ArticlePDF AvailableLiterature Review

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

March 2021
Oncogene 40(12)

March 2021
40(12)

DOI:10.1038/s41388-021-01695-8

Authors:

Li Na Zhao

Agency for Science, Technology and Research (A*STAR)

Mikael Bjorklund

Zhejiang University

Matias Caldez

Osaka University

Show all 5 authorsHide

Most of the drugs currently prescribed for cancer treatment are riddled with substantial side effects. In order to develop more effective and specific strategies to treat cancer, it is of importance to understand the biology of drug targets, particularly the newly emerging ones. A comprehensive evaluation of these targets will benefit drug development with increased likelihood for success in clinical trials. The folate-mediated one-carbon (1C) metabolism pathway has drawn renewed attention as it