μM in HeLa and MeWo cells, we observed an increase in
sensitivity to gamma radiation as assessed by growth curves and senescence-
associated staining. Similar to the results obtained with GATPT pretreatments, we
also saw an increase in DNA damage accumulation after ionizing radiation expo-
sure, as assessed by comet assays, and an increase in pH2AX and pERK1/2 levels.
Also, there was a significant (~50%) reduction in nonhomologous end joining and
homologous recombination pathways of DNA strand break repair in HeLa cells
pretreated with SA3 and following exposure to gamma radiation regimes (Torres
et al. 2017).
The ability that the aforementioned DUSP3/VHR inhibitors present in disrupting
DUSP3-mediated biological effects emphasizes the interest in finding better, safer,
and specific inhibitors for this somewhat elusive enzyme that belongs to the DUSP
subfamily. Obviously, all the published work is just the beginning, and much
remains to be evaluated, namely, the selectivity of these drugs against other
ADUSPs, potential side effects on other key metabolic enzymes, and the availability
of more potent compounds. However, much of these issues would be addressed in
clinical trials, and the properties of these molecules offer a great starting point for the
development of better drugs.
5 Conclusions
DUSP3/VHR-knockout mice have been generated for functional studies of this dual
phosphatase in vivo, and no characteristic disease-associated phenotypes were
observed in these models. However, when these animals or cells derived from
them were challenged with different stimuli or stress, this enzyme deficiency
revealed very promising phenotypes (Amand et al. 2014; Singh et al. 2015;
Vandereyken et al. 2017b). For example, in terms of cancer biology, the majority
of the published reports discussed in this review shows opposite functions for
DUSP3. Whereas DUSP3 loss of function inhibits proliferation and invasiveness
of cancer cell lines in accordance with the oncogene addiction model (Weinstein and
Joe 2008), thus defining this phosphatase as an oncogene that regulates the growth
and survival of cancer cells, in other cases the gain of function of DUSP3 could
generate higher proliferation, migration, and invasiveness capacity of cells; therefore
DUSP3 could be considered a tumor suppressor protein. In addition, the well-
demonstrated effects of DUSP3 on the control of angiogenesis and neoangiogenesis
DUSP3/VHR Is a Potential Drug Target 29
Acknowledgments This project was supported by FAPESP (Grants # 2008/58264-5 and # 2015/
03983-0) and CNPq (Grant # 402230/2016-7) to FLF, head of the Laboratory of Signaling in
Biomolecular Systems (LSBS). LCR is a senior postdoctoral fellow from the CAPES-PNPD
program at the Institute of Chemistry, University of Sao Paulo. JOF is a PhD student fellow of
Fapesp (# 2017/16491-4), and PYFM is a master’s fellow of CNPq, both enrolled at the
postgraduation program in Biochemistry and Molecular Biology, Institute of Chemistry, University
of Sao Paulo. LFM is a master’s fellow of CAPES at the Biotechnology program, also in University
of Sao Paulo. All authors thank BO and JRD for technical assistance in the LSBS laboratory.
References
Ahnstrom M, Nordenskjold B, Rutqvist LE, Skoog L, Stal O (2005) Role of cyclin D1 in ErbB2-
positive breast cancer and tamoxifen resistance. Breast Cancer Res Treat 91:145–151
Alonso A, Saxena M, Williams S, Mustelin T (2001) Inhibitory role for dual specificity phosphatase
VHR in T cell antigen receptor and CD28-induced Erk and Jnk activation. J Biol Chem
276:4766–4771
30 L. F. Monteiro et al.
Chen YR, Chou HC, Yang CH, Chen HY, Liu YW, Lin TY, Yeh CL, Chao WT, Tsou HH, Chuang
HC, Tan TH (2017) Deficiency in VHR/DUSP3, a suppressor of focal adhesion kinase, reveals
its role in regulating cell adhesion and migration. Oncogene 36:6509–6517
Cho SSL, Han J, James SJ, Png CW, Weerasooriya M, Alonso S, Zhang Y (2017) Dual-specificity
phosphatase 12 targets p38 MAP kinase to regulate macrophage response to intracellular
bacterial infection. Front Immunol 8:1259–1269
Christian KJ, Lang MA, Raffalli-Mathieu F (2008) Interaction of heterogeneous nuclear ribonu-
cleoprotein C1/C2 with a novel cis-regulatory element within p53 mRNA as a response to
cytostatic drug treatment. Mol Pharmacol 73:1558–1567
Chung HS, Wang SB, Venkatraman V, Murray CI, Van Eyk JE (2013) Cysteine oxidative
posttranslational modifications: emerging regulation in the cardiovascular system. Circ Res
112:382–392
Daniely Y, Dimitrova DD, Borowiec JA (2002) Stress-dependent nucleolin mobilization mediated
by p53-nucleolin complex formation. Mol Cell Biol 22:6014–6022
Denu JM, Dixon JE (1995) A catalytic mechanism for the dual-specific phosphatases. Proc Natl
Acad Sci U S A 92:5910–5914
Denu JM, Dixon JE (1998) Protein tyrosine phosphatases: mechanisms of catalysis and regulation.
Curr Opin Chem Biol 2:633–641
ElShamy WM, Livingston DM (2004) Identification of BRCA1-IRIS, a BRCA1 locus product. Nat
Cell Biol 6:954–967
Engedal N, Korkmaz CG, Saatcioglu F (2002) C-Jun N-terminal kinase is required for phorbol
ester- and thapsigargin-induced apoptosis in the androgen responsive prostate cancer cell line
LNCaP. Oncogene 21:1017–1027
Eves EM, Shapiro P, Naik K, Klein UR, Trakul N, Rosner MR (2006) Raf kinase inhibitory protein
regulates aurora B kinase and the spindle checkpoint. Mol Cell 23:561–574
Farooq A, Zhou MM (2004) Structure and regulation of MAPK phosphatases. Cell Signal
16:769–779
Feldman BJ, Feldman D (2001) The development of androgen-independent prostate cancer. Nat
Rev Cancer 1:34–45
Foley EA, Kapoor TM (2013) Microtubule attachment and spindle assembly checkpoint signalling
at the kinetochore. Nat Rev Mol Cell Biol 14:25–37
Forti FL (2015) Combined experimental and bioinformatics analysis for the prediction and identi-
fication of VHR/DUSP3 nuclear targets related to DNA damage and repair. Integr Biol (Camb)
7:73–89
Goldstein M, Derheimer FA, Tait-Mulder J, Kastan MB (2013) Nucleolin mediates nucleosome
disruption critical for DNA double-strand break repair. Proc Natl Acad Sci U S A
110:16874–16879
Grisendi S, Mecucci C, Falini B, Pandolfi PP (2006) Nucleophosmin and cancer. Nat Rev Cancer
6:493–505
Hamaguchi T, Sudo T, Osada H (1995) RK-682, a potent inhibitor of tyrosine phosphatase, arrested
the mammalian cell cycle progression at G1phase. FEBS Lett 372:54–58
Hamaguchi T, Masuda A, Morino T, Osada H (1997) Stevastelins, a novel group of immunosup-
pressants, inhibit dual-specificity protein phosphatases. Chem Biol 4:279–286
Hao L, ElShamy WM (2007) BRCA1-IRIS activates cyclin D1 expression in breast cancer cells by
downregulating the JNK phosphatase DUSP3/VHR. Int J Cancer 121:39–46
He R, Zeng LF, He Y, Zhang S, Zhang ZY (2013) Small molecule tools for functional interrogation
of protein tyrosine phosphatases. FEBS J 280:731–750
Hendriks WJ, Elson A, Harroch S, Pulido R, Stoker A, den Hertog J (2013) Protein tyrosine
phosphatases in health and disease. FEBS J 280:708–730
Henkens R, Delvenne P, Arafa M, Moutschen M, Zeddou M, Tautz L, Boniver J, Mustelin T,
Rahmouni S (2008) Cervix carcinoma is associated with an up-regulation and nuclear localiza-
tion of the dual-specificity protein phosphatase VHR. BMC Cancer 8:147
Herbst RS, Heymach JV, Lippman SM (2008) Lung cancer. N Engl J Med 359:1367–1380
32 L. F. Monteiro et al.
Lin CY, Tan BC, Liu H, Shih CJ, Chien KY, Lin CL, Yung BY (2010) Dephosphorylation of
nucleophosmin by PP1beta facilitates pRB binding and consequent E2F1-dependent DNA
repair. Mol Biol Cell 21:4409–4417
Lorenzo PI, Saatcioglu F (2008) Inhibition of apoptosis in prostate cancer cells by androgens is
mediated through downregulation of c-Jun N-terminal kinase activation. Neoplasia 10:418–428
Mandl M, Slack DN, Keyse SM (2005) Specific inactivation and nuclear anchoring of extracellular
signal-regulated kinase 2 by the inducible dual-specificity protein phosphatase DUSP5. Mol
Cell Biol 25:1830–1845
Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S (2002) The protein kinase comple-
ment of the human genome. Science 298:1912–1934
Matta H, Surabhi RM, Zhao J, Punj V, Sun Q, Schamus S, Mazzacurati L, Chaudhary PM (2007)
Induction of spindle cell morphology in human vascular endothelial cells by human herpesvirus
8-encoded viral FLICE inhibitory protein K13. Oncogene 26:1656–1660
Mercader M, Sengupta S, Bodner BK, Manecke RG, Cosar EF, Moser MT, Ballman KV, Wojcik
EM, Kwon ED (2007) Early effects of pharmacological androgen deprivation in human prostate
cancer. BJU Int 99:60–67
Miranda GA, Chokler I, Aguilera RJ (1995) The murine nucleolin protein is an inducible DNA and
ATP binding protein which is readily detected in nuclear extracts of lipopolysaccharide-treated
splenocytes. Exp Cell Res 217:294–308
Mitsui H, Kiecker F, Shemer A, Cannizzaro MV, Wang CQF, Gulati N, Ohmatsu H, Shah KR,
Gilleaudeau P, Sullivan-Whalen M, Cueto I, McNutt NS, Suarez-Farinas M, Krueger JG (2016)
Discrimination of dysplastic nevi from common melanocytic nevi by cellular and molecular
criteria. J Invest Dermatol 136:2030–2040
Mustelin T (2007) A brief introduction to the protein phosphatase families. In: Moorhead G
(ed) Protein phosphatase protocols, vol 365. Springer, Totowa, pp 9–122
Musumeci L, Kuijpers MJ, Gilio K, Hego A, Theatre E, Maurissen L, Vandereyken M, Diogo CV,
Lecut C, Guilmain W, Bobkova EV, Eble JA, Dahl R, Drion P, Rascon J, Mostofi Y, Yuan H,
Sergienko E, Chung TD, Thiry M, Senis Y, Moutschen M, Mustelin T, Lancellotti P,
Heemskerk JW, Tautz L, Oury C, Rahmouni S (2014) Dual-specificity phosphatase 3 deficiency
or inhibition limits platelet activation and arterial thrombosis. Circulation 131:656–668
Nakuci E, Mahner S, Direnzo J, ElShamy WM (2006) BRCA1-IRIS regulates cyclin D1 expression
in breast cancer cells. Exp Cell Res 312:3120–3131
Nicholson RI, Gee JM, Harper ME (2001) EGFR and cancer prognosis. Eur J Cancer 37(Suppl 4):
S9–S15
Nunes-Xavier C, Roma-Mateo C, Rios P, Tarrega C, Cejudo-Marin R, Tabernero L, Pulido R
(2011) Dual-specificity MAP kinase phosphatases as targets of cancer treatment. Anti Cancer
Agents Med Chem 11:109–132
Olson MO, Dundr M, Szebeni A (2000) The nucleolus: an old factory with unexpected capabilities.
Trends Cell Biol 10:189–196
Panico K, Forti FL (2013) Proteomic, cellular, and network analyses reveal new DUSP3 interac-
tions with nucleolar proteins in HeLa cells. J Proteome Res 12:5851–5866
Park J, Fu H, Pei D (2004) Peptidyl aldehydes as slow-binding inhibitors of dual-specificity
phosphatases. Bioorg Med Chem Lett 14:685–687
Park H, Jung SK, Jeong DG, Ryu SE, Kim SJ (2008) Discovery of VHR phosphatase inhibitors
with micromolar activity based on structure-based virtual screening. ChemMedChem
3:877–880
Park S-J, Song M-A, Cho S-Y (2009) Regulation of vaccinia H1-related (VHR) phosphatase
activity by NSC-87877. Bull Kor Chem Soc 30:3098–3100
Patterson KI, Brummer T, O’Brien PM, Daly RJ (2009) Dual-specificity phosphatases: critical
regulators with diverse cellular targets. Biochem J 418:475–489
Pfender S, Kuznetsov V, Pasternak M, Tischer T, Santhanam B, Schuh M (2015) Live imaging
RNAi screen reveals genes essential for meiosis in mammalian oocytes. Nature 524:239–242
Pinol-Roma S (1999) Association of nonribosomal nucleolar proteins in ribonucleoprotein
complexes during interphase and mitosis. Mol Biol Cell 10:77–90
34 L. F. Monteiro et al.
Tonks NK (2013) Protein tyrosine phosphatases – from housekeeping enzymes to master regulators
of signal transduction. FEBS J 280:346–378
Torres TE, Russo LC, Santos A, Marques GR, Magalhaes YT, Tabassum S, Forti FL (2017) Loss of
DUSP3 activity radiosensitizes human tumor cell lines via attenuation of DNA repair pathways.
Biochim Biophys Acta 1861:1879–1894
Ueda K, Usui T, Nakayama H, Ueki M, Takio K, Ubukata M, Osada H (2002) 4-isoavenaciolide
covalently binds and inhibits VHR, a dual-specificity phosphatase. FEBS Lett 525:48–52
Vandereyken M, Jacques S, Van Overmeire E, Amand M, Rocks N, Delierneux C, Singh P,
Singh M, Ghuysen C, Wathieu C, Zurashvili T, Sounni NE, Moutschen M, Gilles C, Oury C,
Cataldo D, Van Ginderachter JA, Rahmouni S (2017a) Dusp3 deletion in mice promotes
experimental lung tumour metastasis in a macrophage dependent manner. PLoS One 12:
e0185786
Vandereyken MM, Singh P, Wathieu CP, Jacques S, Zurashvilli T, Dejager L, Amand M,
Musumeci L, Singh M, Moutschen MP, Libert CRF, Rahmouni S (2017b) Dual-specificity
phosphatase 3 deletion protects female, but not male, mice from endotoxemia-induced and
polymicrobial-induced septic shock. J Immunol 199:2515–2527
Vang T, Miletic AV, Arimura Y, Tautz L, Rickert RC, Mustelin T (2008) Protein tyrosine
phosphatases in autoimmunity. Annu Rev Immunol 26:29–55
Wada T, Penninger JM (2004) Mitogen-activated protein kinases in apoptosis regulation. Oncogene
23:2838–2849
Wagner KW, Alam H, Dhar SS, Giri U, Li N, Wei Y, Giri D, Cascone T, Kim JH, Ye Y, Multani
AS, Chan CH, Erez B, Saigal B, Chung J, Lin HK, Wu X, Hung MC, Heymach JV, Lee MG
(2013) KDM2A promotes lung tumorigenesis by epigenetically enhancing ERK1/2 signaling.
J Clin Invest 123:5231–5246
Wang JY, Yeh CL, Chou HC, Yang CH, Fu YN, Chen YT, Cheng HW, Huang CY, Liu HP, Huang
SF, Chen YR (2011) Vaccinia H1-related phosphatase is a phosphatase of ErbB receptors and is
down-regulated in non-small cell lung cancer. J Biol Chem 286:10177–10184
Weinstein IB, Joe A (2008) Oncogene addiction. Cancer Res 68:3077–3080 discussion 3080
Willard FS, Crouch MF (2001) MEK, ERK, and p90RSK are present on mitotic tubulin in Swiss
3T3 cells: a role for the MAP kinase pathway in regulating mitotic exit. Cell Signal 13:653–664
Wu MH, Chang JH, Chou CC, Yung BY (2002a) Involvement of nucleophosmin/B23 in the
response of HeLa cells to UV irradiation. Int J Cancer 97:297–305
Wu MH, Chang JH, Yung BY (2002b) Resistance to UV-induced cell-killing in nucleophosmin/
B23 over-expressed NIH 3T3 fibroblasts: enhancement of DNA repair and up-regulation of
PCNA in association with nucleophosmin/B23 over-expression. Carcinogenesis 23:93–100
Wu S, Vossius S, Rahmouni S, Miletic AV, Vang T, Vazquez-Rodriguez J, Cerignoli F, Arimura Y,
Williams S, Hayes T, Moutschen M, Vasile S, Pellecchia M, Mustelin T, Tautz L (2009)
Multidentate small-molecule inhibitors of vaccinia H1-related (VHR) phosphatase decrease
proliferation of cervix cancer cells. J Med Chem 52:6716–6723
Yan Q, Sharma-Kuinkel BK, Deshmukh H, Tsalik EL, Cyr DD, Lucas J, Woods CW, Scott WK,
Sempowski GD, Thaden JT, Rude TH, Ahn SH, Fowler VG Jr (2014) Dusp3 and Psme3 are
associated with murine susceptibility to Staphylococcus aureus infection and human sepsis.
PLoS Pathog 10:e1004149
Yang C, Maiguel DA, Carrier F (2002) Identification of nucleolin and nucleophosmin as genotoxic
stress-responsive RNA-binding proteins. Nucleic Acids Res 30:2251–2260
Yano K, Morotomi-Yano K, Wang SY, Uematsu N, Lee KJ, Asaithamby A, Weterings E, Chen DJ
(2008) Ku recruits XLF to DNA double-strand breaks. EMBO Rep 9:91–96
Yuvaniyama J, Denu JM, Dixon JE, Saper MA (1996) Crystal structure of the dual specificity
protein phosphatase VHR. Science 272:1328–1331
Zhou B, Wang ZX, Zhao Y, Brautigan DL, Zhang ZY (2002) The specificity of extracellular signal-
regulated kinase 2 dephosphorylation by protein phosphatases. J Biol Chem 277:31818–31825
Rev Physiol Biochem Pharmacol (2019) 176: 37–64
DOI: 10.1007/112_2018_13
© Springer Nature Switzerland AG 2018
Published online: 5 December 2018
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2 Pathways of Sodium Influx in Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.1 Na+/K+-ATPase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.2 Glutamate Receptor Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.3 Acid-Sensing Ion Channels (ASICs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.4 Na+/H+ Exchanger (NHE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.5 Electrogenic Na+/HCO3 Co-transporter 1 (NBCe1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.6 Transient Receptor Potential (TRP) Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3 Pathways of Chloride Influx in Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.1 Na+-K+-2Cl Co-transporter 1 (NKCC1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4 Pathways of Potassium Efflux in Stroke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.1 KCNQ (Kv7) K+ Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2 ATP-Dependent K+ Channels (KATP Channels) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
5 Aquaporins (AQPs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6 Mitochondrial Dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7 Key Challenges and Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
K. Y. Loh
Duke-NUS Medical School, Singapore, Singapore
Calcium Signalling Laboratory, National Neuroscience Institute, Singapore, Singapore
e-mail: e0011140@u.duke.nus.edu
Z. Wang
Department of Urology, National University Hospital, Singapore, Singapore
e-mail: ziting_wang@nuhs.edu.sg
P. Liao (*)
Duke-NUS Medical School, Singapore, Singapore
Calcium Signalling Laboratory, National Neuroscience Institute, Singapore, Singapore
Singapore Institute of Technology, Singapore, Singapore
e-mail: ping_liao@nni.com.sg
38 K. Y. Loh et al.
Abstract Oncotic cell death or oncosis represents a major mechanism of cell death
in ischaemic stroke, occurring in many central nervous system (CNS) cell types
including neurons, glia and vascular endothelial cells. In stroke, energy depletion
causes ionic pump failure and disrupts ionic homeostasis. Imbalance between the
influx of Na+ and Cl ions and the efflux of K+ ions through various channel proteins
and transporters creates a transmembrane osmotic gradient, with ensuing movement
of water into the cells, resulting in cell swelling and oncosis. Oncosis is a key
mediator of cerebral oedema in ischaemic stroke, contributing directly through
cytotoxic oedema, and indirectly through vasogenic oedema by causing vascular
endothelial cell death and disruption of the blood-brain barrier (BBB). Hence,
inhibition of uncontrolled ionic flux represents a novel and powerful strategy in
achieving neuroprotection in stroke. In this review, we provide an overview of
oncotic cell death in the pathology of stroke. Importantly, we summarised the
therapeutically significant pathways of water, Na+, Cl and K+ movement across
cell membranes in the CNS and their respective roles in the pathobiology of stroke.
Abbreviations
1 Introduction
In stroke, cell death exhibits two types of morphological changes: cell swelling, also
known as oncosis or oncotic cell death (Weerasinghe and Buja 2012), and cell
shrinkage, a hallmark of programmed cell death (Majno and Joris 1995). There are
two widely studied programmed cell death pathways: apoptosis and autophagy. In
oncosis, the root cause is increased membrane permeability, leading to cellular
swelling, organelle swelling and blebbing. In apoptosis, cell shrinkage and fragmen-
tation into membrane-bound apoptotic bodies are typical features. Following oncosis
or apoptosis, cells undergo irreversible necrosis and phagocytosis, which are uni-
versal processes in all cell death pathways (Fig. 1) (Lipton 1999). In stroke, there
have a heterogeneous distribution of oncotic and apoptotic processes across affected
brain regions. Cell apoptosis predominates in regions of reduced severity of ische-
mic damage (penumbra) as compared to the infarct core and following a shorter
period of ischaemia as compared to prolonged ischaemia (Charriaut-Marlangue et al.
1996). Pertinent factors influencing the outcome of cell death include cell type, age
and state of the cell at the point of insult (Martin et al. 1998). This review will discuss
oncotic cell death in stroke. Other types of cell death can be found in previous
excellent reviews (Fricker et al. 2018; Lipton 1999).
All types of cells in the brain including neuron, glia and vascular cells can
undergo oncosis following stroke. Buja et al. summarized the changes of cellular
oncosis into three stages (Buja 2005): In stage 1, ionic pumps fail to maintain ionic
balance across cell membrane due to a depletion of adenosine triphosphate (ATP),
and cell swelling ensues as water fluxes into the cell accompanying ions. It has been
predicted that when energy depletion lowers pump strength to less than 65% of
baseline, membrane potential is no longer preserved, and ionic homeostasis is
disrupted, leading to rapid influx of sodium and chloride ions (Dijkstra et al.
2016). During stage 2, membrane damage is irreversible, and more severe, allowing
large molecules such as propidium iodide (PI) and trypan blue to enter the cells.
Stage 3 represents the eventual physical disruption of the cell membrane, suggesting
that the cells are in the necrotic phase. Stage 3 is not unique to oncosis, but a
common pathway for all types of cell death.
The specific ionic composition of the cytosol is significantly different from
extracellular fluid (Fig. 2a). Cytosolic fluid contains a higher concentration of K+
and negatively charged protein, whereas extracellular fluid contains a higher con-
centration of Na+, Ca2+ and Cl . This ionic gradient is critical for cellular functions
including the formation of membrane potential. Various pumps consume ATP to
maintain this ionic gradient. After stroke, ATP depletion causes the functional failure
of the pumps, leading to an ionic flux down their respective gradients across the
cytosolic membrane (Fig. 2b). Na+, Ca2+ and Cl flow into the cell, whereas K+ ions
flow outside the cell. During the early stage of ATP depletion, cytosolic membrane is
relatively intact and impermeable to the intracellular negatively charged protein.
Therefore, no transmembrane movement of protein occurs. If ionic influx is greater
than efflux, a higher intracellular osmotic pressure is created. Water thus moves into
40 K. Y. Loh et al.
Fig. 1 Two major types of morphological cell changes in cell death: oncosis and apoptosis, as
characterised by cell swelling and shrinkage, respectively, with a common downstream pathway of
necrosis and phagocytosis (Reproduced with permission from Majno and Joris 1995)
the cell to balance the osmotic pressure, leading to cell swelling and oncosis. Given
that the extracellular Ca2+ concentration (1–2 mM) is much lower than that of Na+
(140–145 mM) and Cl (110 mM) concentrations, the osmotic pressure change
caused by Ca2+ influx is expected to be less prominent than Na+ and Cl influx.
Therefore, inhibition of sodium chloride influx following stroke could play a critical
role in preventing oncosis.
Oncotic Cell Death in Stroke 41
a Intracellular Extracellular
Na+ 5-15 mM Na+ 140-145 mM
K+ 140-150 mM K+ 4-5 mM
Ca2+ 0.1-1.5 μM Ca2+ 1.8 mM
Cl- 4-30 mM
Negavely charged proteins Cl- 110 mM
Na+
K+
Cl-
Ca2+
Fig. 2 (a) Ionic composition of intracellular and extracellular environment with the higher ionic
concentrations in bold. (b) Net ionic movement across cell membranes in ischaemia
In stroke, the primary initiating event is the arrest of oxidative phosphorylation and
reduction in ATP production. The resultant deactivation of ATP-dependent Na+
pumps, activation of glutamate-dependent channels, pH-sensitive proton-dependent
channels and other Na+- conducting ion channels culminate in uncontrolled accu-
mulation of intracellular Na+, causing anoxic depolarization. Pharmacological inhi-
bition of Na+ fluxes has exhibited neuroprotective effects early in ischaemic insult up
to 2 h post-ischaemia, demonstrating that the deleterious role of Na+ flux extends
well into the post-ischaemic period (Crumrine et al. 1997). Possible mechanisms of
Na+-induced damage in ischaemia include stimulation of glutamate release, increas-
ing intracellular Ca2+ and depletion of ATP via activation of the Na+/K+ pump
(Lipton 1999). Importantly, the influx of NaCl can generate a transmembrane
osmotic gradient, leading to water movement into the intracellular environment,
causing cell swelling and oncotic cell death (Trump et al. 1997; Weerasinghe and
Buja 2012). Here, we examine the major pathways of sodium influx in stroke which
is of therapeutic significance.
2.1 Na+/K+-ATPase
Mammalian NHEs are integral membrane ion channels that catalyse the exchange of
intracellular H+ for Na+. They perform an essential role in the regulation of intra-
cellular pH, Na+ content, volume and cellular proliferation in epithelial and
non-epithelial cells. Nine NHEs have been identified, and they can be classified as
plasma membrane NHEs (NHE1–5) and organellar NHEs (NHE6–9) (Ohgaki et al.
2011).
NHE1 is expressed ubiquitously in the CNS and functions to regulate intracellular
pH and volume in neurons and glial cells (Ma and Haddad 1997). Increased NHE1
expression was detected in the ischaemic penumbra in a rat model of focal cerebral
ischaemia (Jung et al. 2007). NHE1 activity is enhanced following intracellular
acidosis in stroke as a major pathway of proton extrusion, functioning as a protective
mechanism. The increase in NHE1 expression, on the other hand, contributes to
intracellular Na+ accumulation and resultant cell death in ischaemia. NHE1 inhibitor
cariporide has been successfully shown to reduce neuronal cell death and decrease
intracellular Na+ and Ca2+ concentrations in cultured mouse cortical neurons (Luo
et al. 2005). Furthermore, a reduced infarct volume was observed in NHE1 knockout
mice as compared to wild type following transient ischaemia and reperfusion (Luo
et al. 2005). These evidences support the role of NHE1in ischaemic neuronal injury.
NHE1activity is also enhanced in other ischaemic pathologies, notably, in cardiac
myocytes during ischaemic-reperfusion injury to maintain intracellular pH (Avkiran
2001; Avkiran et al. 2001).
Of note, NHE also functions physiologically to regulate vectorial ion transport
across epithelia, being expressed in both apical and basolateral membranes. In
particular, NHE1 and NHE2 isoforms have been found in the luminal membrane
of endothelial cells constituting the BBB (Wang et al. 2003; Noel et al. 1996;
Mokgokong et al. 2014). BBB Na+ transporters have been implicated in early
ischaemia-induced cerebral and astrocytic oedema in stroke prior to breakdown of
the BBB (Schielke et al. 1991). Studies have shown that NHE1 and NHE2 are
upregulated in cerebral microvascular endothelial cells in response to hypoxia and
46 K. Y. Loh et al.
via multimodal imaging, reduced infarct volume, improved motor function and
demonstrated vascular protective effects (Chen et al. 2018).
Although TRPM4 inhibition has been shown to protect vascular integrity and
reduce brain injury following acute ischaemic stroke, the underlying mechanisms
other than oncosis have not been fully elucidated. Numerous studies have shown that
inflammation plays a crucial role in ischaemic-reperfusion injury in stroke (Anrather
and Iadecola 2016; Pan et al. 2007; Leiva-Salcedo et al. 2017). Under hypoxic
conditions in stroke, necrotic cells and BBB leakage release damage signals which
initiate a cascade of inflammatory response via recruitment of cellular and humoral
components of the immune system, which is facilitated by reperfusion. Conse-
quently, this creates an outburst of pro-inflammatory mediators and reactive
oxidative species surrounding the ischaemic region, resulting in amplification of
ischaemic damage (Anrather and Iadecola 2016; Kim et al. 2014; Lakhan et al. 2009;
Tuttolomondo et al. 2009; Vidale et al. 2017; Wang et al. 2007). TRPM4 inhibition
has been shown to reduce neuroinflammation in animal models of subarachnoid
haemorrhage (Tosun et al. 2013) and experimental autoimmune encephalomyelitis
(Makar et al. 2015). The role of TRPM4 inhibition in neuroinflammation in stroke
needs further investigation.
The chloride co-transporters (CCCs) catalyse the coupled transport of Na+, K+ and
Cl across cell membranes and modulate the chloride electrochemical gradient. The
chloride gradient across neuronal membranes then determines the direction of
chloride current flow through the γ-aminobutyric acid type A (GABAA) receptors,
either leading to cellular hyperpolarization through chloride influx or cellular
depolarization through chloride efflux (Kaila et al. 2014).
The CCCs consist of three members based on their homeostatic functions: Na+-
K -2Cl co-transporters (NKCC) (two isoforms NKCC1 and NKCC2), Na+-Cl
+
channels mediate M-type currents in the CNS (Wang and Li 2016). M-type currents
are time- and voltage dependent and perform a key role in regulation of neuronal
excitability by maintenance of resting membrane potential, modulating excitability
thresholds and spike generations. Activation of M-type currents results in a
hyperpolarising effect, while inhibition produces neuronal excitation (Bierbower
et al. 2015). KCNQ activity is controlled by numerous signalling pathways (Wang
and Li 2016). Retigabine, an M-channel activator and an anti-epileptic drug, has
been known to yield the therapeutic effect via hyperpolarising neuronal membrane
potential (Miceli et al. 2008).
Early in stroke, M-current is activated by reactive oxidative species and is thought
to be neuroprotective in response to cellular hypoxia (Boscia et al. 2006). M-channel
activators achieved neuroprotection, effecting a reduction of infarct volume and
neurological deficits in mouse photothrombotic and MCAO stroke models. This
has been attributed to the hyperpolarising effect of K+ efflux which reduces neuronal
hyper-excitability in ischaemia. Interestingly, in the same study, it was found that the
neuroprotective effects of M-channel activators were more pronounced when admin-
istered early (0–3 h post-stroke compared to 6 h), indicating a probable therapeutic
time window of efficacy (Boscia et al. 2006). This was further supported by another
in vitro study which demonstrated the oxidative stress enhancement of KCNQ-
mediated current, and inhibition of which during 30 min of oxygen-glucose depri-
vation exacerbated cell death. These imply the early neuroprotective effects of
enhanced K+ efflux in the acute phase of excitotoxicity in stroke.
However, other studies have shown that activation of KCNQ-mediated K+-efflux
may contribute to apoptosis in the CNS. KCNQ2/3 activators resulted in dose-
dependent K+ efflux, intracellular decrease in K+, caspase activation and cell death
in hippocampal neurons, which was attenuated by a KCNQ inhibitor. In Chinese
hamster ovarian cells, KCNQ activators initiate the cascade of events leading to cell
apoptosis, which was, again, prevented by KCNQ inhibitors (Zhou et al. 2011; Song
and Yu 2014). It is possible that KCNQ-mediated K+ efflux is beneficial in the acute
stage of stroke by inhibiting neuronal excitotoxicity. However, with prolonged
ischaemia, disruption of cellular membrane integrity and persistent channel activa-
tion could result in dysregulation of K+ homeostasis and yield a pro-apoptotic effect
which overcomes its neuroprotective effect. Such temporal effect of KCNQ inhibi-
tion/activation in stroke requires further clarification.
KATP channels are expressed widely in cells of the CNS, including neurons, astro-
cytes, microglia, vascular smooth muscles and endothelium (Sun and Hu 2010).
KATP channels generally consist of two types of subunit: (1) a member of the K+-
inward rectifier family (Kir6.x) forming the central pore and (2) the SURx which
forms the regulatory subunit (Ashcroft and Gribble 1998). The four subunits present
in the CNS include Kir6.1, Kir6.2, SUR1 and SUR2, and these are heterogeneously
Oncotic Cell Death in Stroke 51
associated to form KATP channels in different CNS cell types. For example, the
Kir6.2 forms the central pore in most neurons, while Kir6.1 is predominantly
expressed in astrocytes and microglia (Thomzig et al. 2001, 2005). KATP regulates
neuronal, astrocytes, microglial and BBB functions in stroke, thus providing a
valuable therapeutic target (Sun and Hu 2010).
Under ischaemic conditions, with a decreased ATP:ADP ratio, activation of KATP
promotes a hyperpolarising K+ efflux current, reduces intracellular Ca2+ accumula-
tion, decreases excitatory glutamate release and the arrest of the ischaemic cascade,
providing a protective effect against neuronal excitotoxicity (Seino 2003). Iptakalim,
a KATP opener, reduced glutaminergic activity, promoted neuro-recovery and
decreased neuronal necrosis and apoptosis in in vivo and in vitro experiments
(Wang et al. 2004). The presence of the Kir6.2 subunit in mice showed a reduction
in period of neuronal depolarization post-ischaemia and neuroprotective effect,
while those without the subunit displayed more severe ischaemic neuronal damage
(Sun et al. 2006) and exquisite sensitivity to hypoxia-induced seizures (Yamada and
Inagaki 2005).
Besides reducing neuronal excitation, KATP may also reduce brain damage by
modulating other key players in the ischaemic cascade. Astrocytic gap junctions are
important in the transfer of toxic and beneficial molecules between cells in the
ischaemic core and penumbra. KATP opening was found to support the function of
gap junctions in astrocytes and may be beneficial in ischaemia (Sun and Hu 2010).
Activated microglia, the resident CNS macrophage, constitutes the early immune
response in stroke and exerts a deleterious effect by producing inflammatory and
cytotoxic mediators (Smith et al. 1998). Diazoxide, a KATP activator, was effective
in reducing microglial activation post-carotid artery occlusion in rat brain (Farkas
et al. 2005). In a Parkinsonian rat model, KATP activation suppressed microglial
inflammatory activity and inhibited the production of pro-inflammatory mediators
including inducible nitric oxide synthase (iNOS), tumour necrosis factor alpha
(TNF-α) and prostaglandin E2 (Zhou et al. 2008). Importantly, KATP may play a
critical role in regulating BBB integrity, as shown by the use of diazoxide post-
ischaemia-reperfusion to ameliorate cerebral oedema (Lenzser et al. 2005).
5 Aquaporins (AQPs)
(AQP6,8,11,12) (Meli et al. 2018). AQPs are widely distributed in cells and tissues
and play essential roles in maintaining water homeostasis.
AQP4 is the major subtype found in the central nervous system. They are found
exclusively in astrocyte membranes, specifically the perivascular end-feet and glial-
limiting membranes at the boundary between the brain parenchyma and the
subarachnoid cerebrospinal fluid (CSF) compartment and below the ependyma at
the boundary of the brain parenchyma and the ventricular CSF compartment. AQP4
have also been found in microglial membranes following lipopolysaccharide
induction (Papadopoulos and Verkman 2007).
AQP4 performs a bimodal role in the pathology of cerebral oedema in ischaemic
stroke. In the early post-ischaemic period, AQP4 contributes to oncotic cell death
and cytotoxic oedema by facilitating water influx into astrocytes following
dysregulation of ionic flux as mentioned previously. This causes extensive astrocyte
swelling which exacerbates brain injury through vascular congestion. Additionally,
astrocyte swelling can increase cellular permeability to excitatory mediators such as
glutamate through opening of volume-regulated ion channels, enhancing excitotoxic
cell damage. Following the initial post-ischaemic period, with increased tissue death
and breakdown of the BBB, leakage of serum proteins set off the process of
vasogenic oedema. This can potentially cause significant brain swelling, leading to
catastrophic brain herniation and death. AQP4 performs a protective role in resolu-
tion of vasogenic oedema by mediating transcellular movement of water from the
astrocyte cell membranes of the glial limitans into the subarachnoid CSF, and from
the ependyma and sub-ependymal astrocyte membranes into ventricular CSF, and
lastly through astrocyte pericapillary foot processes into serum. Hence, inhibition of
AQP4 early post-ischaemia may attenuate the deleterious effects of cytotoxic
oedema and astrocyte swelling, while promotion of AQP4 activity in late ischaemic
stroke or haemorrhagic conversion may enhance oedema clearance and decrease
neuroinflammation by protecting BBB integrity with decreased activation of inflam-
matory cells. AQP4 has also been implicated in other neuropathologies including
Parkinson’s disease and neuromyelitis optica (Vella et al. 2015).
The potential therapeutic effect of targeting AQP4 in ischaemic stroke has been
shown in various studies. AQP4 knockout mice has been associated with decreased
mortality and improved motor recovery 3–14 days post-MCAO with decreased
infarct volume, neuronal cell death and neuroinflammation compared to wild type.
However, oedema was not shown to be reduced in AQP4 knockout mice (Hirt et al.
2017). Another study employed the use of AQP4 inhibitor TGN020 15 min post-
MCAO in a rat model. TGN020 successfully reduced oedema, gliosis and apoptosis
3 and 7 days post-occlusion (Pirici et al. 2017). The discrepancy in effects on
cerebral oedema in the above-mentioned studies may be explained by the bimodal
effects of AQP4 in different phases post-stroke. Persistent cerebral oedema in AQP4
knockout mice may highlight the importance of AQP4 in oedema resolution in the
delayed post-ischaemic phase.
Although AQP4 is a promising target in the treatment of ischaemic stroke, it is
pertinent to further clarify the time point at which AQP4 inhibition or activation
provides the optimal therapeutic outcome. Furthermore, heterogeneity in AQP4
Oncotic Cell Death in Stroke 53
Table 1 Summary of ion channels and transporters involved in oncotic cell death in ischaemic
stroke
Ion
channels/
transporters Conduction Cell type Pharmacological agent
Sodium-conducting channels
Na+/K+- Na+, K+ Neurons, glial Cardiac glycosides (Liu et al. 2013)
ATPase
AMPA Na+, K+, Ca2+ Neurons, glial Memantine (Chen et al. 1992)
NMDA
ASIC1a Na+, Ca2+ Neurons Flurbiprofen (Mishra et al. 2010)
Amiloride (Miao et al. 2010)
NHE1 Na+, H+ Neurons, glial, vascular Cariporide (Luo et al. 2005)
endothelial SM20220 (Suzuki et al. 2002)
NBCe1 Na+, HCO3 Neurons, astrocytes S0859 (Yao et al. 2016)
TRPM4 Na+, K+, Cs+ Neurons, astrocytes, Glibenclamide (Simard et al. 2009)
vascular endothelial TRPM4-siRNA (Loh et al. 2014;
Chen et al. 2018)
Potassium-conducting channels
KCNQ K+ Neurons Retigabine (Miceli et al. 2008)
(Kv7)
KATP K+ Neurons, glial, vascular Iptakalim (Wang et al. 2004)
endothelial, vascular
smooth muscles
Chloride-conducting channels
NKCC1 Cl , Na+, K+ Neurons, glial, vascular Bumetanide (Chen et al. 2005;
endothelial, choroid plexus Yan et al. 2003; Wang et al. 2014;
epithelial Xu et al. 2017)
Aquaporins
AQP4 Water Astrocytes, microglial TGN020 (Pirici et al. 2017)
6 Mitochondrial Dysfunction
Mitochondria are the main source of cellular energy production in the form of ATP
by oxidative phosphorylation, alongside other important roles in cell metabolism,
calcium regulation, apoptosis and free radical production. Mitochondria in the
central nervous system are exquisitely susceptible to ischaemic damage. In early
54 K. Y. Loh et al.
stages of reduced cerebral perfusion, even in the absence of energy deficits, mito-
chondrial respiratory activities can exhibit severe alterations (Fiskum et al. 1999).
In ischaemic cell death, mitochondrial dysfunction disrupts the process of oxida-
tive phosphorylation and ATP production, increases generation of reactive oxygen
species (ROS) and impairs calcium buffering. In addition, the opening of mitochon-
drial permeability transition pores facilitates the release of pro-apoptotic mitochon-
drial proteins into the cytoplasm which activates both caspase-dependent
(cytochrome c) and caspase-independent (calpain) apoptotic pathways (Lipton
1999; Watts et al. 2013). In the ischaemic brain, mitochondrial dysfunction repre-
sents the key step to oncotic cell death. Accounting for more than 90% of cellular
ATP production, mitochondrial dysfunction results in the significant reduction of
ATP availability. This causes the disruption of normal ionic and water flux across
membranes, causing cell swelling and the ensuing process of oncosis in ischaemic
stroke (Mills et al. 2002).
Given the vital role of the electrochemical gradient across the mitochondrial
membranes in oxidative phosphorylation, it has been proposed that a loss in the
electrochemical gradient heralds the cellular changes in oncosis (Mills et al. 2002).
The uncoupling protein (UCP) family performs a central role in regulation of the
mitochondrial membrane potential. These proteins are expressed in the inner mito-
chondrial membranes. They regulate the membrane potential by facilitating entry of
protons which are driven out from the matrix during oxidative respiration. Hence,
they reduce the proton gradient required for generation of ATP in the oxidative
phosphorylation process (Klingenberg et al. 2001). The UCP-1 is highly expressed
in the inner mitochondrial membrane of brown adipocytes in animals. UCP-1
reduces ATP production and functions to divert energy use for cold-induced ther-
mogenesis (Ricquier and Bouillaud 2000b). Another member of the UCP family,
UCP-2, which is expressed widely in other tissues, has been examined for its role in
regulation of the mitochondrial membrane potential (Ricquier and Bouillaud 2000a).
In vitro studies have revealed that an increased expression of UCP-2 significantly
reduces mitochondrial membrane potential and ATP production and results in a
mechanism of cell death which is morphologically similar to oncosis (Mills et al.
2002). Inhibition of these uncoupling proteins has the potential to prevent oncotic
cell death by maintaining mitochondrial membrane potential and ATP production in
hypoxic conditions.
However, the role of UCP-2 in ischaemic stroke remains controversial. In vivo
studies with UCP-2 knockout mice paradoxically resulted in increased infarct
volume associated with suppression of anti-oxidant, cell-cycle and DNA-repair
genes and increased expression of inflammatory mediators (Haines et al. 2010).
The purported neuroprotective effect of UCP-2 has been attributed to its anti-oxidant
effect which reduces oxidative stress during cerebral ischaemia-reperfusion. UCP-2
overexpressing mice demonstrated decreased neuronal damage as compared to wild
type in cerebral ischaemia-reperfusion and had lower ROS levels (Haines and Li
2012). The role of UCP-2 on oncosis remains poorly understood.
The diverse functions of mitochondrial and pathological roles in cerebral ischae-
mia provide a potential novel target in stroke treatment. In addition to UCP-2, other
Oncotic Cell Death in Stroke 55
The myriad of possible therapeutic targets in oncotic cell death represent viable
alternatives in treatment of ischaemic stroke. However, despite promising results in
many preclinical studies, none of the experimented agents have successfully
progressed down the drug developmental pathways. To facilitate the development
of new therapies in stroke, several key issues need to be addressed.
Firstly, the expression and functions of the implicated ion channels exhibit
dynamic temporal changes following ischaemic stroke. The activation or inhibition
of the channels may or may not produce the desired neuroprotective effect
depending on the phases of ischaemic stroke when pharmacological intervention is
initiated. The temporal expression and role of the target channels at different phases
of stroke needs to be better characterised. This will aid in defining the optimal time
period post-stroke in targeting these ion channels.
Secondly, these ion channels are not unique to the central nervous system and are
expressed in other tissues. Ensuring selectivity in targeting the CNS represents
another key challenge. Many of the pharmacological agents that have been used in
preclinical studies represent currently available drugs used for other medical indi-
cations. Examples include glibenclamide, a diabetic medication, in targeting
TRPM4, and the use of bumetanide, a loop diuretic, in targeting NKCC1. Potential
adverse effects on other organ systems need to be carefully evaluated before the use
of these drugs in ischaemic stroke. While a more expensive option compared to
using existing medication, future directions in development of stroke therapies could
involve designing targeted therapies in the form of small molecules or other novel
blockers. Additionally, it is important to achieve good CNS penetration and stability
in the pharmacological design of the drugs.
The last issue involves the translation of preclinical stroke research to clinical
trials. Despite promising results in preclinical trials, many of the novel
neuroprotective agents failed at clinical trials. These have been attributed to poor
trial designs and questionable preclinical data. More stringent regulation of preclin-
ical animal studies is key to providing accurate and reproducible data to facilitate
progress to clinical trials. The Stroke Therapy and Academic Industry Roundtable
has implemented guidelines for preclinical stroke studies to ensure study quality.
Even so, compliance of studies to guidelines remains a problem. In addition,
inappropriate animal stroke models used may influence success at human trials
despite full compliance to the guidelines. Future efforts should also focus on
evaluating the suitability of various animal stroke models in mimicking the human
diseased state (Herson and Traystman 2014).
56 K. Y. Loh et al.
8 Conclusion
Oncosis represents a major type of cell death in stroke, occurring directly after
energy depletion. Being closely linked to transmembrane electrolyte movements,
blocking or delaying such movements will ameliorate oncotic cell death. As the
process involves multiple ions and their respective transmembrane transporters, a
better strategy is to target multiple pathways simultaneously, yielding a synergistic
effect in preventing oncotic cell death. In comparison to apoptosis and autophagy,
where key players of the death pathway are located intracellularly, the mediators of
oncosis largely consist of transmembrane channels and transporters, suggesting that
they are a superior and more accessible class of target for intervention. Furthermore,
as oncosis is involved in cell death in major CNS cell types, namely, astrocytes,
neurons and vascular cells, inhibition of oncosis will ameliorate both cytotoxic
and vasogenic oedema. With a stabilised BBB, reduction of subsequent neuro-
inflammation is expected. It should be noted that most ion channels and transporters
are constitutively expressed in the healthy brain and other non-CNS tissues and
organs. The development of drugs that selectively act on hypoxic CNS cells without
affecting healthy tissues remains a challenge. With further advancements in drug
discovery, targeting of oncosis may form a powerful strategy in achieving
neuroprotection in a post-stroke setting.
References
Farkas E, Timmer NM, Domoki F, Mihaly A, Luiten PG, Bari F (2005) Post-ischemic administra-
tion of diazoxide attenuates long-term microglial activation in the rat brain after permanent
carotid artery occlusion. Neurosci Lett 387(3):168–172. https://doi.org/10.1016/j.neulet.2005.
06.036
Fiskum G, Murphy AN, Beal MF (1999) Mitochondria in neurodegeneration: acute ischemia and
chronic neurodegenerative diseases. J Cereb Blood Flow Metab 19(4):351–369. https://doi.org/
10.1097/00004647-199904000-00001
Frenguelli BG (2013) Glutamate receptor-dependent synaptic plasticity. Neuropharmacology 74:1.
https://doi.org/10.1016/j.neuropharm.2013.05.009
Fricker M, Tolkovsky AM, Borutaite V, Coleman M, Brown GC (2018) Neuronal cell death.
Physiol Rev 98(2):813–880. https://doi.org/10.1152/physrev.00011.2017
Gerelsaikhan T, Turner RJ (2000) Transmembrane topology of the secretory Na+-K+-2Cl-
cotransporter NKCC1 studied by in vitro translation. J Biol Chem 275(51):40471–40477.
https://doi.org/10.1074/jbc.M007751200
Geter-Douglass B, Witkin JM (1999) Behavioral effects and anticonvulsant efficacies of
low-affinity, uncompetitive NMDA antagonists in mice. Psychopharmacology 146(3):280–289
Glushchenko TS, Izvarina NL (1997) Na+,K(+)-ATPase activity in neurons and glial cells of the
olfactory cortex of the rat brain during the development of long-term potentiation. Neurosci
Behav Physiol 27(1):49–52
Goldberg WJ, Watson BD, Busto R, Kurchner H, Santiso M, Ginsberg MD (1984) Concurrent
measurement of (Na+, K+)-ATPase activity and lipid peroxides in rat brain following reversible
global ischemia. Neurochem Res 9(12):1737–1747
Gomez-Angelats M, Bortner CD, Cidlowski JA (2000) Cell volume regulation in immune cell
apoptosis. Cell Tissue Res 301(1):33–42
Haines B, Li PA (2012) Overexpression of mitochondrial uncoupling protein 2 inhibits inflamma-
tory cytokines and activates cell survival factors after cerebral ischemia. PLoS One 7(2):e31739.
https://doi.org/10.1371/journal.pone.0031739
Haines BA, Mehta SL, Pratt SM, Warden CH, Li PA (2010) Deletion of mitochondrial uncoupling
protein-2 increases ischemic brain damage after transient focal ischemia by altering gene
expression patterns and enhancing inflammatory cytokines. J Cereb Blood Flow Metab
30(11):1825–1833. https://doi.org/10.1038/jcbfm.2010.52
Hasbani MJ, Hyrc KL, Faddis BT, Romano C, Goldberg MP (1998) Distinct roles for sodium,
chloride, and calcium in excitotoxic dendritic injury and recovery. Exp Neurol 154(1):241–258.
https://doi.org/10.1006/exnr.1998.6929
Herson PS, Traystman RJ (2014) Animal models of stroke: translational potential at present and in
2050. Future Neurol 9(5):541–551. https://doi.org/10.2217/fnl.14.44
Hirt L, Fukuda AM, Ambadipudi K, Rashid F, Binder D, Verkman A, Ashwal S, Obenaus A,
Badaut J (2017) Improved long-term outcome after transient cerebral ischemia in aquaporin-4
knockout mice. J Cereb Blood Flow Metab 37(1):277–290. https://doi.org/10.1177/
0271678X15623290
Hu HJ, Song M (2017) Disrupted ionic homeostasis in ischemic stroke and new therapeutic targets.
J Stroke Cerebrovasc Dis 26(12):2706–2719. https://doi.org/10.1016/j.jstrokecerebrovasdis.
2017.09.011
Jung YW, Choi IJ, Kwon TH (2007) Altered expression of sodium transporters in ischemic
penumbra after focal cerebral ischemia in rats. Neurosci Res 59(2):152–159. https://doi.org/
10.1016/j.neures.2007.06.1470
Kaila K, Price TJ, Payne JA, Puskarjov M, Voipio J (2014) Cation-chloride cotransporters in
neuronal development, plasticity and disease. Nat Rev Neurosci 15(10):637–654. https://doi.
org/10.1038/nrn3819
Kang TC, An SJ, Park SK, Hwang IK, Suh JG, Oh YS, Bae JC, Won MH (2002) Alterations in
Na+/H+ exchanger and Na+/HCO3- cotransporter immunoreactivities within the gerbil hippo-
campus following seizure. Brain Res Mol Brain Res 109(1–2):226–232
Oncotic Cell Death in Stroke 59
Ma E, Haddad GG (1997) Expression and localization of Na+/H+ exchangers in rat central nervous
system. Neuroscience 79(2):591–603
MacMillan V (1982) Cerebral Na+,K+-ATPase activity during exposure to and recovery from acute
ischemia. J Cereb Blood Flow Metab 2(4):457–465. https://doi.org/10.1038/jcbfm.1982.52
Majno G, Joris I (1995) Apoptosis, oncosis, and necrosis. An overview of cell death. Am J Pathol
146(1):3–15
Makar TK, Gerzanich V, Nimmagadda VK, Jain R, Lam K, Mubariz F, Trisler D, Ivanova S, Woo
SK, Kwon MS, Bryan J, Bever CT, Simard JM (2015) Silencing of Abcc8 or inhibition of newly
upregulated Sur1-Trpm4 reduce inflammation and disease progression in experimental autoim-
mune encephalomyelitis. J Neuroinflammation 12:210. https://doi.org/10.1186/s12974-015-
0432-3
Martin LJ, Al-Abdulla NA, Brambrink AM, Kirsch JR, Sieber FE, Portera-Cailliau C (1998)
Neurodegeneration in excitotoxicity, global cerebral ischemia, and target deprivation: a
perspective on the contributions of apoptosis and necrosis. Brain Res Bull 46(4):281–309
Martin-Aragon Baudel MA, Poole AV, Darlison MG (2017) Chloride co-transporters as possible
therapeutic targets for stroke. J Neurochem 140(2):195–209. https://doi.org/10.1111/jnc.13901
Meli R, Pirozzi C, Pelagalli A (2018) New perspectives on the potential role of aquaporins (AQPs)
in the physiology of inflammation. Front Physiol 9:101. https://doi.org/10.3389/fphys.2018.
00101
Merello M, Nouzeilles MI, Cammarota A, Leiguarda R (1999) Effect of memantine (NMDA
antagonist) on Parkinson’s disease: a double-blind crossover randomized study. Clin
Neuropharmacol 22(5):273–276
Miao Y, Zhang W, Lin Y, Lu X, Qiu Y (2010) Neuroprotective effects of ischemic preconditioning
on global brain ischemia through up-regulation of acid-sensing ion channel 2a. Int J Mol Sci
11(1):140–153. https://doi.org/10.3390/ijms11010140
Miceli F, Soldovieri MV, Martire M, Taglialatela M (2008) Molecular pharmacology and thera-
peutic potential of neuronal Kv7-modulating drugs. Curr Opin Pharmacol 8(1):65–74. https://
doi.org/10.1016/j.coph.2007.10.003
Mills EM, Xu D, Fergusson MM, Combs CA, Xu Y, Finkel T (2002) Regulation of cellular oncosis
by uncoupling protein 2. J Biol Chem 277(30):27385–27392. https://doi.org/10.1074/jbc.
M111860200
Mishra V, Verma R, Raghubir R (2010) Neuroprotective effect of flurbiprofen in focal cerebral
ischemia: the possible role of ASIC1a. Neuropharmacology 59(7–8):582–588. https://doi.org/
10.1016/j.neuropharm.2010.08.015
Mokgokong R, Wang S, Taylor CJ, Barrand MA, Hladky SB (2014) Ion transporters in brain
endothelial cells that contribute to formation of brain interstitial fluid. Pflugers Arch
466(5):887–901. https://doi.org/10.1007/s00424-013-1342-9
Moseley AE, Lieske SP, Wetzel RK, James PF, He S, Shelly DA, Paul RJ, Boivin GP, Witte DP,
Ramirez JM, Sweadner KJ, Lingrel JB (2003) The Na,K-ATPase alpha 2 isoform is expressed in
neurons, and its absence disrupts neuronal activity in newborn mice. J Biol Chem
278(7):5317–5324. https://doi.org/10.1074/jbc.M211315200
Noel J, Roux D, Pouyssegur J (1996) Differential localization of Na+/H+ exchanger isoforms
(NHE1 and NHE3) in polarized epithelial cell lines. J Cell Sci 109(Pt 5):929–939
O’Donnell ME, Lam TI, Tran L, Anderson SE (2004) The role of the blood-brain barrier Na-K-2Cl
cotransporter in stroke. Adv Exp Med Biol 559:67–75
Ohgaki R, van IJzendoorn SC, Matsushita M, Hoekstra D, Kanazawa H (2011) Organellar Na+/H+
exchangers: novel players in organelle pH regulation and their emerging functions. Biochem-
istry 50(4):443–450. https://doi.org/10.1021/bi101082e
Pan J, Konstas AA, Bateman B, Ortolano GA, Pile-Spellman J (2007) Reperfusion injury following
cerebral ischemia: pathophysiology, MR imaging, and potential therapies. Neuroradiology
49(2):93–102. https://doi.org/10.1007/s00234-006-0183-z
Papadopoulos MC, Verkman AS (2007) Aquaporin-4 and brain edema. Pediatr Nephrol
22(6):778–784. https://doi.org/10.1007/s00467-006-0411-0
Oncotic Cell Death in Stroke 61
Pfeffer CK, Stein V, Keating DJ, Maier H, Rinke I, Rudhard Y, Hentschke M, Rune GM, Jentsch
TJ, Hubner CA (2009) NKCC1-dependent GABAergic excitation drives synaptic network
maturation during early hippocampal development. J Neurosci 29(11):3419–3430. https://doi.
org/10.1523/JNEUROSCI.1377-08.2009
Pietrini G, Matteoli M, Banker G, Caplan MJ (1992) Isoforms of the Na,K-ATPase are present in
both axons and dendrites of hippocampal neurons in culture. Proc Natl Acad Sci U S A
89(18):8414–8418
Pignataro G, Simon RP, Xiong ZG (2007) Prolonged activation of ASIC1a and the time window for
neuroprotection in cerebral ischaemia. Brain 130(Pt 1):151–158. https://doi.org/10.1093/brain/
awl325
Pirici I, Balsanu TA, Bogdan C, Margaritescu C, Divan T, Vitalie V, Mogoanta L, Pirici D, Carare
RO, Muresanu DF (2017) Inhibition of aquaporin-4 improves the outcome of ischaemic stroke
and modulates brain paravascular drainage pathways. Int J Mol Sci 19(1). https://doi.org/10.
3390/ijms19010046
Radu BM, Banciu A, Banciu DD, Radu M (2016) Acid-sensing ion channels as potential pharma-
cological targets in peripheral and central nervous system diseases. Adv Protein Chem Struct
Biol 103:137–167. https://doi.org/10.1016/bs.apcsb.2015.10.002
Remillard CV, Yuan JX (2004) Activation of K+ channels: an essential pathway in programmed
cell death. Am J Physiol Lung Cell Mol Physiol 286(1):L49–L67. https://doi.org/10.1152/
ajplung.00041.2003
Ricquier D, Bouillaud F (2000a) Mitochondrial uncoupling proteins: from mitochondria to the
regulation of energy balance. J Physiol 529(Pt 1):3–10
Ricquier D, Bouillaud F (2000b) The uncoupling protein homologues: UCP1, UCP2, UCP3,
StUCP and AtUCP. Biochem J 345(Pt 2):161–179
Rothman SM (1985) The neurotoxicity of excitatory amino acids is produced by passive chloride
influx. J Neurosci 5(6):1483–1489
Schielke GP, Moises HC, Betz AL (1991) Blood to brain sodium transport and interstitial fluid
potassium concentration during early focal ischemia in the rat. J Cereb Blood Flow Metab
11(3):466–471. https://doi.org/10.1038/jcbfm.1991.89
Schrodl-Haussel M, Theparambil SM, Deitmer JW, Roussa E (2015) Regulation of functional
expression of the electrogenic sodium bicarbonate cotransporter 1, NBCe1 (SLC4A4), in mouse
astrocytes. Glia 63(7):1226–1239. https://doi.org/10.1002/glia.22814
Seino S (2003) Physiology and pathophysiology of K(ATP) channels in the pancreas and cardio-
vascular system: a review. J Diabetes Complicat 17(2 Suppl):2–5
Seki G, Horita S, Suzuki M, Yamazaki O, Usui T, Nakamura M, Yamada H (2013) Molecular
mechanisms of renal and extrarenal manifestations caused by inactivation of the electrogenic
Na(+)-HCO3( ) cotransporter NBCe1. Front Physiol 4:270. https://doi.org/10.3389/fphys.
2013.00270
Simard JM, Chen M, Tarasov KV, Bhatta S, Ivanova S, Melnitchenko L, Tsymbalyuk N, West GA,
Gerzanich V (2006) Newly expressed SUR1-regulated NC(Ca-ATP) channel mediates cerebral
edema after ischemic stroke. Nat Med 12(4):433–440. https://doi.org/10.1038/nm1390
Simard JM, Yurovsky V, Tsymbalyuk N, Melnichenko L, Ivanova S, Gerzanich V (2009) Protec-
tive effect of delayed treatment with low-dose glibenclamide in three models of ischemic stroke.
Stroke 40(2):604–609. https://doi.org/10.1161/STROKEAHA.108.522409
Simard JM, Woo SK, Gerzanich V (2012) Transient receptor potential melastatin 4 and cell death.
Pflugers Arch 464(6):573–582. https://doi.org/10.1007/s00424-012-1166-z
Simard JM, Sheth KN, Kimberly WT, Stern BJ, del Zoppo GJ, Jacobson S, Gerzanich V (2014)
Glibenclamide in cerebral ischemia and stroke. Neurocrit Care 20(2):319–333. https://doi.org/
10.1007/s12028-013-9923-1
Small DL, Morley P, Buchan AM (1999) Biology of ischemic cerebral cell death. Prog Cardiovasc
Dis 42(3):185–207
Smith ME, van der Maesen K, Somera FP (1998) Macrophage and microglial responses to
cytokines in vitro: phagocytic activity, proteolytic enzyme release, and free radical production.
J Neurosci Res 54(1):68–78. https://doi.org/10.1002/(SICI)1097-4547(19981001)54:1<68::
AID-JNR8>3.0.CO;2-F
62 K. Y. Loh et al.
Yu CY, Ng G, Liao P (2013) Therapeutic antibodies in stroke. Transl Stroke Res 4(5):477–483.
https://doi.org/10.1007/s12975-013-0281-2
Zhang E, Liao P (2015) Brain transient receptor potential channels and stroke. J Neurosci Res
93(8):1165–1183. https://doi.org/10.1002/jnr.23529
Zhou F, Yao HH, Wu JY, Ding JH, Sun T, Hu G (2008) Opening of microglial K(ATP) channels
inhibits rotenone-induced neuroinflammation. J Cell Mol Med 12(5A):1559–1570. https://doi.
org/10.1111/j.1582-4934.2007.00144.x
Zhou X, Wei J, Song M, Francis K, Yu SP (2011) Novel role of KCNQ2/3 channels in regulating
neuronal cell viability. Cell Death Differ 18(3):493–505. https://doi.org/10.1038/cdd.2010.120
Rev Physiol Biochem Pharmacol (2019) 176: 65–106
DOI: 10.1007/112_2018_15
© Springer Nature Switzerland AG 2018
Published online: 8 November 2018
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
1.1 Toward the Identification of Mammalian/Human Mg2+ Transporters/Mg2+
Homeostatic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2 Novel Mammalian Magnesiotropic Genes Encode for Electrogenic Mg2+ Transporters
(Channels) When Expressed in Xenopus laevis Oocytes but Not in Homologous Expression
Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3 Expression Profiles of Novel Magnesiotropic Genes and Their Protein Products . . . . . . . . . . 73
4 Cellular Compartmentalization and Functional Characterization of Proteins Encoded by
the Putative and Confirmed Mg2+ Transporters/Mg2+ Homeostatic Factors . . . . . . . . . . . . . . . . 73
4.1 Mg2+ Transporters and/or Mg2+ Homeostatic Factors Localized to Mitochondria . . . 75
4.2 Mg2+ Transporters and/or Mg2+ Homeostatic Factors Localized to Endoplasmic
Reticulum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.3 Mg2+ Transporters and/or Mg2+ Homeostatic Factors Localized to the Golgi
Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.4 Mg2+ Transporters and/or Mg2+ Homeostatic Factors Localized to Lysosomes . . . . . . 84
4.5 Mg2+ Transport Across Nuclear Membrane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.6 Putative and Confirmed Mg2+ Transporters/Mg2+ Homeostatic Factors with Unclear
Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.7 Putative and Confirmed Mg2+ Transporters/Mg2+ Homeostatic Factors with Plasma
Membrane Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4.8 MagT1 (Magnesium Transporter Subtype 1) and TUSC3/N33 (Tumor Suppressor
Candidate 3): Mg2+ Transporter Candidates Which Turned to Have Other but
Not Mg2+ Transport Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Abstract Magnesium research has boomed within the last 20 years. The real
breakthrough came at the start of the new millennium with the discovery of
a plethora of possible Mg homeostatic factors that, in particular, included putative
Mg2+ transporters. Until that point, Mg research was limited to biochemical and
physiological work, as no target molecular entities were known that could be used to
explore the molecular biology of Mg homeostasis at the level of the cell, tissue,
organ, or organism and to translate such knowledge into the field of clinical medicine
and pharmacology. Because of the aforementioned, Mg2+ and Mg homeostasis, both
of which had been heavily marginalized within the biomedical field in the twentieth
century, have become overnight a focal point of many studies ranging from primary
biomedical research to translational medicine.
The amount of literature concerning cellular Mg2+ transport and cellular Mg
homeostasis is increasing, together with a certain amount of confusion, especially
about the function(s) of the newly discovered and, in the majority of instances, still
only putative Mg2+ transporters/Mg2+ homeostatic factors. Newcomers to the field
of Mg research will thus find it particularly difficult to orient themselves.
Here, we briefly but critically summarize the status quo of the current under-
standing of the molecular entities behind cellular Mg2+ homeostasis in mammalian/
human cells other than TRPM6/7 chanzymes, which have been universally accepted
as being unspecific cation channel kinases allowing the flux of Mg2+ while consti-
tuting the major gateway for Mg2+ to enter the cell.
1 Introduction
2007; Romani 2007, 2011). Many of the functions of Mg2+ result from its being a
natural antagonist of Ca2+ in a plethora of cellular processes (Nishizawa et al. 2007).
The basal concentration of bioavailable intracellular Mg2+ ([Mg2+]i) ranges from
0.2 to 1 mM in various mammalian cell types, a concentration that is enormous when
compared with resting [Ca2+]i, which ranges from only 10s to ~100 nM (Romani
2007, 2011; Romani and Scarpa 2000; Delva et al. 1996, 2006; Zhou and Clapham
2009; Dragileva et al. 1999).
Basal [Mg2+]i is highly stable. Even when exposed to unphysiologically high
(e.g., 10–30 mM) or low (0–0.3 mM) extracellular [Mg2+] ([Mg2+]e), the cell
maintains its [Mg2+]i close to its basal value. Furthermore, the electrochemical
equilibrium potential (VeqMg2þ ) would be reached at the [Mg2+]i of ~50 mM in most
eukaryotes under resting conditions, but in reality, cells keep [Mg2+]i 50–100 times
lower than that. This suggests the existence of effective cellular regulatory (Mg2+
transport) mechanisms that maintain intracellular Mg2+ homeostasis (Nishizawa
et al. 2007; Romani 2007, 2011; Schweigel-Röntgen and Kolisek 2014; Flatman
1984).
The [Mg2+]i (cytosol) of most mammalian cells types is almost equimolar to
the concentration of [Mg2+]e (extracellular fluid; [Mg2+]e ¼ approximately
0.8–1 mM vs. [Mg2+]i ¼ approximately 0.2–1 mM) (Nishizawa et al. 2007; Romani
2007, 2011; Romani and Scarpa 2000; Delva et al. 1996; Schweigel-Röntgen and
Kolisek 2014). Similarly, a Mg2+ concentration gradient is almost nonexistent
between [Mg2+]i (cytosol) and [Mg2+]io (intracellular organelles) (Nishizawa et al.
2007; Romani 2011; Rutter et al. 1990). This emphasizes the importance of the
electrical component of the electrochemical gradient for Mg2+ transport and further
suggests that a primary active (ATP-dependent) or secondary active (dependent on
Na+- or H+-motive forces) Mg2+ transport mechanism would be of great importance
for the maintenance of cellular Mg2+ homeostasis and transcellular Mg2+ transport
(Nishizawa et al. 2007; Romani 2011).
Several Mg2+ transport mechanisms have been proposed to exist based on
biochemical and physiological evidence. Among these are the following: Mg2+ ion
channel(s), Na+/Mg2 and H+/Mg2+ exchangers, Mn2+/Mg2+ and Ca2+/Mg2+
exchangers, choline/Mg2+ exchanger, and Mg2+ HCO3 a Mg2+ Cl symporter
(Romani 2011; Schweigel-Röntgen and Kolisek 2014; Günther et al. 1990; Ebel
et al. 2002; McGuigan et al. 2002; Günther 1993).
Pioneering work in the world of Mg2+ transporters in prokaryotes was carried out by
Michael E. Maguire and his colleagues; they described three prominent classes of
bacterial Mg2+ transporters (CorA, MgtA, and MgtE) long before the discovery of
eukaryotic Mg2+ transporters (Hmiel et al. 1986, 1989; Smith et al. 1995). The major
68 M. Kolisek et al.
bacterial Mg2+ transporters CorA and MgtE have a wide phylogenetic distribution,
whereas MgtA occurs only in some bacteria (Maguire 2006; Groisman et al. 2013).
MgtA and MgtB are P-type ATPases that mediate Mg2+ influx. Both are exclu-
sive to bacteria (Groisman et al. 2013; Maguire 1992). On the other hand, eukaryotic
homologs of CorA and MgtE have been identified (Groisman et al. 2013).
The first Mg2+ transporters ever described in eukaryotic cells (yeast:
Saccharomyces cerevisiae) were the CorA homologs Mrs2 and Alr1 (Bui et al.
1999; Gregan et al. 2001a; Graschopf et al. 2001). Whereas the Mrs2 gene is nuclear,
the Mrs2 protein operates in mitochondria, namely, in the inner mitochondrial
membrane (IMM; Fig. 1) (Gregan et al. 2001a; Wiesenberger et al. 1992; Kolisek
et al. 2003; Schindl et al. 2007). Mrs2 is now widely accepted as being an essential
component of the mitochondrial Mg2+ homeostatic network from yeast to man
(Kolisek et al. 2003; Schindl et al. 2007; Zsurka et al. 2001; Piskacek et al. 2009;
Sponder et al. 2013a; Yamanaka et al. 2016; Merolle et al. 2018). Alr1 is a Mg2+
transporter integral to the cytoplasmic membrane of S. cerevisiae (Graschopf et al.
2001).
Fig. 1 Cellular Mg2+ transport mechanisms and/or Mg2+ homeostatic factors. The confirmed Mg2+
transporters/Mg2+ homeostatic factors are shown in color. The putative Mg2+ transporters/Mg2+
homeostatic factors are shown in white. MagT1 and TUSC3/N33 are crossed (red cross) because
they were already disproved to function as Mg2+ transporters. SLC41A2, CNNM1, CNNM3, and
CNNM4 are not depicted due to ongoing controversy either on their localization or function(s) or
both. CNNM2 is depicted as a Mg2+ homeostatic factor
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 69
Almost in parallel with Mrs2, chanzymes TRPM6/7 were discovered and iden-
tified as unique natural fusion proteins comprising a functional channel domain and a
functional kinase domain (Montell 2003; Nadler et al. 2001; Schlingmann et al.
2002). The abilities of TRPM6/7 to conduct Mg2+ transport have been robustly
researched over the last 15 years, and many reviews are available online. Hence, in
the following text, we will only mention TRPM6/7 when relevant.
The real momentum to the otherwise rather small field of Mg2+ transport research
was provided by the work of Angela Goytain and Gary A. Quamme. At the turn of
the millennium, they designed an elegant set of Mg2+ starvation experiments
utilizing mice and MDCT (mouse distal convoluted tubule) cells. By using oligo-
nucleotide microarray technology, quantitative real-time RT-PCR, and in silico
similarity searches, they identified magnesiotropic genes (MgG(s), namely, genes
reactive to changing [Mg2+]i at the level of gene expression) encoding for 15 novel
putative Mg2+ transporters (npMgT; Fig. 1) in mouse (SLC41A1, SLC41A2,
SLC41A3, NIPA1, NIPA2, NIPA3 (human NIPAL1), NIPA4 (human NIPAL4),
MagT1, TUSC3/N33, MMgT1, MMgT2, CNNM2, HIP14 and HIP14L, and
MagC1) (Quamme 2010). All of these mouse MgGs, except for MMgT2, also
have human homologs (Table 1; https://www.genecards.org; https://www.uniprot.
org). Most of the proteins encoded by mouse MgGs have been characterized as
encoding for “Mg2+ transporters with channel-like properties” (Quamme 2010;
Goytain and Quamme 2005a, b, c, d, 2008; Goytain et al. 2007, 2008a, b). Thus,
the group led by Goytain and Quamme revolutionized the whole field of mamma-
lian/human magnesium research by suggesting molecular identities of the entities
constituting the cellular Mg homeostatic machinery.
structures, Lys lysosome, MgCh Mg2+ channel, MgChz Mg2+ chanzyme, MgHF Mg homeostatic
factor, MgT Mg2+ transporter, N nucleus, NME Na+/Mg2+ exchanger, OST oligosaccharyl-
transferase, PA palmitoyl acyltransferase, PMgP putative Mg2+ pump, PMgT putative Mg2+
transporter, SCaMC short Ca2+-binding mitochondrial carrier, UCChZ unspecific cation chanzyme
a
Magnesiotropic genes identified by the group around Quamme
b
Predicted in silico based on similarity search
c
Other function attributed, but not Mg2+ transport
d
Not present in human (identified in mouse)
Table 2 Permeation profiles established for Mg2+ transporters by two-electrode voltage clamp
(green) or by patch clamp technique (violet)
Protein Permeation profile Citation
TRPM6 Px/PCa (Hs): Ba2+ ≥ Ni2+ > Mg2+ > Ca2+ Voets et al. (2004)
TRPM7 Px/PCa (Hs): Zn2+ ≈ Ni2+ Ba2+ > Co2+ > Mg2+ ≥ Mn2+ ≥≥ Sr2+ ≥ Cd2+ ≥ Ca2+ Penner and Fleig (2007) and
Monteilh-Zoller et al. (2003)
Mrs2 Px/PMg (Sc): Mg2+ > Ni2+ Schindl et al. (2007)
NT (not transported): Ca2+, Mn2+, Co2+
SLC41A1 Px/PMg (Mm): Mg2+ > Sr2+ ≈ Fe2+ > Ba2+ > Cu2+ > Zn2+ ≈ Co2+ > Cd2+ > Mn2+ Quamme (2010) and Goytain
NT: Ni2+, Ca2+, Gd3+ and Quamme (2005a)
SLC41A3a Px/PMg (Mm): Ba2+ > Mg2+ > Ni2+ ≈ Zn2+ > Sr2+ ≈ Fe2+ > Mn2+ > Cu2+ ≈ Co2+ Quamme (2010)
NT: Ca2+, Cd2+
SLC41A2 Px/PMg (Mm): Mg2+ > Ba2+ > Ni2+ > Co2+ > Sr2+ ≈ Fe2+ > Mn2+ Quamme (2010) and Goytain
NT: Ca2+, Cu2+, Zn2+, Cd2+ and Quamme (2005b)
MagT1 (Mm) Highly selective for Mg2+ Quamme (2010) and Goytain
and Quamme (2005c)
TUSC3/N33 a Px/PMg (Mm): Mg2+ > Mn2+ > Cu2+ > Fe2+ > Ba2+ ≈ Co2+ > Sr2+ ≈ Zn2+ ≥ Ni2+ > Ca2+ Quamme (2010)
NIPA1 Px/PMg (Mm): Mg2+ > Sr2+ > Co2+ > Zn2+ ≈ Fe2+ > Ni2+ ≈ Ca2+ ≈ Ba2+ ≈ Cu2+ ≈ Mn2+ Quamme (2010) and Goytain
NT: Cd2+ et al. (2007, 2008a)
NIPA2 Px/PMg (Mm): Mg2+ Sr2+ ≈ Co2+ ≈ Zn2+ ≈ Fe2+ ≈ Cd2+ ≈ Ni2+ ≈ Ca2+ ≈ Ba2+ ≈ Cu2+ ≈ Mn2+ Quamme (2010) and Goytain
et al. (2008a)
2+
NPAL3 Px/PMg (Mm): Mg > Sr2+ > Ba2+ > Fe2+ ≈ Mn2+ > Cu2+ ≈ Co2+ > Zn2+ ≈ Cd2+ > Ni2+ ≈ Ca2+ Quamme (2010) and Goytain
et al. (2008a)
NPAL4/Ichthyin Px/PMg (Mm): Mg2+ > Ba2+ > Sr2+ > Fe2+ > Cu2+ ≈ Ca2+ ≈ Zn2+ ≥ Co2+ ≈ Mn2+ ≈ Ni2+ Quamme (2010) and Goytain
NT: Cd2+ et al. (2008a)
MMgT1 Px/PMg (Mm): Mg2+ ≥ Sr2+ ≥ Fe2+ > Co2+ > Cu2+ > Ba2+ > Ca2+ > Zn2+ > Mn2+ > Ni2+ Quamme (2010) and Goytain
and Quamme (2008)
MMgT2 Px/PMg (Mm): Mg2+ ≥ Sr2+ > Ba2+ > Cu2+ > Mn2+ > Co2+ > Ni2+ > Fe2+ > Zn2+ ≥ Ca2+ Quamme (2010) and Goytain
and Quamme (2008)
CNNM2 Px/PMg (Mm): Mg2+ ≥ Co2+ > Ba2+ ≥ Mn2+ ≈ Gd3+ ≥ Sr2+ > Cu2+ ≥ Fe2+ Quamme (2010) and Goytain
NT: Zn2+, Cd2+, Ni2+, Ca2+ and Quamme (2005d)
The permeation profile in original paper (Goytain and Quamme 2005d) does not match
the permeation profile published in review (Quamme 2010)
CNNM1 Px/PMg: not available
CNNM3a Px/PMg (Mm): Mg2+ > Fe2+ > Cu2+ > Co2+> Ni2+ > Ca2+ Quamme (2010)
NT: Sr2+, Ba2+, Mn2+, Zn2+, Cd2+
CNNM4 Px/PMg: not available
HIP14 Px/PMg (Mm): Mg2+ ≈ Sr2+ > Ni2+ > Ba2+ > Zn2+ ≥ Mn2+ > Fe2+ Quamme (2010) and Goytain
NT: Co2+, Cu2+, Ca2+ et al. (2008b)
HIP14L Px/PMg (Mm): Mg2+ ≈ Sr2+ > Ni2+ ≥ Mn2+ > Cu2+ ≈ Ba2+ ≈ Zn2+ Quamme (2010) and
NT: Fe2+, Co2+, Ca2+ Goytain et al. (2008b)
a
Indicates that the data were introduced in the review paper and not in original research paper.
Please note that the npMgT permeation profiles were reconstructed mostly from the graphical
content published in indicated papers of the group around Quamme (green) (Quamme 2010;
Goytain and Quamme 2005a, b, c, d, 2008; Goytain et al. 2007, 2008a, b). Due to frequent lack
of statistics in the original works, it is not being reflected in our interpretation of these permeation
profiles
The expression profiles of the novel magnesiotropic genes in various human organs
or tissues are summarized in Table 3. At the level of mRNA, the majority of the
novel magnesiotropic genes are expressed in most of the tested organs/tissues
(Table 3). A ubiquitous and/or almost ubiquitous expression of MgGs might indicate
their importance for overall cellular physiology. However, at the protein level,
TRPM6, CNNM1, CNNM4, NIPAL4, HIP14, and HIP14L are only detected in
specific organs/tissues that have absorptive functions, re-absorptive functions, bar-
rier functions, or secretory functions, or in those that are metabolically highly active
(https://www.proteinatlas.org) (Romani 2011; Quamme 2010).
influx mechanism. Essentially similar conclusions have been drawn from the studies
of Jung et al. (1990, 1997). As an alternative to a regulated Mg2+ influx mechanism,
Kolisek and colleagues have proposed the existence of a Mg2+ efflux mechanism
capable of balancing for Mg2+ influx (Kolisek et al. 2003).
Kolisek and colleagues have identified Mrs2 (mitochondrial RNA splicing 2) to
be a channel responsible for Mg2+ flux mediation into mitochondria (Figs. 1, 3, and
4) (Kolisek et al. 2003). Furthermore, they have demonstrated that the Mrs2-
mediated Mg2+ influx is entirely driven by the mitochondrial membrane potential
(ΔψIMM) and inhibited by cobalt(III)hexaammine, an inhibitor of CorA (distant
prokaryotic homolog of Mrs2) (Kolisek et al. 2003). These findings have been
verified by the study of Schindl and coworkers who have shown by means of
electrophysiology that Mrs2 fulfills all the prerequisites for a superconductive
mitochondrial Mg2+ channel (Schindl et al. 2007). The depolarization of IMM
(and thus a decrease of ΔψIMM) leads to the inevitable suspension of Mrs2 function
(Kolisek et al. 2003). This is coherent with Mg2+ release in response to the depo-
larization of IMM (Kubota et al. 2005). Trapani and Wolf have quoted, in their
review, the work of Kubota and colleagues stating that Mrs2-mediated Mg2+ efflux
might take place in mitochondria upon the depolarization of IMM (Kubota et al.
2005; Trapani and Wolf 2015). Considering the enormous Δψ on IMM, a very
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 77
Fig. 3 Recycling of Mg2+, Na+, and Ca2+ on inner mitochondrial membrane. MCU mitochondrial
Ca2+ uniporter, NCE Na+/Ca2+ exchanger, NHE Na+/H+ exchanger, NME Na+/Mg2+ exchanger,
Mrs2 mitochondrial Mg2+ channel
strong ΔμMg and an almost negligible Δ[Mg2+] between the intermembrane space
[Mg2+]is and the matrix [Mg2+]m, scenario in which Mrs2 act as a Mg2+-efflux
mechanism is almost impossible, even during an event of drastic depolarization
(Kolisek et al. 2003; Vergun et al. 2003; Corkey et al. 1986). Furthermore, the
overall structure of the Mrs2 channel, the organization of the selective filter, and the
complex gating mechanism are not in support of the hypothesis that the channel may
conduct Mg2+ efflux under certain conditions (Sponder et al. 2013a; Khan et al.
2013).
The functionally inactive, mutant Mrs2 causes in dmy/dmy rats a mitochondrial
disease hallmarked by the demyelination of the neurons (Kuramoto et al. 2011;
Kuwamura et al. 2011).
The proton (H+) gradient across the IMM generates the major motive force
powering the transport of a plethora of solutes across this membrane. Thus, the
mitochondrial Mg2+ efflux system was assumed to be directly or indirectly coupled
to H+ influx. The experimental work of Rutter and colleagues using rat heart
mitochondria supports the existence of a mitochondrial H+/Mg2+ exchanger
(Fig. 4) (Rutter et al. 1990). Furthermore, long-chain fatty acids induce the
rapid release of Mg2+ from rat liver mitochondria in alkaline media, presumably
via a Mg2+/Me+ or an H+/Mg2+ exchanger (Schönfeld et al. 2002). However, until
78 M. Kolisek et al.
Fig. 4 The mitochondrial (IMM) Mg2+ transport circuit (Mrs2, SLC41A3, APC, and HME). APC
ATP-Mg/Pi carrier, HME H+/Mg2+ exchanger (?, mechanism encoded by yet unknown molecular
entity), IMM inner mitochondrial membrane, Mrs2 Mg2+ channel, NME Na+/Mg2+ exchanger
now the molecular identity of the mitochondrial transporter capable of Mg2+ efflux
via a mechanism of H+/Mg2+ exchange remains elusive.
Recently, the group of Sponder has provided experimental evidence that
npMgT SLC41A3 is a Na+-coupled Mg2+ efflux system (very likely a Na+/Mg2+
exchanger) that resides in the IMM (Figs. 1, 3, and 4) (Mastrototaro et al. 2016). This
finding is also supported by the observation that the 25Mg2+ uptake in wild type
(SLC41A3+/+) and SLC41A3/ mice is alike (de Baaij et al. 2016). The exact nature
of the Na+-Mg2+ transport coupling via SLC41A3 and the stoichiometry of this
process remains to be addressed. Perhaps, the further characterization of SLC4A3
will shed more light on the observation of Zhang and Melwin that the elevation of
[Na+]i induces an efflux of Mg2+ from intracellular pool(s) (mitochondria) in rat
sublingual mucous acini (Zhang and Melvin 1996).
Moreover, an increase of the classic cytosolic secondary messenger cAMP has
been reported to activate the release of Mg2+ from mitochondria (Romani et al.
1991). cAMP is an important cofactor/activator of PKA. Whether PKA plays a role
in the activation of SLC41A3 remains to be examined. However, the cAMP-
dependent activation of PKA followed by the phosphorylation of the plasma mem-
brane Na+/Mg2+ exchanger SLC41A1 has clearly been demonstrated as being an
essential step toward the release of Mg2+ from the cell (Mastrototaro et al. 2015;
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 79
Kolisek et al. 2013a; Sponder et al. 2017). Therefore, keeping in mind the shared
ancestry of SLC41A1 and SLC41A3, it would be unsurprising if SLC41A3 also
undergoes PKA-mediated activation (Fig. 5). Recently, this hypothesis has become
even more feasible because of the discovery of A-kinase-anchoring proteins
(AKAP), namely, AKAP121 and SPHKAP/SKIP (sphingosine kinase type-1
anchor/interacting protein), which are known to tether PKA to the outer mitochon-
drial membrane (OMM) and intermembrane space within the proximity of its local
targets (Fig. 5) (Feliciello et al. 2005; Kovanich et al. 2010; Means et al. 2011;
Lefkimmiatis et al. 2013).
The existence of an independent intramitochondrial cAMP signaling circuit
consisting of the bicarbonate-activated soluble adenylyl cyclase (sAC), PKA holo-
enzyme, and phosphodiesterase 2a (PDE2A) has also been reported (Lefkimmiatis
et al. 2013; Wuttke et al. 2001; Zippin et al. 2003; Sardanelli et al. 2006; Acin-Perez
et al. 2009, 2011a, b). Hence, strictly hypothetically, even a matrix-based cAMP-
PKA regulation of SLC41A3-mediated Mg2+ would be possible (Fig. 5).
Fig. 5 Strictly hypothetical model of SLC41A3 regulation via cAMP-PKA and Akt/PKB signaling
pathways in mitochondria. AC adenylyl cyclase, Akt/PKB protein kinase B, AKAP1 A-kinase-
anchoring protein 1, APC ATP-Mg/Pi carrier, CM cytoplasmic membrane, G glucagon, GP G
protein, HME H+/Mg2+ exchanger (?, mechanism encoded by yet unknown molecular entity), IMM
inner mitochondrial membrane, INS insulin, Mrs2 mitochondrial Mg2+ channel, NME Na+/Mg2+
exchanger, OMM outer mitochondrial membrane, PDE2a phosphodiesterase 2a, PKA protein
kinase A, RTK receptor tyrosine kinase, sAC soluble adenylyl cyclase, SKIP sphingosine kinase
type 1 interacting protein
80 M. Kolisek et al.
Insulin (INS) decreases the concentration of the cytosolic cAMP via the classic
INS-signaling cascade IR-PI3K-Akt/PKB by the activation of phosphodiesterase 3b
(PDE3b) and consequently limits the activation of the cytosolic PKA (Mastrototaro
et al. 2015).
INS promotes the efflux of Mg2+ from mitochondria (and perhaps also other
intracellular Mg2+ stores) via a cAMP-PKA-independent mechanism concomitant
with the translocation of PKB into mitochondria (Fig. 5) (Mastrototaro et al. 2015;
Bijur and Jope 2003). The exact nature of the INS action on the process of
mitochondrial Mg2+ release and its correlation with the translocation/accumulation
of Akt/PKB into mitochondria remain elusive. Without doubt, PKA and Akt/PKB
play fundamental roles in the regulation of cytosolic and mitochondrial [Mg2+] and
the regulation of mitochondrial homeostasis per se. Furthermore, both of these
kinases significantly influence cellular energetics and metabolic activity. Thus,
further research aimed at explaining the role of PKA and Akt/PKB in the regulation
of the mitochondrial accumulation of Mg2+ and its release from mitochondria is
urgently needed (Zhang et al. 2017a; Manning and Toker 2017).
Recently, van Ooijen and colleagues have explained a possible molecular back-
ground behind the involvement of [Mg2+]i and its oscillations in the regulation of the
circadian clock in eukaryotes (Feeney et al. 2016). Interestingly, PKA and Akt/PKB
have both been substantially implicated in the regulation of the circadian clock
(Huang et al. 2007; Noguchi et al. 2018; Luciano et al. 2018). Thus, we can assume
that except for SLC41A1, also the components of mitochondrial Mg2+ homeostasis
(Mrs2, SLC41A3 and APC) and their regulatory network (very likely PKA and
Akt/PKB) form an intricate biological mechanism making mitochondria not only
“a battery” but also “a tuning mechanism” for the circadian clock. Further research
into the intersection between mitochondrial homeostasis and cellular/mitochondrial
Mg2+ homeostasis is undoubtedly of great importance, especially for the field
of chronomedicine and for research into degenerative diseases, in which
CHRONO-component plays an important role.
ATP-Mg/Pi carrier (APC; or short Ca2+-binding mitochondrial carrier, SCaMC;
or solute carrier family 25 member A24, A25, A23, A41 (SLC25A24, SLC25A25,
SLC25A23, SLC25A41)), which is localized in the IMM, facilitates an
electroneutral reversible exchange between MgATP2 and HPO42 (Figs. 1 and
4) (Run et al. 2015; Traba et al. 2009; Joyal and Aprille 1992). APC remains inactive
unless stimulated by a cytosolic Ca2+ signal (Nosek et al. 1990). While under
physiological conditions, the preferred substrates for the APC are MgATP2 and
Pi (HPO42), but in the absence of Mg2+, ADP (HADP2) can also be transported
via the APC (Joyal and Aprille 1992; Tewari et al. 2012). Nevertheless, this
transporter does not transport ionized Mg2+ alone; complexed with ATP, its impact
on the modulation of mitochondrial or cytosolic [Mg2+] might be significant (Tewari
et al. 2012; Kun 1976). However, no information has been acquired about any
possible crosstalk between APC and other mitochondrial Mg2+ transporters.
The ADP/ATP carrier (AAC, or solute carrier family 25 member A4, A5, A6
(SLC25A4–6)) is the major carrier that exports ATP out of the matrix for energy
consumption while importing ADP for the production of new ATP by the ATP
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 81
synthase, and its functional defects can be detrimental for the cell (Run et al. 2015;
Fiore et al. 1998; Klingenberg 2008). Whereas AAC accounts for the bulk of
ADP/ATP recycling in the matrix, APC is important for mitochondrial activities in
the matrix that depend on adenine nucleotides, such as gluconeogenesis and mito-
chondrial biogenesis (Run et al. 2015). Quoting Romani, Trapani, and Wolf have
stated that, although the AAC substrate of choice is ATP, it might change to MgATP
in some cases, e.g., following cAMP or thyroid hormone stimulation (Romani 2011;
Trapani and Wolf 2015). However, this statement is contradicted by several works
concluding that MgATP or MgADP is not transported through AAC (Run et al.
2015; Pfaff et al. 1969; Nury et al. 2006). Therefore, currently, questions remain as
to whether AAC, alone or in the mitochondrial permeability transition pore complex,
contributes to Mg2+ transport across IMM (Karch and Molkentin 2014).
Pharmacological experiments with SNAP, 8-Br-cGMP, diazoxide, and several
inhibitors, performed by Yamanaka and colleagues, have revealed that the NO –
cGMP – protein kinase G (PKG) signaling pathway triggers an increase in [Mg2+]i
and that Mg2+ mobilization is attributable to the Mg2+ release from mitochondria
induced by mitoKATP channel opening (Yamanaka et al. 2013). Furthermore, Mg2+
release is potentiated by the positive feedback loop including mitoKATP channel
opening, mitochondrial depolarization, and PKC activation (Yamanaka et al. 2013).
Despite the ongoing debate about the physical existence of the mitoKATP channel,
the recreation of the experiments of Yamanaka and colleagues in rodent or human-
derived cells/cell lines with silenced expression or overexpression of mitochondrial
Mg2+ extruder SLC41A3 would be of interest (Yamanaka et al. 2013; Garlid and
Halestrap 2012).
Yeast (S. cerevisiae) Lpe10 (or MFM1; Mrs2 function modulating factor 1)
belongs to the CorA superfamily of Mg2+ transporters. Moreover, it is homologous
to yeast Mrs2 (approximately 32% sequence homology and presence of the G-M-N
Mg2+ binding motive) (Gregan et al. 2001b). Sponder and colleagues have utilized
an intriguing set of complementation experiments in order to demonstrate that both
Lpe10 and Mrs2 are functionally related and that they form complexes together but
cannot substitute for each other (Sponder et al. 2010a). Deletion of Lpe10 leads to a
rapid loss of the membrane potential on IMM, a phenomenon otherwise not seen
when only Mrs2 is deleted (Sponder et al. 2010a). Lpe10 alone is not able to mediate
the high-capacity Mg2+ influx otherwise seen with Mrs2. When coexpressed with
Mrs2, they yield a unique reduced Mg2+ conductance in comparison with that of
Mrs2 channels (Sponder et al. 2010a). A homolog of Lpe10 is not found in
mammalian/human mitochondria. However, we cannot exclude that, during the
evolution, the Mrs2-modulating role of Lpe10 was overtaken by an as yet unknown
mitochondrial protein innate to mammalian/human mitochondria. Only further
targeted research may address this issue.
Recently, a mitochondrial Mg2+ efflux system constituted by Mme1 (mitochon-
drial Mg2+ exporter 1) protein has been described in S. cerevisiae, followed by the
discovery of its ortholog, dMme1 protein, in Drosophila melanogaster (Cui et al.
2015, 2016). No human ortholog of Mme1 or dMme1, if any, has yet been identified.
An alteration in the expression of dMme1, although only resulting in a change of
82 M. Kolisek et al.
The high metabolic activity and the large estimated ΔψER of 75 to 95 mV across
the membrane of ER make this cellular compartment, similar to mitochondria, not
only “a Mg2+ consumer” but also a suitable compartment for storing Mg2+ (with a
total Mg concentration estimated as being between 14 and 18 mM) (Fig. 2) (Romani
2011; Qin et al. 2011). Unfortunately, among npMgTs characterized by Goytain and
Quamme’s group, no candidates have been predicted to localize to ER compartment
(Quamme 2010).
In the regnum of plants, Li and colleagues have mapped the Mg2+ transporter
AtMGT4 (Arabidopsis thaliana Mg2+ Transporter 4), a distant homolog of yeast
Mrs2 and a member of the CorA superfamily, to ER (Li et al. 2015). There is no
reason to suggest that the ER of animal cells differ from the ER of plant cells with
respect to the existence of ER-localized Mg2+ transporters/Mg2+ homeostatic factors.
Presently, many indirect indices argue for the existence of an ER-localized Mg2+
transporter in mammalian/human cellular systems; however, only future experimen-
tal attempts will uncover their identities.
ATP13A4, a member of the subfamily of P5-type ATPases, is thought to act as a
cation transporter and can serve as an example of a promising candidate (Schultheis
et al. 2004). Subcellularly, it localizes to the ER when expressed in COS-7 cells
(Fig. 1) (Vallipuram et al. 2010). Will and colleagues have hypothesized that
ATP13A4 transports Mg2+ (Will et al. 2010). However, the same authors stress
that the substrate specificity of ATP13A4 is as yet only poorly understood.
In contrast to ER, a screen performed by Goytain and Quamme’s group has revealed
the identities of four genes encoding for proteins putatively localized to the GA and
post-Golgi vesicles, namely, MMgT1, MMgT2 (not found in the human genome but
present in mouse), and HIP14 and HIP14L (Quamme 2010; Goytain and Quamme
2008; Goytain et al. 2008b).
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 83
In their most recent work, Maeshima and colleagues have concluded that Mg2+
released from the MgATP complex after its hydrolysis contributes to mitotic
chromosome condensation with increased rigidity, suggesting a novel regulatory
mechanism for higher-order chromatin organization by the intracellular MgATP
complex balance (Maeshima et al. 2018). Because of DNA and RNA metabolism,
which is known to be Mg2+-dependent, Mg2+ is of high demand by the nucleus
(Nishizawa et al. 2007). The major nuclear Mg2+ influx/efflux mechanism is
believed to be constituted by the nuclear pore that allows Mg2+ to move freely
across the porous nuclear membranes simply by diffusion (Romani 2011).
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 85
Member A2 of the solute carriers family 41 (SLC41A2), which like other members
of this family is distantly homologous to the bacterial Mg2+ transporter MgtE, was
predicted by Sahni and colleagues to reside in cytoplasmic membrane, as based on
results from experimental evidence (Sahni et al. 2007). The protein comprises
11 transmembrane helices with an “N-terminus outside and C-terminus inside”
orientation of its termini (Sahni et al. 2007). Since the orientation of SLC41A2
appears to be the opposite of that predicted by the structure of prokaryotic MgtE
proteins or SLC41A1, the same group has subsequently speculated that the observed
plasma membrane localization of SLC41A2 reflects aberrant cell surface targeting
attributable to the overexpression of the protein (Sahni et al. 2007; Sahni and
Scharenberg 2013). Despite the targeting and placement of SLC41A2 to the cyto-
plasmic membrane having not been completely rejected, the hypothesis that
SLC41A2 plays its roles in intracellular membranous compartments and vesicles
is currently better accepted (Sahni et al. 2007; Sahni and Scharenberg 2013).
Similar to other npMgTs functionally characterized by the group of Goytain
and Quamme, SLC41A2, when overexpressed in X. laevis oocytes, also
conducts the electrogenic transport of Mg2+ and of other cations (Table 2) (Quamme
2010; Goytain and Quamme 2005b). Sahni and colleagues have attempted to
identify Mg2+-specific currents in DT40 cells overexpressing human SLC41A2
with patch clamp; however, these attempts remain unsuccessful (Sahni et al. 2007).
Although the data on the mode of Mg2+ transport via SLC41A2 are puzzling, we
can safely say that, with high probability, SLC41A2 is a Mg2+ transporter. This
assumption is underpinned by the ability of SLC41A2 to complement the growth/
proliferation defect of DT40 TRPM7-KO cells, when they are cultured at physio-
logical [Mg2+]e. TRPM7-deficient DT40 cells induced to express SLC41A2 prolif-
erate more slowly than wild-type DT40 cells in culture medium containing
physiological levels of Mg2+, but in contrast to the TRPM7-deficient cells, they
are able to proliferate continuously (Sahni et al. 2007). Furthermore, Sahni
and colleagues have determined the Mg2+ uptake (measured by the ratio of 26Mg2+/
24
Mg2+) to be approximately twofold to threefold higher in SLC41A2-
overexpressing cells as compared with the cells without induced expression of
transgenic SLC41A2 and even greater than that observed in the wild-type DT40
cells (Sahni et al. 2007). The last mentioned results provide direct evidence that
SLC41A2 overexpression correlates with enhanced Mg2+ accumulation.
Whether the gene SLC41A2 is truly magnesiotropic remains controversial.
Goytain and Quamme have used real-time RT-PCR analysis of MDCT cells cultured
in medium with [Mg2+]e of 1 mM or in Mg2+ free medium for 16 h and found no
change in SLC41A2 expression; the same result was found to be the case with kidney
cortical tissue harvested from mice fed with a normal or low Mg2+ diet for 5 days
86 M. Kolisek et al.
4.6.2 The CNNM (Cyclin and CBS Domain Divalent Metal Cation
Transport Mediator) Family, “Enfant terrible” Among Mg2+
Transporters/Mg2+ Homeostatic Factors
The CNNM family consists of four members (CNNM1–CNNM4) that were for-
merly known as ancient conserved domain proteins (ACDP). Their original name
was based on the observation that these genes/proteins possessed a highly conserved
domain also found in other species from bacteria to mammals. The most conserved
regions of CNNMs are the two cystathionine beta-synthase (CBS) domains and the
DUF21 domain that are both also found in bacterial CorC (Wang et al. 2003).
The high homology of ACD proteins to the bacterial CorC protein which is
involved in Mg2+ and Co2+ efflux early led to the speculation that these proteins
are involved in ion transport (Gibson et al. 1991; Wang et al. 2004). Although
sequence conservation among the family members is remarkably high (for the
human and mouse family members: 55.3% of amino acid identity and 83.3% of
amino acid homology in the conserved region), their expression pattern, localization
within the cell (Fig. 1), and their function appear to be divergent and are still a matter
of debate (Wang et al. 2004).
CNNM2 is the best studied representative of this family and, at the same time, is the
most controversial. Wang and colleagues have demonstrated the expression of
CNNM2 in most mouse tissues; however, the expression levels are generally low,
except in the brain, kidney, and liver (Wang et al. 2003). This finding has subse-
quently been confirmed in mice by Goytain and Quamme who have found the
highest expression in kidney and brain (Goytain and Quamme 2005d). Abundant
amounts have also been detected in the heart and liver, with lower expression in the
small intestine and colon. As mentioned previously, the expression of CNNM2 is
influenced by the food Mg content and displays increased expression in the kidneys
of mice fed with a low Mg diet and in cultured MDCT cells cultured in low Mg
medium (Goytain and Quamme 2005d). Mouse CNNM2, when expressed in
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 87
Xenopus oocytes, exhibits currents for Mg2+ and other divalent cations (Table 2) and
therefore behaves as a rather unspecific cation transporter. However, as the Km was
within a physiological range only for Mg2+, Goytain and Quamme concluded that
CNNM2 primarily acts as a Mg2+ transporter (Quamme 2010; Goytain and Quamme
2005d). The assumption that CNNM2 acts as a Mg2+ transporter seems to be
supported by a study of Sponder and colleagues who have demonstrated that the
longest splice variant of human CNNM2 (875 amino acids) partially complements
the Mg-dependent growth defect of Salmonella strains MM281 (deficient for MgtA,
MgtB, and CorA). Only the somewhat shorter splice variant 2 fails to do so (Sponder
et al. 2010b).
Further evidence for the involvement of CNNM2 in Mg2+ transport has
come from a study that has identified a connection between CNNM2 and
dominant hypomagnesemia. This rare disorder is characterized by severely lowered
serum Mg2+ levels most probably caused by defective tubular reabsorption. Such a
notion is also supported by the localization of CNNM2 on the basolateral membrane
of DCT cells and its strong expression in the kidney (Stuiver et al. 2011; de Baaij
et al. 2012).
Interestingly, the electrophysiological characterization of CNNM2 in HEK293
cells yields Mg2+-sensitive Na+ currents and thereby divergent results from those
observed in Xenopus oocytes. The authors have consequently concluded that
CNNM2 is a regulatory factor rather than a direct mediator of Mg2+ transport
(Stuiver et al. 2011). Currents for Na+ have also been reported by another group
when CNNM2 is overexpressed in HEK293 cells (Yamazaki et al. 2013).
Arjona and colleagues have used zebrafish as a model system to study the
function of CNNM2 and have reported disturbed brain development and reduced
body Mg content as consequences of the deletion of the gene. They have furthermore
demonstrated, with the aid of stable isotope 25Mg2+, that the overexpression of
mouse CNNM2 in HEK293 cells increases the cellular Mg content. However,
since this effect is abolished by the TRPM7 inhibitor 2 APB, the authors have
concluded that this effect is indirect and presumably mediated by a regulatory
effect of CNNM2 on the Mg2+ permeable channel TRPM7 (Arjona et al. 2014).
This is in contrast to the data obtained by Hirata and colleagues with the Mg2+
indicator Magnesium Green, which has shown the strong and remarkably fast efflux
of Mg2+ within seconds when CNNM2 is expressed in HEK293 cells. They have
further found the CBS domains that directly bind ATP in a Mg2+-dependent manner
as being indispensable for Mg2+ efflux (Hirata et al. 2014).
The same group of Miki has subsequently used mice to investigate the function of
CNNM2 in an animal model. Homozygous knockout mice have an embryonic lethal
phenotype; heterozygous mice are viable and exhibit impaired Mg reabsorption with
reduced serum Mg levels. The same phenotype is observed in mice with a kidney-
specific deletion of CNNM2 (Funato et al. 2017).
In contrast to these functional data from mice, Sponder and colleagues have
employed mag-fura-2, Mg2+-sensitive fluorescent dye, in an in vitro model to
investigate the Mg2+ transport activity of two of the three known splice variants of
human CNNM2 under conditions favoring both Mg2+ influx and efflux. Although
88 M. Kolisek et al.
the same expression system, HEK293 cells as in previous studies, were used neither
electroneutral nor was electrogenic Mg2+ transport detected for the two splice
variants in the overexpressing cells. The authors have concluded that CNNM2 acts
as Mg2+ homeostatic factor without being a Mg2+ transporter itself (Sponder et al.
2016).
Recently, a role of CNNM2 in tumorigenesis has been proposed. The protein
interacts with PRL-1, a member of the so-called phosphatases of regenerating liver;
these phosphatases exhibit high expression in most solid tumors and hematological
cancers and are considered to be highly oncogenic. Binding between the two
proteins is mediated via the interaction of an amino acid in the CBS domain of
CNNM2 and the catalytic domain of the phosphatase. The authors speculate that this
interaction increases the cellular magnesium content, thereby aggravating tumor
progression and metastasis formation (Giménez-Mascarell et al. 2017).
In summary, great efforts have been undertaken to elucidate the molecular
function of CNNM2. In view of the connection between mutations in CNNM2
and patients suffering from hypomagnesemia and knockdown experiments in
mouse and zebrafish, the involvement of CNNMs in cellular Mg2+ homeostasis is
unequivocal. However, the strongly divergent behavior of CNNM2 in various
expression systems and experimental setups complicates rather than clarifies the
situation and has led to much conflicting and contradictory data. A clear conclusion
cannot therefore be drawn about the function of this protein in cellular Mg2+
homeostasis.
CNNM4 has a relatively broad expression pattern with its highest expression in
intestinal epithelia (Wang et al. 2003; de Baaij et al. 2012).
The first indication for a function in metal ion homeostasis came from the
observation that the expression of the protein in HEK293 cells resulted in an
increased sensitivity to copper, manganese, and cobalt. The toxicity of these metal
ions was aggravated by the coexpression of CNNM4 together with COX11
suggesting a possible functional dependence of CNNM4 on other proteins (Guo
et al. 2005). Mutations in CNNM4 do not influence blood Mg concentrations and
cause Jalili syndrome in humans, a combination of recessively inherited cone-rod
dystrophy and amelogenesis imperfecta. Given the broad expression pattern, the
findings that only the retina and ameloblasts are affected by the expression of mutant
variants of CNNM4 and that therefore phenotypic consequences are restricted to
retinal function and tooth biomineralization are of interest (Parry et al. 2009).
CNNM4 has also appeared together with CNNM2 and CNNM3 in a genome-
wide association study on common single nucleotide polymorphisms SNPs being
associated with serum magnesium concentrations (Meyer et al. 2010).
Direct evidence for an involvement of CNNM4 in Mg2+ homeostasis has come
from Miki’s group who have characterized CNNM4 in mice. Expression of the
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 89
protein has been found to be high in the intestine where it localizes to the basolateral
membrane. No expression in the kidney has been detected. CNNM4 knockout mice
are viable with no obvious phenotype. However, the animals exhibit lower serum
Mg concentration, a result that has been attributed to impaired intestinal Mg2+
absorption. The transport function was directly assessed by the use of the Mg2+
indicator Magnesium Green. When CNNM4 was expressed in HEK293 cells,
the protein mediated a very rapid efflux of Mg2+ that was dependent on extracellular
Na+. Similar to CNNM2, the CBS domains are also essential for the transport
activity of CNNM4 (Hirata et al. 2014). From their data, the authors conclude that
CNNM4 forms a high capacity Mg2+ efflux system that is localized in the basolateral
membrane of the intestine and that acts as a counterpart to the TRPM6/7 channel-
mediated Mg2+ uptake system on the apical side. Of note is the observation that the
knockout of CNNM4 in mice results in amelogenesis imperfecta. However, in
contrast to descriptions of Jalili syndrome in humans, the retina remains unaffected
(Yamazaki et al. 2013).
CNNM1 and CNNM3 (Cyclin and CBS Domain Divalent Metal Cation
Transport Mediator3 and 4)
Knowledge about the two other members of this family, namely, CNNM1 and
CNNM3, is still scarce. Wang et al. have reported the high expression of CNNM1
in the brain, whereas only low levels have been detected in the kidney, testis, and
most other tested tissues. For CNNM3, the highest expression has been observed in
the brain, kidney, liver, and heart and very low levels in skeletal muscles (Wang et al.
2003). Together with CNNM2 and CNNM4, CNNM3 has been also found in the
aforementioned genome-wide association study as being linked to the serum mag-
nesium concentration (Meyer et al. 2010). Studies directly investigating the possible
transport activity of the two proteins are rare. In a study on the pufferfish Takifugu
obscurus, CNNM3 was upregulated in the kidney when the animals were kept in salt
water. Immunohistochemical investigations revealed expression in the proximal
tubule where CNNM3 was localized to the lateral membrane. The expression of
CNNM3 in Xenopus oocytes resulted in a significant decrease of the cellular Mg2+
concentration (Islam et al. 2014).
The only study in a mammalian model system was performed in the laboratory of
Miki. They investigated the importance of the CBS domains for all four members of
the CNNM family.
CNNMs were expressed in HEK293 cells, and Magnesium Green was used for
intracellular Mg2+ detection. In contrast to the aforementioned strong activity of
CNNM2 and CNNM4, only a weak efflux was detected for CNNM1, and no activity
in cells expressing CNNM3 was observed in this experimental setup (Hirata et al.
2014).
Interestingly, for CNNM3 an interaction with two members of the aforemen-
tioned phosphatases of regenerating liver (PRLs), namely, PRL-2 and PRL-3, has
also been reported suggesting a role of CNNM3 in tumor development and progres-
sion (Zhang et al. 2017b).
90 M. Kolisek et al.
Several Mg2+ influx and/or efflux mechanisms have been foreseen to exist in the
cytoplasmic membrane (Nishizawa et al. 2007; Romani 2011; Schweigel et al.
2000). Presently, only chanzymes TRPM6/7 and Na+/Mg2+ exchanger SLC41A1
are well characterized at the molecular level (Schweigel-Röntgen and Kolisek 2014;
Penner and Fleig 2007; Sponder et al. 2017; Cabezas-Bratesco et al. 2015). It is
universally accepted that both chanzymes represent a channel component responsi-
ble for the transport of the majority of extracellular Mg2+ into the cell.
In vitro, also SLC41A1 has been shown to mediate Mg2+ uptake under conditions
strongly supporting Mg2+ influx. However, it has to be stressed out that these
conditions were far from physiological to most of the tested cell types (Schweigel-
Röntgen and Kolisek 2014; Kolisek et al. 2008, 2012; Fleig et al. 2013). Whether
SLC41A1 mediates Mg2+ influx also in vivo under physiological conditions is
unclear and a subject of ongoing debates.
Unequivocally, confirmed by several independent and unbiased studies, SLC41A1
was shown to be an ubiquitously expressed (Table 3), major cellular Mg2+ efflux
system functionally conserved from fish to Man (Schweigel-Röntgen and Kolisek
2014; Kolisek et al. 2008, 2012; Fleig et al. 2013; Islam et al. 2013; Hurd et al. 2013;
Lin et al. 2014).
Thus, it could be stated with confidence that TRPM6/7 chanzymes together with
Mg2+ efflux-mediating carrier SLC41A1 constitute the Mg2+ transport circuit of
cytoplasmic membrane (Fig. 6).
Fig. 6 Current model of the Mg2+ transport circuit on cytoplasmic membrane (CM). It consists of a
Mg2+ ion channel allowing for Mg2+ influx (chanzymes TRPM6/7) and a Na+/Mg2+ exchanger
(NME; SLC41A1) mediating Mg2+ efflux
4.7.2 KelI/XK
Although primarily expressed in erythroid tissues, Kell and XK are also present in
many other tissues (Table 3) (Rivera et al. 2013; Lee et al. 2000). XK and Kell are
linked close to the membrane surface by a single disulfide bond between Kell
cysteine 72 and XK cysteine 347 (Fig. 1) (Lee et al. 2000). While Kell is an
endothelin-3-converting enzyme, XK has structural characteristics of prokaryotic
and eukaryotic transport proteins (Ho et al. 1994; Zhu et al. 2009). Rivera and
colleagues studied the activity of Na+/Mg2+ exchanger in erythrocytes of Kell-KO,
XK-KO, and Kell/XK-KO mice (Rivera et al. 2013). They found that Na+/Mg2+
exchange was significantly reduced by the absence of XK and increased in Kell-KO
animals compared to wild-type mice. However, in Kell/XK-KO, the Na+/Mg2+
exchange activity resembled that of Kell-KO Na+/Mg2+ exchange activity
suggesting that Kell may act as a Na+/Mg2+ exchanger regulatory unit (Rivera
et al. 2013).
Deletion or loss of function mutations within XK leads to McLeod syndrome.
It is a rare and progressive disease that shares important similarities with Huntington
disease but has widely varied neurologic, neuromuscular, and cardiologic manifes-
tations. Patients with McLeod syndrome have a distinct hematologic presentation
with specific transfusion requirements (Roulis et al. 2018).
The human NIPA family comprises four members, namely, NIPA1, NIPA2,
NIPAL1 (NIPA3), and NIPAL4 (NIPA4). Members 1–4 (mouse) of the NIPA
family have been characterized as putative Mg2+ transporters with channel-like
properties when expressed in X. laevis oocytes and characterized with TEV and
mag-fura-2-assisted fluorescence spectrometry (Table 1) (Quamme 2010; Goytain
et al. 2007, 2008a). NIPA1–NIPA4 mediate Mg2+ uptake that is electrogenic,
94 M. Kolisek et al.
voltage-dependent, and saturable with a Km ¼ 0.31, 0.66, 0.9, and 0.36, respectively
(Quamme 2010; Goytain et al. 2007, 2008a). All four NIPA proteins exhibit quite
broad substrate specificity and thus behave like unspecific cation channels (Table 2)
(Quamme 2010; Goytain et al. 2007, 2008a). NIPA1–NIPA4 seem to be ubiqui-
tously expressed, but NIPA1 has its highest expression level in the brain (Table 3).
The alteration of [Mg2+]i induces the redistribution of NIPA1 and NIPA2 between
the endosomal compartment and the plasma membrane (Fig. 1). High [Mg2+]e results
in diminished cell surface NIPA1 and NIPA2, whereas low [Mg2+]e leads to an
accumulation of NIPA1 and NIPA2 in early endosomes and their recruitment to the
plasma membrane (Quamme 2010; Goytain et al. 2007, 2008a).
Unfortunately, all NIPA-related Mg2+ transport data have been acquired in a
heterologous expression system, and thus we cannot exclude that NIPA1–NIPA4 do
not transport Mg2+ in homologous expression systems.
NIPA1 has been implicated in autosomal-dominant hereditary spastic paraplegia
and NIPA2 in childhood absence epilepsy (Svenstrup et al. 2011; Xie et al. 2014). If
the Mg2+ transport function of NIPA proteins is confirmed in a homologous
expression system, this would represent direct evidence of the involvement of
cellular Mg homeostasis in the pathoetiology of these serious neurological ailments.
5 Summary
However, with hindsight, the selection of X. laevis oocytes for the characteriza-
tion of mammalian npMgTs (encoded by MgGs) in combination with TEV was not
the best move. Unfortunately, not a single mammalian/human npMgT assumed to be
an ion channel transporting Mg2+ (apart of other cations) based on TEV character-
ization in X. laevis oocytes has proven to be one in homologous expression systems.
Thus, strictly speaking, out of the total of 24 candidate Mg2+ transporters/Mg2+
homeostatic factors, only approximately one third has been experimentally
established as being directly involved in the Mg2+ homeostatic network of the cell,
either as Mg2+ transporters or as Mg2+ homeostatic factors. The other two thirds have
been studied only poorly, sometimes being limited to a single report.
Some of the pnMgTs have over the time proven not to be Mg2+ transporters at all.
For example, MagT1 and TUSC3/N33 are still often quoted as being Mg2+ channels,
despite the fact that they have never been shown to be channels in a homologous
expression system. Moreover, large amounts of solid evidence exist showing that
MagT1 and TUSC3/N33 are oxidoreductases of the ER-localized OST complex.
Nowadays, the proteins constituting the cytoplasmic membrane Mg2+ transport
circuit (TRPM6/7 and SLC41A1; Fig. 6) and the mitochondrial (IMM) Mg2+
transport circuit (Mrs2, SLC41A3 and APC/SCaMC; Figs. 3 and 4) have been
demonstrated by independent studies to be the primary constituents of the cellular
Mg2+ homeostasis. They all act like Mg2+ transporters (channels, carriers/
exchangers). The position of CNNM2 and CNNM4 in cellular Mg2+ homeostasis
is a matter of controversy, and the debate continues as to whether these proteins
are Mg2+ transporters per se or whether they play the role of “true” Mg2+ homeo-
static factors without the ability to transport Mg2+ (meaning sensors of [Mg2+]e
and/or [Mg2+]i and regulators of other components of Mg2+ homeostatic network
that are transporting Mg2+).
Nevertheless, no doubt exists that CNNM2 is crucial for Mg2+ homeostasis at the
level of the cell and also of the organism, as mutations in the gene encoding for
CNNM2 have been identified that lead to severe systemic hypomagnesaemia and
renal Mg2+ wasting.
The other putative Mg2+ transporters/homeostatic factors described in this man-
uscript must be robustly researched before uncertainties can be lifted to enable a
comprehensive elucidation of their functions with respect to cellular Mg2+
homeostasis.
Acknowledgment Our gratitude is due to Dr. Theresa Jones for linguistic corrections. This work
was supported by the grant “Return Home” issued by the Government of Slovak Republic to MK
and also by the project “Creating a New Diagnostic Algorithm for Selected Cancer Diseases”
(ITMS: 26220220022) co-financed from EU sources and the European Regional
Development Fund.
The authors declare no conflict of interests. All authors read and approved the final version of
the manuscript.
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 97
Contributions MK and GS wrote the manuscript, and IP, MC, ZT, TW, and PR contributed to the
manuscript writing.
References
Acin-Perez R, Salazar E, Kamenetsky M, Buck J, Levin LR, Manfredi G (2009) Cyclic AMP
produced inside mitochondria regulates oxidative phosphorylation. Cell Metab 9(3):265–276
Acin-Perez R, Russwurm M, Günnewig K, Gertz M, Zoidl G, Ramos L, Buck J, Levin LR,
Rassow J, Manfredi G, Steegborn C (2011a) A phosphodiesterase 2A isoform localized to
mitochondria regulates respiration. J Biol Chem 286(35):30423–30432
Acin-Perez R, Gatti DL, Bai Y, Manfredi G (2011b) Protein phosphorylation and prevention of
cytochrome oxidase inhibition by ATP: coupled mechanisms of energy metabolism regulation.
Cell Metab 13(6):712–719
Arjona FJ, de Baaij JH, Schlingmann KP, Lameris AL, van Wijk E, Flik G, Regele S, Korenke GC,
Neophytou B, Rust S, Reintjes N, Konrad M, Bindels RJ, Hoenderop JG (2014) CNNM2
mutations cause impaired brain development and seizures in patients with hypomagnesemia.
PLoS Genet 10(4):e1004267
Bai Y, Dong L, Huang X, Zheng S, Qiu P, Lan F (2017) Associations of rs823128, rs1572931, and
rs823156 polymorphisms with reduced Parkinson’s disease risks. Neuroreport 28(14):936–941
Bijur GN, Jope RS (2003) Rapid accumulation of Akt in mitochondria following
phosphatidylinositol 3-kinase activation. J Neurochem 87(6):1427–1435
Bui DM, Gregan J, Jarosch E, Ragnini A, Schweyen RJ (1999) The bacterial magnesium transporter
CorA can functionally substitute for its putative homologue Mrs2p in the yeast inner mitochon-
drial membrane. J Biol Chem 274(29):20438–20443
Butland SL, Sanders SS, Schmidt ME, Riechers SP, Lin DT, Martin DD, Vaid K, Graham RK,
Singaraja RR, Wanker EE, Conibear E, Hayden MR (2014) The palmitoyl acyltransferase
HIP14 shares a high proportion of interactors with huntingtin: implications for a role in the
pathogenesis of Huntington’s disease. Hum Mol Genet 23(15):4142–4160
Cabezas-Bratesco D, Brauchi S, González-Teuber V, Steinberg X, Valencia I, Colenso C (2015)
The different roles of the channel-kinases TRPM6 and TRPM7. Curr Med Chem
22(25):2943–2953
Cherepanova NA, Gilmore R (2016) Mammalian cells lacking either the cotranslational or
posttranslocational oligosaccharyltransferase complex display substrate-dependent defects in
asparagine linked glycosylation. Sci Rep 6:20946
Cherepanova N, Shrimal S, Gilmore R (2016) N-linked glycosylation and homeostasis of the
endoplasmic reticulum. Curr Opin Cell Biol 41:57–65
Corkey BE, Duszynski J, Rich TL, Matschinsky B, Williamson JR (1986) Regulation of free and
bound magnesium in rat hepatocytes and isolated mitochondria. J Biol Chem 261(6):2567–2574
Cui Y, Zhao S, Wang J, Wang X, Gao B, Fan Q, Sun F, Zhou B (2015) A novel mitochondrial
carrier protein Mme1 acts as a yeast mitochondrial magnesium exporter. Biochim Biophys Acta
1853(3):724–732
Cui Y, Zhao S, Wang X, Zhou B (2016) A novel Drosophila mitochondrial carrier protein acts as a
Mg2+ exporter in fine-tuning mitochondrial Mg2+ homeostasis. Biochim Biophys Acta
1863(1):30–39
de Baaij JH, Stuiver M, Meij IC, Lainez S, Kopplin K, Venselaar H, Müller D, Bindels RJ,
Hoenderop JG (2012) Membrane topology and intracellular processing of cyclin M2
(CNNM2). J Biol Chem 287(17):13644–13655
de Baaij JH, Hoenderop JG, Bindels RJ (2015) Magnesium in man: implications for health and
disease. Physiol Rev 95(1):1–46
98 M. Kolisek et al.
de Baaij JH, Arjona FJ, van den Brand M, Lavrijsen M, Lameris AL, Bindels RJ, Hoenderop JG
(2016) Identification of SLC41A3 as a novel player in magnesium homeostasis. Sci Rep
6:28565
Delva PT, Pastori C, Degan M, Montesi GD, Lechi A (1996) Intralymphocyte free magnesium in a
group of subjects with essential hypertension. Hypertension 28(3):433–439
Delva P, Degan M, Trettene M, Lechi A (2006) Insulin and glucose mediate opposite intracellular
ionized magnesium variations in human lymphocytes. J Endocrinol 190(3):711–718
Dragileva E, Rubinstein S, Breitbart H (1999) Intracellular Ca2+-Mg2+-ATPase regulates calcium
influx and acrosomal exocytosis in bull and ram spermatozoa. Biol Reprod 61(5):1226–1234
Ducker CE, Stettler EM, French KJ, Upson JJ, Smith CD (2004) Huntingtin interacting protein 14 is
an oncogenic human protein: palmitoyl acyltransferase. Oncogene 23(57):9230–9237
Ebel H, Hollstein M, Günther T (2002) Role of the choline exchanger in Na+-independent Mg2+
efflux from rat erythrocytes. Biochim Biophys Acta 1559(2):135–144
Feeney KA, Hansen LL, Putker M, Olivares-Yañez C, Day J, Eades LJ, Larrondo LF, Hoyle NP,
O’Neill JS, van Ooijen G (2016) Daily magnesium fluxes regulate cellular timekeeping and
energy balance. Nature 532(7599):375–379
Feliciello A, Gottesman ME, Avvedimento EV (2005) cAMP-PKA signaling to the mitochondria:
protein scaffolds, mRNA and phosphatases. Cell Signal 17(3):279–287
Fiore C, Trézéguet V, Le Saux A, Roux P, Schwimmer C, Dianoux AC, Noel F, Lauquin GJ,
Brandolin G, Vignais PV (1998) The mitochondrial ADP/ATP carrier: structural, physiological
and pathological aspects. Biochimie 80(2):137–150
Flatman PW (1984) Magnesium transport across cell membranes. J Membr Biol 80(1):1–14
Fleig A, Schweigel-Röntgen M, Kolisek M (2013) Solute carrier family SLC41, what do we really
know about it? WIREs Membr Transport Signaling 2(6). https://doi.org/10.1002/wmts.95
Funato Y, Yamazaki D, Miki H (2017) Renal function of cyclin M2 Mg2+ transporter maintains
blood pressure. J Hypertens 35(3):585–592
Garlid KD, Halestrap AP (2012) The mitochondrial K(ATP) channel – fact or fiction? J Mol Cell
Cardiol 52(3):578–583
Gibson MM, Bagga DA, Miller CG, Maguire ME (1991) Magnesium transport in Salmonella
typhimurium: the influence of new mutations conferring Co2+ resistance on the CorA Mg2+
transport system. Mol Microbiol 5(11):2753–2762
Giménez-Mascarell P, Oyenarte I, Hardy S, Breiderhoff T, Stuiver M, Kostantin E, Diercks T, Pey
AL, Ereño-Orbea J, Martínez-Chantar ML, Khalaf-Nazzal R, Claverie-Martin F, Müller D,
Tremblay ML, Martínez-Cruz LA (2017) Structural basis of the oncogenic interaction of
phosphatase PRL-1 with the magnesium transporter CNNM2. J Biol Chem 292(3):786–801
Goldin AL (2006) Expression of ion channels in Xenopus oocytes. In: Clare JJ, Trezise DJ (eds)
Expression and analysis of recombinant ion channels. Wiley, Weinheim
Goytain A, Quamme GA (2005a) Functional characterization of human SLC41A1, a Mg2+ trans-
porter with similarity to prokaryotic MgtE Mg2+ transporters. Physiol Genomics 21(3):337–342
Goytain A1, Quamme GA (2005b) Functional characterization of the mouse solute carrier,
SLC41A2. Biochem Biophys Res Commun 330(3):701–705
Goytain A, Quamme GA (2005c) Identification and characterization of a novel mammalian Mg2+
transporter with channel-like properties. BMC Genomics 6:48
Goytain A, Quamme GA (2005d) Functional characterization of ACDP2 (ancient conserved
domain protein), a divalent metal transporter. Physiol Genomics 22(3):382–389
Goytain A, Quamme GA (2008) Identification and characterization of a novel family of membrane
magnesium transporters, MMgT1 and MMgT2. Am J Physiol Cell Physiol 294(2):C495–C502
Goytain A, Hines RM, El-Husseini A, Quamme GA (2007) NIPA1(SPG6), the basis for autosomal
dominant form of hereditary spastic paraplegia, encodes a functional Mg2+ transporter. J Biol
Chem 282(11):8060–8068
Goytain A, Hines RM, Quamme GA (2008a) Functional characterization of NIPA2, a selective
Mg2+ transporter. Am J Physiol Cell Physiol 295(4):C944–C953
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 99
Goytain A, Hines RM, Quamme GA (2008b) Huntingtin-interacting proteins, HIP14 and HIP14L,
mediate dual functions, palmitoyl acyltransferase and Mg2+ transport. J Biol Chem 283
(48):33365–33374
Graschopf A, Stadler JA, Hoellerer MK, Eder S, Sieghardt M, Kohlwein SD, Schweyen RJ (2001)
The yeast plasma membrane protein Alr1 controls Mg2+ homeostasis and is subject to Mg2+-
dependent control of its synthesis and degradation. J Biol Chem 276(19):16216–16222
Gregan J, Kolisek M, Schweyen RJ (2001a) Mitochondrial Mg2+ homeostasis is critical for group II
intron splicing in vivo. Genes Dev 15(17):2229–2237
Gregan J, Bui DM, Pillich R, Fink M, Zsurka G, Schweyen RJ (2001b) The mitochondrial inner
membrane protein Lpe10p, a homologue of Mrs2p, is essential for magnesium homeostasis and
group II intron splicing in yeast. Mol Gen Genet 264(6):773–781
Groisman EA, Hollands K, Kriner MA, Lee EJ, Park SY, Pontes MH (2013) Bacterial Mg2+
homeostasis, transport, and virulence. Annu Rev Genet 47:625–646
Günther T (1993) Mechanisms and regulation of Mg2+ efflux and Mg2+ influx. Miner Electrolyte
Metab 19(4–5):259–265
Günther T (2006a) Concentration, compartmentation and metabolic function of intracellular free
Mg2+. Magnes Res 19(4):225–236
Günther T (2006b) Mechanisms, regulation and pathologic significance of Mg2+ efflux from
erythrocytes. Magnes Res 19(3):190–198
Günther T (2007) Na+/Mg2+ antiport in non-erythrocyte vertebrate cells. Magnes Res 20(2):89–99
Günther T, Vormann J, Cragoe EJ Jr (1990) Species-specific Mn2+/Mg2+ antiport from Mg2+-
loaded erythrocytes. FEBS Lett 261(1):47–51
Guo D, Ling J, Wang MH, She JX, Gu J, Wang CY (2005) Physical interaction and functional
coupling between ACDP4 and the intracellular ion chaperone COX11, an implication of the role
of ACDP4 in essential metal ion transport and homeostasis. Mol Pain 1:15
Hirata Y, Funato Y, Takano Y, Miki H (2014) Mg2+-dependent interactions of ATP with the
cystathionine-β-synthase (CBS) domains of a magnesium transporter. J Biol Chem 289
(21):14731–14739
Hmiel SP, Snavely MD, Miller CG, Maguire ME (1986) Magnesium transport in Salmonella
typhimurium: characterization of magnesium influx and cloning of a transport gene. J Bacteriol
168(3):1444–1450
Hmiel SP, Snavely MD, Florer JB, Maguire ME, Miller CG (1989) Magnesium transport in
Salmonella typhimurium: genetic characterization and cloning of three magnesium transport
loci. J Bacteriol 171(9):4742–4751
Ho M, Chelly J, Carter N, Danek A, Crocker P, Monaco AP (1994) Isolation of the gene for
McLeod syndrome that encodes a novel membrane transport protein. Cell 77(6):869–880
Huang G, Chen S, Li S, Cha J, Long C, Li L, He Q, Liu Y (2007) Protein kinase A and casein
kinases mediate sequential phosphorylation events in the circadian negative feedback loop.
Genes Dev 21(24):3283–3295
Hurd TW, Otto EA, Mishima E, Gee HY, Inoue H, Inazu M, Yamada H, Halbritter J, Seki G,
Konishi M, Zhou W, Yamane T, Murakami S, Caridi G, Ghiggeri G, Abe T, Hildebrandt F
(2013) Mutation of the Mg2+ transporter SLC41A1 results in a nephronophthisis-like pheno-
type. J Am Soc Nephrol 24(6):967–977
Islam Z, Hayashi N, Yamamoto Y, Doi H, Romero MF, Hirose S, Kato A (2013) Identification and
proximal tubular localization of the Mg2+ transporter, Slc41a1, in a seawater fish. Am J Physiol
Regul Integr Comp Physiol 305(4):R385–R396
Islam Z, Hayashi N, Inoue H, Umezawa T, Kimura Y, Doi H, Romero MF, Hirose S, Kato A
(2014) Identification and lateral membrane localization of cyclin M3, likely to be involved in
renal Mg2+ handling in seawater fish. Am J Physiol Regul Integr Comp Physiol 307(5):R525–
R537
Iwatsuki H, Lu YM, Yamaguchi K, Ichikawa N, Hashimoto T (2000) Binding of an intrinsic
ATPase inhibitor to the F1F0ATPase in phosphorylating conditions of yeast mitochondria.
J Biochem 128(4):553–559
100 M. Kolisek et al.
Joyal JL, Aprille JR (1992) The ATP-Mg/Pi carrier of rat liver mitochondria catalyzes a divalent
electroneutral exchange. J Biol Chem 267:19198–19203
Jung DW, Apel L, Brierley GP (1990) Matrix free Mg2+ changes with metabolic state in isolated
heart mitochondria. Biochemistry 29(17):4121–4128
Jung DW, Panzeter E, Baysal K, Brierley GP (1997) On the relationship between matrix free Mg2+
concentration and total Mg2+ in heart mitochondria. Biochim Biophys Acta 1320(3):310–320
Karch J, Molkentin JD (2014) Identifying the components of the elusive mitochondrial permeability
transition pore. Proc Natl Acad Sci U S A 111(29):10396–10397
Khan MB, Sponder G, Sjöblom B, Svidová S, Schweyen RJ, Carugo O, Djinović-Carugo K (2013)
Structural and functional characterization of the N-terminal domain of the yeast Mg2+ channel
Mrs2. Acta Crystallogr D Biol Crystallogr 69(Pt 9):1653–1664
Klingenberg M (2008) The ADP and ATP transport in mitochondria and its carrier. Biochim
Biophys Acta 1778(10):1978–2021
Koivusalo M, Steinberg BE, Mason D, Grinstein S (2011) In situ measurement of the electrical
potential across the lysosomal membrane using FRET. Traffic 12(8):972–982
Kolisek M, Zsurka G, Samaj J, Weghuber J, Schweyen RJ, Schweigel M (2003) Mrs2p is an
essential component of the major electrophoretic Mg2+ influx system in mitochondria. EMBO J
22(6):1235–1244
Kolisek M, Launay P, Beck A, Sponder G, Serafini N, Brenkus M, Froschauer EM, Martens H,
Fleig A, Schweigel M (2008) SLC41A1 is a novel mammalian Mg2+ carrier. J Biol Chem 283
(23):16235–16247
Kolisek M, Nestler A, Vormann J, Schweigel-Röntgen M (2012) Human gene SLC41A1 encodes
for the Na+/Mg2+ exchanger. Am J Physiol Cell Physiol 302(1):C318–C326
Kolisek M, Sponder G, Mastrototaro L, Smorodchenko A, Launay P, Vormann J, Schweigel-
Röntgen M (2013a) Substitution p.A350V in Na+/Mg2+ exchanger SLC41A1, potentially
associated with Parkinson’s disease, is a gain-of-function mutation. PLoS One 8(8):e71096
Kolisek M, Galaviz-Hernández C, Vázquez-Alaniz F, Sponder G, Javaid S, Kurth K, Nestler A,
Rodríguez-Moran M, Verlohren S, Guerrero-Romero F, Aschenbach JR, Vormann J (2013b)
SLC41A1 is the only magnesium responsive gene significantly overexpressed in placentas of
preeclamptic women. Hypertens Pregnancy 32(4):378–389
Kovanich D, van der Heyden MA, Aye TT, van Veen TA, Heck AJ, Scholten A (2010) Sphingosine
kinase interacting protein is an A-kinase anchoring protein specific for type I cAMP-dependent
protein kinase. Chembiochem 11(7):963–971
Kubota T, Shindo Y, Tokuno K, Komatsu H, Ogawa H, Kudo S, Kitamura Y, Suzuki K, Oka K
(2005) Mitochondria are intracellular magnesium stores: investigation by simultaneous fluores-
cent imagings in PC12 cells. Biochim Biophys Acta 1744(1):19–28
Kun E (1976) Kinetics of ATP-dependent Mg2+ flux in mitochondria. Biochemistry 15
(11):2328–2336
Kuramoto T, Kuwamura M, Tokuda S, Izawa T, Nakane Y, Kitada K, Akao M, Guénet JL,
Serikawa T (2011) A mutation in the gene encoding mitochondrial Mg2+ channel MRS2 results
in demyelination in the rat. PLoS Genet 7(1):e1001262
Kuwamura M, Inumaki K, Tanaka M, Shirai M, Izawa T, Yamate J, Franklin RJ, Kuramoto T,
Serikawa T (2011) Oligodendroglial pathology in the development of myelin breakdown in the
dmy mutant rat. Brain Res 1389:161–168
Lambie EJ, Tieu PJ, Lebedeva N, Church DL, Conradt B (2013) CATP-6, a C. elegans ortholog of
ATP13A2 PARK9, positively regulates GEM-1, an SLC16A transporter. PLoS One 8(10):
e77202
Lee S, Russo D, Redman CM (2000) The Kell blood group system: Kell and XK membrane
proteins. Semin Hematol 37(2):113–121
Lefkimmiatis K, Leronni D, Hofer AM (2013) The inner and outer compartments of mitochondria
are sites of distinct cAMP/PKA signaling dynamics. J Cell Biol 202(3):453–462
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 101
Noguchi M, Hirata N, Suizu F (2018) AKT keeps the beat in CLOCK’s circadian rhythm. J Biol
Chem 293(23):9137–9138
Nosek MT, Dransfield DT, Aprille JR (1990) Calcium stimulates ATP-Mg/Pi carrier activity in rat
liver mitochondria. J Biol Chem 265(15):8444–8450
Nury H, Dahout-Gonzalez C, Trézéguet V, Lauquin GJ, Brandolin G, Pebay-Peyroula E (2006)
Relations between structure and function of the mitochondrial ADP/ATP carrier. Annu Rev
Biochem 75:713–741
Parry DA, Mighell AJ, El-Sayed W, Shore RC, Jalili IK, Dollfus H, Bloch-Zupan A, Carlos R, Carr
IM, Downey LM, Blain KM, Mansfield DC, Shahrabi M, Heidari M, Aref P, Abbasi M,
Michaelides M, Moore AT, Kirkham J, Inglehearn CF (2009) Mutations in CNNM4 cause
Jalili syndrome, consisting of autosomal-recessive cone-rod dystrophy and amelogenesis
imperfecta. Am J Hum Genet 84(2):266–273
Penner R, Fleig A (2007) The Mg2+ and Mg2+-nucleotide-regulated channel-kinase TRPM7. Handb
Exp Pharmacol 179:313–328
Pfaff E, Heldt HW, Klingenberg M (1969) Adenine nucleotide translocation of mitochondria.
Kinetics of the adenine nucleotide exchange. Eur J Biochem 10(3):484–493
Piskacek M, Zotova L, Zsurka G, Schweyen RJ (2009) Conditional knockdown of hMRS2 results
in loss of mitochondrial Mg2+ uptake and cell death. J Cell Mol Med 13(4):693–700
Qin Y, Dittmer PJ, Park JG, Jansen KB, Palmer AE (2011) Measuring steady-state and dynamic
endoplasmic reticulum and Golgi Zn2+ with genetically encoded sensors. Proc Natl Acad Sci
U S A 108(18):7351–7356
Quamme GA (2010) Molecular identification of ancient and modern mammalian magnesium
transporters. Am J Physiol Cell Physiol 298(3):C407–C429
Rivera A, Kam SY, Ho M, Romero JR, Lee S (2013) Ablation of the Kell/Xk complex alters
erythrocyte divalent cation homeostasis. Blood Cells Mol Dis 50(2):80–85
Rodríguez-Zavala JS, Moreno-Sánchez R (1998) Modulation of oxidative phosphorylation by Mg2+
in rat heart mitochondria. J Biol Chem 273(14):7850–7855
Romani AM (2007) Magnesium homeostasis in mammalian cells. Front Biosci 12:308–331
Romani AMP (2011) Cellular magnesium homeostasis. Arch Biochem Biophys 512(1):1–23
Romani AM, Scarpa A (2000) Regulation of cellular magnesium. Front Biosci 5:D720–D734
Romani A, Dowell E, Scarpa A (1991) Cyclic AMP-induced Mg2+ release from rat liver
hepatocytes, permeabilized hepatocytes, and isolated mitochondria. J Biol Chem 266
(36):24376–24384
Roulis E, Hyland C, Flower R, Gassner C, Jung HH, Frey BM (2018) Molecular basis and clinical
overview of McLeod syndrome compared with other Neuroacanthocytosis syndromes: a
review. JAMA Neurol. https://doi.org/10.1001/jamaneurol.2018.2166
Run C, Yang Q, Liu Z, OuYang B, Chou JJ (2015) Molecular basis of MgATP selectivity of the
mitochondrial SCaMC carrier. Structure 23(8):1394–1403
Rutter GA, Osbaldeston NJ, McCormack JG, Denton RM (1990) Measurement of matrix free Mg2+
concentration in rat heart mitochondria by using entrapped fluorescent probes. Biochem J 271
(3):627–634
Sahni J, Scharenberg AM (2013) The SLC41 family of MgtE-like magnesium transporters. Mol
Asp Med 34(2–3):620–628
Sahni J, Nelson B, Scharenberg AM (2007) SLC41A2 encodes a plasma-membrane Mg2+ trans-
porter. Biochem J 401(2):505–513
Sardanelli AM, Signorile A, Nuzzi R, Rasmo DD, Technikova-Dobrova Z, Drahota Z, Occhiello A,
Pica A, Papa S (2006) Occurrence of A-kinase anchor protein and associated cAMP-dependent
protein kinase in the inner compartment of mammalian mitochondria. FEBS Lett 580
(24):5690–5696
Schapiro FB, Grinstein S (2000) Determinants of the pH of the Golgi complex. J Biol Chem 275
(28):21025–21032
Schindl R, Weghuber J, Romanin C, Schweyen RJ (2007) Mrs2p forms a high conductance Mg2+
selective channel in mitochondria. Biophys J 93(11):3872–3883
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 103
Tewari SG, Dash RK, Beard DA, Bazil JN (2012) A biophysical model of the mitochondrial
ATP-Mg/Pi carrier. Biophys J 103(7):1616–1625
Traba J, Satrústegui J, del Arco A (2009) Characterization of SCaMC-3-like/slc25a41, a novel
calcium-independent mitochondrial ATP-Mg/Pi carrier. Biochem J 418(1):125–133
Trapani V, Wolf FI (2015) Mitochondrial magnesium to the rescue. Magnes Res 28(2):79–84
Tsao YT, Shih YY, Liu YA, Liu YS, Lee OK (2017) Knockdown of SLC41A1 magnesium
transporter promotes mineralization and attenuates magnesium inhibition during osteogenesis
of mesenchymal stromal cells. Stem Cell Res Ther 8(1):39
Vallipuram J, Grenville J, Crawford DA (2010) The E646D-ATP13A4 mutation associated with
autism reveals a defect in calcium regulation. Cell Mol Neurobiol 30(2):233–246
Vergun O, Votyakova TV, Reynolds IJ (2003) Spontaneous changes in mitochondrial membrane
potential in single isolated brain mitochondria. Biophys J 85(5):3358–3366
Voets T, Nilius B, Hoefs S, van der Kemp AW, Droogmans G, Bindels RJ, Hoenderop JG (2004)
TRPM6 forms the Mg2+ influx channel involved in intestinal and renal Mg2+ absorption. J Biol
Chem 279(1):19–25
Wabakken T, Rian E, Kveine M, Aasheim HC (2003) The human solute carrier SLC41A1 belongs
to a novel eukaryotic subfamily with homology to prokaryotic MgtE Mg2+ transporters.
Biochem Biophys Res Commun 306(3):718–724
Wang CY, Shi JD, Yang P, Kumar PG, Li QZ, Run QG, Su YC, Scott HS, Kao KJ, She JX (2003)
Molecular cloning and characterization of a novel gene family of four ancient conserved domain
proteins (ACDP). Gene 306:37–44
Wang CY, Yang P, Shi JD, Purohit S, Guo D, An H, Gu JG, Ling J, Dong Z, She JX (2004)
Molecular cloning and characterization of the mouse Acdp gene family. BMC Genomics 5(1):7
Wiesenberger G, Waldherr M, Schweyen RJ (1992) The nuclear gene MRS2 is essential for the
excision of group II introns from yeast mitochondrial transcripts in vivo. J Biol Chem 267
(10):6963–6969
Will C, Breiderhoff T, Thumfart J, Stuiver M, Kopplin K, Sommer K, Günzel D, Querfeld U, Meij
IC, Shan Q, Bleich M, Willnow TE, Müller D (2010) Targeted deletion of murine Cldn16
identifies extra- and intrarenal compensatory mechanisms of Ca2+ and Mg2+ wasting. Am J
Physiol Renal Physiol 298(5):F1152–F1161
Wuttke MS, Buck J, Levin LR (2001) Bicarbonate-regulated soluble adenylyl cyclase. JOP 2
(4 Suppl):154–158
Xie H, Zhang Y, Zhang P, Wang J, Wu Y, Wu X, Netoff T, Jiang Y (2014) Functional study of
NIPA2 mutations identified from the patients with childhood absence epilepsy. PLoS One 9
(10):e109749
Yamanaka R, Shindo Y, Hotta K, Suzuki K, Oka K (2013) NO/cGMP/PKG signaling pathway
induces magnesium release mediated by mitoKATP channel opening in rat hippocampal
neurons. FEBS Lett 587(16):2643–2648
Yamanaka R, Tabata S, Shindo Y, Hotta K, Suzuki K, Soga T, Oka K (2016) Mitochondrial Mg2+
homeostasis decides cellular energy metabolism and vulnerability to stress. Sci Rep 6:30027
Yamazaki D, Funato Y, Miura J, Sato S, Toyosawa S, Furutani K, Kurachi Y, Omori Y,
Furukawa T, Tsuda T, Kuwabata S, Mizukami S, Kikuchi K, Miki H (2013) Basolateral Mg2+
extrusion via CNNM4 mediates transcellular Mg2+ transport across epithelia: a mouse model.
PLoS Genet 9(12):e1003983
Yanai A, Huang K, Kang R, Singaraja RR, Arstikaitis P, Gan L, Orban PC, Mullard A, Cowan CM,
Raymond LA, Drisdel RC, Green WN, Ravikumar B, Rubinsztein DC, El-Husseini A, Hayden
MR (2006) Palmitoylation of huntingtin by HIP14 is essential for its trafficking and function.
Nat Neurosci 9(6):824–831
Yu N, Jiang J, Yu Y, Li H, Huang X, Ma Y, Zhang L, Zou J, Zhang B, Chen S, Liu P (2014)
SLC41A1 knockdown inhibits angiotensin II-induced cardiac fibrosis by preventing Mg2+
efflux and Ca2+ signaling in cardiac fibroblasts. Arch Biochem Biophys 564:74–82
Zhang GH, Melvin JE (1996) Na+-dependent release of Mg2+ from an intracellular pool in rat
sublingual mucous acini. J Biol Chem 271(46):29067–29072
Magnesium Extravaganza: A Critical Compendium of Current Research into. . . 105
Zhang J, Wang Y, Liu X, Dagda RK, Zhang Y (2017a) How AMPK and PKA interplay to regulate
mitochondrial function and survival in models of ischemia and diabetes. Oxidative Med Cell
Longev 2017:4353510
Zhang H, Kozlov G, Li X, Wu H, Gulerez I, Gehring K (2017b) PRL3 phosphatase active site is
required for binding the putative magnesium transporter CNNM3. Sci Rep 7(1):48
Zhou H, Clapham DE (2009) Mammalian MagT1 and TUSC3 are required for cellular magnesium
uptake and vertebrate embryonic development. Proc Natl Acad Sci U S A 106
(37):15750–15755
Zhu X, Rivera A, Golub MS, Peng J, Sha Q, Wu X, Song X, Kumarathasan P, Ho M, Redman CM,
Lee S (2009) Changes in red cell ion transport, reduced intratumoral neovascularization, and
some mild motor function abnormalities accompany targeted disruption of the Mouse Kell gene
(Kel). Am J Hematol 84(8):492–498
Zippin JH, Chen Y, Nahirney P, Kamenetsky M, Wuttke MS, Fischman DA, Levin LR, Buck J
(2003) Compartmentalization of bicarbonate-sensitive adenylyl cyclase in distinct signaling
microdomains. FASEB J 17(1):82–84
Zsurka G, Gregán J, Schweyen RJ (2001) The human mitochondrial Mrs2 protein functionally
substitutes for its yeast homologue, a candidate magnesium transporter. Genomics 72
(2):158–168
Rev Physiol Biochem Pharmacol (2019) 176: 107–130
DOI: 10.1007/112_2018_11
© Springer International Publishing AG, part of Springer Nature 2018
Published online: 5 May 2018
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
2 Curcumin in the Treatment of Gynecological Cancers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
The original version of this chapter was revised. A correction to this chapter is available at
DOI: 10.1007/112_2018_14.
A. A. Momtazi-Borojeni (*)
Nanotechnology Research Center, Bu-Ali Research Institute, Mashhad University of Medical
Sciences, Mashhad, Iran
Department of Medical Biotechnology, Student Research Committee, Faculty of Medicine,
Mashhad University of Medical Sciences, Mashhad, Iran
e-mail: momtaziaa921@mums.ac.ir; abbasmomtazi@yahoo.com
J. Mosafer
Research Center of Advanced Technologies in Medicine, Torbat Heydarieh University of
Medical Sciences, Torbat Heydarieh, Iran
B. Nikfar (*)
Pars Advanced and Minimally Invasive Medical Manners Research Center, Pars Hospital, Iran
University of Medical Sciences, Tehran, Iran
e-mail: banafsheh.nikfar@gmail.com
M. Ekhlasi-Hundrieser
Werlhof-Institut, Hannover, Germany
S. Chaichian
Minimally Invasive Techniques Research Center in Women, Tehran Medical Sciences Branch,
Islamic Azad University, Tehran, Iran
A. Mehdizadehkashi
Endometriosis Research Center, Iran University of Medical Sciences, Tehran, Iran
A. Vaezi
Department of Community Medicine, School of Medicine, Isfahan University of Medical
Sciences, Isfahan, Iran
108 A. A. Momtazi-Borojeni et al.
Abbreviations
1 Introduction
varying growth factors, and cell membrane receptors (Hajavi et al. 2017; Kasinski
et al. 2008; Pan et al. 2008; Soflaei et al. 2017; Watson et al. 2010; Zhang et al.
2017). Curcumin with very wide pleiotropic functions holds the key to modifying
the trend of cancer therapy and advanced development of current cancer therapy
modalities. Regarding gynecological cancers, curcumin has demonstrated opposing
typical cervical cancer risk factors in advancing molecular alterations toward
cancer incidence or progression, including human papilloma virus infections (typi-
cally HPV16 and 18), estrogen, smoking, and obesity (Maruthur et al. 2009; Zaman
et al. 2016). For instance, it has been shown that curcumin could inhibit the
expression of E6 and E7 oncoprotein, reduce estrogen-induced DNA damage, and
mitigate adipose-related inflammation and estrogen production (Zaman et al. 2016).
The current review presents an assortment of studies on curcumin ameliorating
functions in gynecological cancer progression and metastasis. The sensitizing influ-
ence of curcumin treatment, when combined with routine chemotherapeutic agents
like cisplatin and paclitaxel as well as irradiation, has been addressed, together with
the molecular pathways associated with the drug-induced resistance which curcumin
counteracts. Moreover, the approaches adopted to advance curcumin anticancer
potential, stability, and bioavailability have also been discussed thoroughly in
terms of effectiveness, which includes various curcumin formulations and curcumin
derivatives.
It has been shown that curcumin could act as an anti-metastatic agent and inhibit
endometrial carcinoma (EC) cell migration and invasion in vitro through decreasing
the expression and activity of the matrix metalloproteinases (MMP)-2 and MMP-9.
These enzymes that degrade the extracellular matrix in tumors make the metastasis
of cancer cells possible and are believed to drive deep myometrial cancer invasion
and metastasis in lymph node in type II EC. The reduced expression of these
enzymes by curcumin was also found to occur through suppression of the ERK
signaling pathway (Chen et al. 2015b). Curcumin-induced apoptosis in ovarian
cancer cells was found to be independent of p35, as it displayed the same cytotoxic
activity in cells with reduced or knockdown p53 expression, as shown in the
wild-type p53 cells. Nuclear condensation and fragmentation, DNA fragmentation,
and poly(ADP-ribose) polymerase-1 cleavage were the cell features in HEY cells
treated with curcumin which denoted cell apoptosis. Furthermore, it was found
that both the intrinsic and extrinsic pathways of apoptosis could be activated by
curcumin. The enhanced activity of p38 mitogen-activated protein kinases (MAPK)
reduced the expression of antiapoptotic regulators of survivin and Bcl-2, and the
suppression of prosurvival Akt signaling was also found to be involved in curcumin-
mediated anticancer cell death in various ovarian cancer cells (Watson et al. 2010).
It has been shown that curcumin could partially suppress urokinase-type plas-
minogen activator (uPA) expression in the highly invasive human ovarian cancer
cell line, HRA, which is involved in cancer cell metastasis. The expression of the
Curcumin Effect on Gynecological Cancers 111
Platinum drugs such as cisplatin and oxaliplatin are the first-line chemotherape-
utics against ovarian, bladder, and testicular cancers, and their administrations are
frequently faced with the development of resistant tumors (Montopoli et al. 2009).
The loss of platinum uptake by cells through gated channel-facilitated diffusion, p53
gene implication in DNA damage repair, and enhanced intracellular level of gluta-
thione, responsible for platinum inactivation and removal, are among the underlying
mechanisms for the drug resistance. To evade the platinum drug-induced resistance,
extensive studies have been conducted, in which a combination therapy with phyto-
chemicals has been shown to be highly effective. In this regard, curcumin has been
utilized in combination with platinum drugs like cisplatin and oxaliplatin to enhance
their anticancer properties.
It has been shown that curcumin could resensitize cisplatin-resistant ovarian
cancer cells, and it suppresses DNA damage responses against these DNA cross-
linking agents. It has been found that curcumin treatment downregulates the Fanconi
anemia (FA)/BRCA pathway-related DNA damage repair responses, such as
FANCD2 protein mono-ubiquitination, which is the prerequisite step for the DNA
damage repair complex to form and relocate into chromatin of the DNA lesion sites
(Chen et al. 2015a). Therefore, curcumin could reverse the acquired resistance in
cancer cells, which lies in the enhanced activation of the FA/BRCA pathway, in
response to DNA cross-linking agents in long-term administration (Fig. 1).
Moreover, curcumin could suppress cisplatin resistance development through
extracellular vesicle-mediated cell-cell communication. It is believed that the extra-
cellular vesicles, known as exosome, transfer some proteins, mRNAs, and non-
coding RNAs from donor cells to recipient cells, and this communication leads
to the development of a drug-resistant cell population in various cancers (Zhang
et al. 2017). Curcumin could limit such exosome-mediated chemoresistance by
changing their contents. For instance, curcumin treatment has been shown to be
accompanied with the restoration of MEG3 long noncoding (lnc) RNA levels
(Fig. 1), upregulation of miR-29a and miR-185, and downregulation of miR-124
112 A. A. Momtazi-Borojeni et al.
pathway where the ROS intracellular level is decreased, whereas high curcumin
concentrations (>15 μM) were found to enhance ROS levels in these tissues.
Conversely, in cancer cells with previously high ROS concentrations, curcumin
was shown to enhance ROS levels by inactivating the NF-kB pathway (Sreekanth
et al. 2011; Yunos et al. 2011). Moreover, it appears that the reduced ROS and the
increased GSH basal levels are the main hallmark of the development of resistance to
cisplatin in cisplatin-resistant ovarian cancer cells, which could be linked to the more
active NF-kB pathway in these cells. In summary, it is plausible for curcumin to
contribute much greater to induce ROS generation in cisplatin-treated cancer
cells than in non-treated ones, indicating that curcumin could act as a modifier in
chemotherapy (Yunos et al. 2011) (Fig. 1).
In addition to the NF-kB transcription factor discussed above, it was found
that the expression of many other proteins was altered upon curcumin treatment
(Nessa et al. 2012). A total of 59 proteins were found to be associated with platinum
resistance in ovarian cancer cells, juxtaposing 2D gel electrophoresis from A2780
tumor model with that of resistant tumor cells (Huq et al. 2014). These included
cytoskeletal proteins involved in cell invasion and metastasis, stress-related proteins
and molecular chaperones, proteins involved in detoxification and metabolic pro-
cesses, as well as a set of mRNA processing proteins (for a complete list, refer to
Huq et al. 2014). The inhibition of inflammatory cytokines (e.g., TNF-α, IL-1, IL-6)
and enzymes (e.g., cyclooxygenase or COX-2 and inducible nitric oxide synthase or
iNOS), suppression of angiogenic factors (e.g., vascular endothelial growth factor or
VEGF), and modulation of other signaling proteins [e.g., the upregulation of serine/
threonine-specific protein kinase (AKT)] have also been reported in curcumin-
treated cancer cells (Nessa et al. 2012).
It has also been shown that curcumin-mediated sensitization to cisplatin is
associated with its anti-inflammatory activities in resistant cancer cells. It has
been revealed that IL-6 reduction in curcumin-treated CAOV3 and SKOV3 ovarian
cancer cell lines is accompanied by increased sensitivity to cisplatin, where it is
believed that the overproduction of pro-inflammatory cytokines as such by the tumor
cells drives drug resistance and tumor invasion (Chan et al. 2003). The production
of IL-6 could also induce drug resistance in other cancer cells, including myeloma,
lung, breast, prostate, and colon cancer cells. It has been found that multiple
molecular targets are affected in IL-6-induced platinum resistance in various tumor
cells. Mechanistically, IL-6 could reduce cisplatin accumulation in tumor cells
through induction of multidrug-resistant proteins (MRPs) and P-glycoprotein
(Pgp) in human hepatoma and renal carcinoma cells, induce the expression of
glutathione S-transferase involved in ROS scavenging in breast cancer cells, and
enhance the expression of metal-detoxifying protein of metallothionein in ovarian
cancer cells. It has also been proposed to be involved in enhancing the invasiveness
of ovarian cancer cells, where the induced transcription factor NF-kB results to
the expression of additional inflammatory cytokines (Chan et al. 2003).
The modulation of epigenetic regulators could lead to the emergence of cancer
cells, where curcumin has also been shown to counteract them (Roy and Mukherjee
2014). In cervical cancer cases, the human papilloma virus is putatively known as
114 A. A. Momtazi-Borojeni et al.
the causative agent of cancer emergence and development, in which curcumin has
been effective in suppressing the expression of viral oncoproteins of HPV-E6 and
HPV-E7. Curcumin treatment has also been reported to mitigate cell molecular
modulations derived from the activity of HPV-E6 and HPV-E7 proteins. Curcumin
could inhibit p53 ubiquitin-dependent proteasomal degradation driven by HPV-E6
and act against HPV-E7-reduced pRb functionality. Curcumin could result in cell
cycle arrest at the G1/S phase through the modification of regulatory proteins
involved in cell cycle. It could inhibit the histone deacetylases (HDACs) that are
activated by HPV. The acetylation and upregulation of p53 proteins; increased pRb,
p21, and p27; and the corresponding suppression of cylin D1 and CDK4 have also
been shown in both cisplatin-sensitive and cisplatin-resistant cancer cell line SiHa
upon curcumin treatment, which are known to occur during cancer cell apoptosis
(Roy and Mukherjee 2014).
Figure 2 depicts the summary of the abovementioned multiple pathways and
molecular targets where curcumin exerts opposing influence over cisplatin-induced
resistance. In summary, it appears that curcumin is a modifier of multiple cellular
pathways deregulated during cancer cell progression and, therefore, it could con-
tribute to the advanced antiproliferative responses of other chemoagents applied for
cervical and ovarian cancers. Such an improvement might hinder the development of
resistance toward these agents. Paclitaxel is another chemotherapeutic agent that is
administered for different sorts of gynecological cancers, and it has been shown that
co-treatment with curcumin and paclitaxel could promote antitumor responses in
comparison with paclitaxel alone. In this context, curcumin formulations could alter
the expression of multiple cellular proteins and provide resistance to paclitaxel,
which are discussed in the following two sections.
MEG3 IKK
lat
DNMT3a
thy
PP
me
mi-R29a
A
DNMT1
po
DN
mRNA
hy
PP p21
mi-R185 Cur ReLA p50
PIP2 027
PI–3K Cyclin D1
ne
mi-R124
PTEN
ge
CDK4
Cell survival
proliferation
et
PIP3
mi-R-21
rg
Ac
Ta
mobility
se
Ac
era
RE
Akt P53
NA
pathway
lym
IL-1 TNF-a Ac Ac mRNA
rD
po
a
Ac
cle
A
RN
Nu
TGF-Beta IL-6 HDACs
iNOS
E6
MT Cox-2
D
Dr rug P-gp E7
ug d E2F
ef eto GST Rb E2F
flu xif
x ica transcription
tio E7
n P53 Degradation factor
Rb
Fig. 2 A summary of the molecular pathways by which curcumin treatment could lead to the
sanitization of cancer cells in the cisplatin-induced resistant cancer cells. Curcumin could modulate
the content of exosomes that contain molecular messengers driving the development of resistance
toward cisplatin in the recipient cells (yellow sect). By inactivating the NF-kB pathway, curcumin
could modify many molecular targets that are involved in cancer progression and metastasis (pink
sect). Curcumin could reduce inflammatory cytokine secretion and the enzymes producing inflam-
matory compounds. Through IL-6 downregulation, moreover, curcumin could reduce the expres-
sion of metallothionein, glutathione S-transferase and P-glycoproteins, involved in the scavenging
of superoxide radicals, drug detoxification, and drug efflux from cells, respectively (Green section).
Finally, by counteracting HPV-E6 and HPV-E7 oncoprotein, curcumin could restore the level of
antiapoptotic protein p53, inactivate histone deacetylase involved in chromatin condensation, and
limit E2F transcription factor translocation to the nucleus, where the expression of target genes
drives cell division and growth
the NF-kB activation status and the expression of NF-kB target genes involved
in inflammation and tumor aggressiveness (such as Cox-2, ICAM-1, cyclin D1,
VEGF, MMP-2, and MMP-9); the expression of antiapoptotic proteins that are
transactivated by NF-kB (Bcl-2, c-IA P1, survivin, and XIAP); the expression and
activation of three vital MAP kinases – i.e., c-Jun-NH2 kinase (JNK), extracellular
signal-regulated protein kinase (ERK), and p38; as well as the cleavage and activity
of procaspases 9, 8, 7, and 3. All these molecular evidences indicated that curcumin
could tackle cancerous tumors by modulating various cell signaling pathways and
kinases (Sreekanth et al. 2011), which could promote the efficiency of treatment
in combination with paclitaxel. Similarly, it has been reported that the combination
of curcumin (5 μM) and paclitaxel (5 nM) could augment anticancer responses more
116 A. A. Momtazi-Borojeni et al.
efficiently than paclitaxel alone in HeLa cells, without any synergistic effect on
normal cervical cells, the 293 cell line (Bava et al. 2005). It has been proposed that
the curcumin-induced sensitization to paclitaxel could be related to the opposite
effect of curcumin on the NF-kB activation status. It was identified that curcumin
could suppress NF-kB and Akt pathways, augment the activation of caspases and
cytochrome c release. Moreover, it was discovered that curcumin opposed the
NF-kB activation induced by paclitaxel and reduced the phosphorylation of Akt,
which is a survival signal regulated by NF-kB (Bava et al. 2005). However, at low
concentrations (5 μM), curcumin could not interfere with the tubulin-polymerization
action of paclitaxel and could not further augment the cell cycle protein Cdc2, which
increased during paclitaxel-induced mitotic arrest. This indicates that paclitaxel-
induced resensitization by curcumin is independent of the classic function of taxols
(or paclitaxel).
To exert the abovementioned influences in cancer cells in vivo, it is required to
improve the efficiency of drug delivery to these cells through the application of
various drug delivery systems. Curcumin and paclitaxel have been shown to have
poor pharmacokinetic profiles, which necessitates the use of appropriate formula-
tions to help in attaining the required dose of the drug at tumor sites.
drug and curcumin, the thin phospholipid interfacial layer, and the PEG hydro-
philic outer layer. It has been found out that this system was more effective in
controlling drug release compared to simple PLGA systems and could retain pacli-
taxel up to 72 h in PBS. Moreover, it has also been shown that the PLGA nano-
particles containing curcumin and paclitaxel are more efficacious in decreasing the
expression of P-gp compared to free curcumin (Liu et al. 2016). Polyethylene glycol-
phosphatidylethanolamine (PEG-PE) micelles targeted with transferrin (TF) are
another example of nanoparticles that have been used to promote paclitaxel and
curcumin delivery to tumor sites and enhance the efficacy of tumor therapy. These
micelles were evaluated against resistant ovarian cancers in a cancer cell culture
grown in multicellular three-dimensional spheroids and in vivo tumors. When
paclitaxel was co-delivered with curcumin in the form of micelles, an increase was
recorded in the cytotoxicity of paclitaxel. In addition, transferrin modification of the
micelles could assist in significantly deeper micelle penetration into the spheroids
and tumors (Sarisozen et al. 2014).
These studies all stated that curcumin could significantly enhance the antitumor
potential of paclitaxel against resistant cancer cells, when curcumin is added into the
chemotherapy regimen. Radiotherapy is also applied along with platinum-based
agents in the treatment of advanced ovarian and cervical cancers. As previously
discussed, curcumin could overcome these drug-induced resistances, and it has been
shown to overcome radiotherapy-induced resistance as well.
with more anticancer potency, and basic studies to unravel the molecular pathways
modulated with curcumin treatment in order to find a combination therapy capable
of tackling various malignancies.
Table 1 Rational approaches to overcome curcumin insufficient efficacy in cervical and ovarian
cancer cells
Curcumin derivatives Description Ref.
3,5-bis (2-flurobenzylidene) Increased cytotoxicity, inhibition of Kasinski et al.
piperidin-4-one NF-kB nuclear translocation, TNF-a- (2008)
induced IkB phosphorylation and
degradation, and IKK inactivation
1,5-bis(22-hydroxyl)21,4- In silico study proposed improved Singh and Misra
pentadiene interaction with HPV16-E6 protein (2013)
active site and p53 restoration
1,5-bis(2-hydroxyphenyl)2 Increased cytotoxicity, DNA Wang et al. (2011)
1,4-pentadiene-3-one fragmentation, and decreased HPV16-
and HPV18-associated E6 and E6
oncoproteins
Dimethoxycurcumin Increased cytotoxicity and Wang et al. (2011)
downregulation of cyclin D1
Curcumin conjugation
Curcumin-piperic acid In silico studies assumed increased Mishra et al.
toxicity in cervical cancers (2005b)
Dipiperoyl and diglycinoyl Increased cytotoxicity and ROS Mishra et al.
curcumin generation in histiocytoma cells, but it (2005a)
may be efficient against cervical and
ovarian cancer cells?
Curcumin-chlorogenic acid In silico study proposed increased Singh and Misra
cytotoxicity and HPV15-E6 (2013)
downregulation
Curcumin nanoformulations
Liposomal curcumin Including DDAB, cholesterol, and a Saengkrit et al.
nonionic surfactant like Montanov82 (2014)
increased cytotoxicity and cell pene-
tration of curcumin
Niosomal curcumin Including nonionic surfactants of Singh and Misra
Span80, Tween80, and Poloxamer (2013)
188 enhanced cytotoxicity and
controlled curcumin release
Milk-derived exosome Tumor growth inhibition following Aqil et al. (2017)
oral administration
PLGA nanoparticles conjugated Targeted delivery of curcumin to Punfa et al. (2012)
to anti-Pgp proteins cervical cancer cell line of KB-V1,
expressing highly P-gp
Naïve PLGA nanoparticles Enhanced cell apoptosis, reduced Zaman et al. (2016)
tumor burden, and suppressed
HPV-E6 and HPV-E7 oncoprotein
expression
Fe3O4 nanoparticles coated Enhanced curcumin entrapment in the Kumar et al. (2014)
layer by layer with dextran and particles, increased cell penetration, and Mancarella
polylysine films and enhanced cytotoxicity et al. (2015)
122 A. A. Momtazi-Borojeni et al.
balance was shifted toward ceramide accumulation which pushes cancer cells toward
apoptosis and may be useful to cumulatively enhance antiproliferative response in
combination with curcumin.
In addition to ceramide accumulation, curcumin has been shown to result in the
modulations of other cell signaling molecules. It has been found that the activation of
AMP-activated protein kinase (AMPK) could induce cell death and suppress cell
progression in a variety of cancer cells, and CaOV3 ovarian cancer cell pretreatment
with an AMPK inhibitor attenuates curcumin-induced cell death. Moreover, p38
activation and Akt inhibition are other changes which occur in apoptotic cancer cells
treated with curcumin. Considering all the mentioned cell signaling effectors, every
agent that could contribute to these modulations has been proven to enhance the
anticancer potential of curcumin (Pan et al. 2008), and maybe their topical admin-
istration combined with curcumin as an ointment could exhibit therapeutic response
in gynecological cancers that is worth being investigated.
9 Curcumin Derivatives
Molecular docking studies of curcumin analogs with various functional group sub-
stitutions were conducted on prospective targets like EGFR tyrosine kinases, where
the potential analogs were tested on various cancer cells with the hope of unraveling
the relationship between curcumin structure and its activity (Sharma et al. 2015).
Sometimes, these studies culminate in the discovery of more potent analogs as
compared to curcumin.
As discussed previously, the NF-kB signaling pathway plays a central role
in governing cancer cell progression and metastasis, where curcumin has exhibited
cancer-therapeutic values via NF-kB inactivation. In this regard, Kasinski et al.
(2008) presented a synthetic monoketone analog of curcumin-termed 3,5-bis(2-
flurobenzylidene) piperidin-4-one – with enhanced anticancer activity against a
variety of cancer cells, including ovarian and cervical cancer cells. In comparison
with curcumin, this analog exhibited enhanced cancer cell growth inhibition up to
tenfold in comparison with curcumin. Likewise, the analog rapidly inhibited the
nuclear translocation of NF-kB at a dose tenfold lower than that of curcumin. In
mechanism, NF-kB inhibition was found to result from the strong analog-IKK
interaction which resulted in cancer cell apoptosis.
1,5-bis(22-hydroxyl)21,4-pentadiene, as a curcumin derivative lacking a diketone
site and methoxy functional groups, has been found to exert more antiproliferative
effect than curcumin on different cervical cancer cells (Singh and Misra 2013).
The curcumin analog 1,5-bis(2-hydroxyphenyl)2 1,4-pentadiene-3-one could induce
apoptosis more efficiently than curcumin, and it downregulates the expression of
oncogenes E6 and E7 in HPV16- and HPV18-infected cervical cancer cells,
known as risk factors of cervical cancers (Paulraj et al. 2015). It has been shown
that dimethoxycurcumin is a more stable analog of curcumin in physiological media
and could exert improved anticancer effect on multiple cervical cancer cells
Curcumin Effect on Gynecological Cancers 123
(Teymouri et al. 2018). These are shining examples of curcumin derivatives with
enhanced efficacy, which begin with in silico studies on curcumin analogs with
successful in vitro improved potency. However, the translation of such a potency to a
real clinical setting is yet to be fully fulfilled. The low water stability and in vivo
bioavailability of curcumin are the main setbacks of curcumin therapy. It has been
shown that curcumin conjugation to hydrophilic molecules like amino acid, piperic
acid, and chlorogenic acid could increase the stability of curcumin in physiological
media (Singh and Misra 2013). Curcumin conjugation to piperic acid could enhance
the cell penetrability of curcumin, and its administration with chlorogenic acid might
fully restrict cancer cell proliferation in estrogen-responsive cervical cancer cells,
where curcumin has been found to be partially effective in comparison (Mishra et al.
2005b; Singh and Misra 2013).
Apart from the conjugation of curcumin to small molecules, it has been shown
that curcumin entrapment in various nanoparticulate systems could improve its
efficacy and tissue distribution in cervical and ovarian cancer cells, as discussed in
the following section.
10 Curcumin Nanoformulations
determined to be rapidly removed by the neighboring blood cells at the injection site
when they are intravenously administered. As a result, special consideration should
be given to the route intended for liposomal curcumin administration. If liposomal
curcumin is administered intravenously, where the liposomes are required to travel a
long distance before they reach tumor and accumulate there, negative-to-neutral-
charged liposomes would be probably more successful in reaching the tumor
(Teymouri et al. 2016; Teymouri et al. 2015). However, when the intention is to
enhance curcumin delivery via topical application, for example, as cream or an
ointment in cervical cancers, positively charged liposomal curcumin would promote
curcumin delivery to tissues as well as the therapeutic outcome due to increasing cell
internalization of liposomal curcumin (Debata et al. 2013; Song and Kim 2006).
Another issue that should be taken carefully into account is that the DDAB-
containing liposomes per se have been proven toxic to both cancerous and normal
cells. DDAB together with DOPE at DDAB/DOPE at 40 μM have been found to be
potentially harmful to CasKi cells. It is necessary to investigate whether these
liposomal components will cause unwanted suffering before deciding about their
application in clinics, so as to avoid the possible undesired side effects (Saengkrit
et al. 2014).
The curcumin-niosome system has also been shown to possess high entrapment
efficiency and lead to superior apoptotic rate in ovarian cancer A2780 cells com-
pared with the free curcumin dispersed in dimethyl sulfoxide (Xu et al. 2016). The
niosome system consisted of a nonionic surfactant of Span 80, Tween 80, and
Poloxamer 80 plus additives of cholesterol was examined in terms of entrapment
efficiency and curcumin delivery. It was found that the system is a highly superior
version of liposomal curcumin in terms of curcumin entrapment efficiency. How-
ever, whether niosomal curcumin would be a safer and more successful delivery
system, impart improved pharmacokinetic profiles, and result in higher tumor
accumulation of curcumin than liposomal curcumin is an interesting contrastive
study to be undertaken. Moreover, the surfactant-related hemolysis should be con-
sidered when optimizing these formulations of curcumin (Xu et al. 2016). Given
such serious complications that liposome or noisome ingredients might carry for
clinical application, searching for drug delivery systems that are perfectly safe is
highly desirable.
Milk-derived exosomes loaded with curcumin have been demonstrated to surpass
the low bioavailability of oral curcumin and resulted in three to five times increased
delivery of curcumin to various organs, as compared to free curcumin (Aqil et al.
2017). The exosomal curcumin exhibited increased antiproliferative and anti-
inflammatory activity against multiple cancer cell lines, including breast, lung, and
cervical cancer cells. The underlying mechanism for the promoted efficacy of
curcumin might lie in the fact that exosomes enter via endocytosis and go through
the endosomal pathway, where curcumin activity would be preserved in the desir-
able acidic media of endosomes. Apart from this, exosomes alone have been shown
to possess a moderate intrinsic antiproliferative and anti-inflammatory activity
leading to tumor growth inhibition, which is hardly believed to be achieved by
immune factors, miRNAs, and proteins derived from exosomes. Furthermore, high
Curcumin Effect on Gynecological Cancers 125
11 Conclusion
Acknowledgment The authors would like to say special thanks to Dr. Amir Saberi-Demneh and
Dr. Leila Ghalichi for their guidance and kindness.
Conflict of Interest The authors declare that they have no conflicts of interest about this report.
References
Abdollahi E, Momtazi AA, Johnston TP, Sahebkar A (2018) Therapeutic effects of curcumin in
inflammatory and immune-mediated diseases: a nature-made jack-of-all-trades? J Cell Physiol
233:830–848
Abouzeid AH, Patel NR, Sarisozen C, Torchilin VP (2014) Transferrin-targeted polymeric micelles
co-loaded with curcumin and paclitaxel: efficient killing of paclitaxel-resistant cancer cells.
Pharm Res 31:1938–1945
Aggarwal BB, Harikumar KB (2009) Potential therapeutic effects of curcumin, the anti-
inflammatory agent, against neurodegenerative, cardiovascular, pulmonary, metabolic, autoim-
mune and neoplastic diseases. Int J Biochem Cell Biol 41:40–59
Ak T, Gülçin İ (2008) Antioxidant and radical scavenging properties of curcumin. Chem Biol
Interact 174:27–37
Aqil F, Munagala R, Jeyabalan J, Agrawal AK, Gupta R (2017) Exosomes for the enhanced tissue
bioavailability and efficacy of curcumin. AAPS J 19:1691–1702
Bava SV, Puliappadamba VT, Deepti A, Nair A, Karunagaran D, Anto RJ (2005) Sensitization
of taxol-induced apoptosis by curcumin involves down-regulation of nuclear factor-κB and
the serine/threonine kinase Akt and is independent of tubulin polymerization. J Biol Chem
280:6301–6308
Chan MM, Fong D, Soprano KJ, Holmes WF, Heverling H (2003) Inhibition of growth and
sensitization to cisplatin-mediated killing of ovarian cancer cells by polyphenolic chemopre-
ventive agents. J Cell Physiol 194:63–70
Chen P, Li J, Jiang H-G, Lan T, Chen Y-C (2015a) Curcumin reverses cisplatin resistance
in cisplatin-resistant lung cancer cells by inhibiting FA/BRCA pathway. Tumor Biol
36:3591–3599
Chen Q, Gao Q, Chen K, Wang Y, Chen L, Li X (2015b) Curcumin suppresses migration and
invasion of human endometrial carcinoma cells. Oncol Lett 10:1297–1302
Cheng SCS, Luo D, Xie Y (2001) Taxol induced BCL-2 protein phosphorylation in human
hepatocellular carcinoma QGY-7703 cell line. Cell Biol Int 25:261–265
Debata PR, Castellanos MR, Fata JE, Baggett S, Rajupet S, Szerszen A, Begum S, Mata A,
Murty VV, Opitz LM (2013) A novel curcumin-based vaginal cream Vacurin selectively
eliminates apposed human cervical cancer cells. Gynecol Oncol 129:145–153
Curcumin Effect on Gynecological Cancers 127
Momtazi AA, Shahabipour F, Khatibi S, Johnston TP, Pirro M, Sahebkar A (2016) Curcumin as a
MicroRNA regulator in cancer: a review. Rev Physiol Biochem Pharmacol 171:1–38
Momtazi-Borojeni AA, Haftcheshmeh SM, Esmaeili S-A, Johnston TP, Abdollahi E, Sahebkar A
(2017) Curcumin: a natural modulator of immune cells in systemic lupus erythematosus.
Autoimmun Rev 17:125–135
Montopoli M, Ragazzi E, Froldi G, Caparrotta L (2009) Cell-cycle inhibition and apoptosis induced
by curcumin and cisplatin or oxaliplatin in human ovarian carcinoma cells. Cell Prolif
42:195–206
Nessa MU, Beale P, Chan C, Yu JQ, Huq F (2012) Studies on combination of platinum drugs
cisplatin and oxaliplatin with phytochemicals anethole and curcumin in ovarian tumour models.
Anticancer Res 32:4843–4850
Pan W, Yang H, Cao C, Song X, Wallin B, Kivlin R, Lu S, Hu G, Di W, Wan Y (2008) AMPK
mediates curcumin-induced cell death in CaOV3 ovarian cancer cells. Oncol Rep 20:1553–1559
Panahi Y, Ahmadi Y, Teymouri M, Johnston TP, Sahebkar A (2018) Curcumin as a potential
candidate for treating hyperlipidemia: a review of cellular and metabolic mechanisms. J Cell
Physiol 233:141–152
Paulraj F, Abas F, Lajis NH, Othman I, Hassan SS, Naidu R (2015) The curcumin analogue 1, 5-bis
(2-hydroxyphenyl)-1, 4-pentadiene-3-one induces apoptosis and downregulates E6 and E7
oncogene expression in HPV16 and HPV18-infected cervical cancer cells. Molecules
20:11830–11860
Peng S, Xu Q, Ling XB, Peng X, Du W, Chen L (2003) Molecular classification of cancer types
from microarray data using the combination of genetic algorithms and support vector machines.
FEBS Lett 555:358–362
Punfa W, Yodkeeree S, Pitchakarn P, Ampasavate C, Limtrakul P (2012) Enhancement of cellular
uptake and cytotoxicity of curcumin-loaded PLGA nanoparticles by conjugation with anti-P-
glycoprotein in drug resistance cancer cells. Acta Pharmacol Sin 33:823–831
Rezaee R, Momtazi AA, Monemi A, Sahebkar A (2016) Curcumin: a potentially powerful tool to
reverse cisplatin-induced toxicity. Pharmacol Res 117:218–227
Roy M, Mukherjee S (2014) Reversal of resistance towards cisplatin by curcumin in cervical cancer
cells. Asian Pac J Cancer Prev 15:1403–1410
Saengkrit N, Saesoo S, Srinuanchai W, Phunpee S, Ruktanonchai UR (2014) Influence of
curcumin-loaded cationic liposome on anticancer activity for cervical cancer therapy. Colloids
Surf B Biointerfaces 114:349–356
Sarisozen C, Abouzeid AH, Torchilin VP (2014) The effect of co-delivery of paclitaxel and
curcumin by transferrin-targeted PEG-PE-based mixed micelles on resistant ovarian cancer in
3-D spheroids and in vivo tumors. Eur J Pharm Biopharm 88:539–550
Sharma R, Jadav SS, Yasmin S, Bhatia S, Khalilullah H, Ahsan MJ (2015) Simple, efficient, and
improved synthesis of Biginelli-type compounds of curcumin as anticancer agents. Med Chem
Res 24:636–644
Singh AK, Misra K (2013) Human papilloma virus 16 E6 protein as a target for curcuminoids,
curcumin conjugates and congeners for chemoprevention of oral and cervical cancers.
Interdiscip Sci Comput Life Sci 5:112
Soflaei S, Momtazi A, Majeed M, Derosa G, Maffioli P, Sahebkar A (2017) Curcumin: a natural
pan-HDAC inhibitor in cancer. Curr Pharm Des 24:123–129
Song Y-K, Kim C-K (2006) Topical delivery of low-molecular-weight heparin with surface-
charged flexible liposomes. Biomaterials 27:271–280
Sreekanth C, Bava S, Sreekumar E, Anto R (2011) Molecular evidences for the chemosensitizing
efficacy of liposomal curcumin in paclitaxel chemotherapy in mouse models of cervical cancer.
Oncogene 30:3139–3152
Tanaka Y, Kobayashi H, Suzuki M, Kanayama N, Terao T (2004) Transforming growth factor-β1-
dependent urokinase up-regulation and promotion of invasion are involved in Src-MAPK-
dependent signaling in human ovarian cancer cells. J Biol Chem 279:8567–8576
Curcumin Effect on Gynecological Cancers 129
Correction to:
Chapter “Curcumin in Advancing Treatment
for Gynecological Cancers with Developed Drug-
and Radiotherapy-Associated Resistance” in:
A. A. Momtazi-Borojeni et al., Rev Physiol Biochem
Pharmacol,
DOI: 10.1007/112_2018_11
The affiliation of the 6th author Dr. Abolfazl Mehdizadehkashi was incorrect. It has
been corrected to Endometriosis Research Center, Iran University of Medical Sci-
ences, Tehran, Iran.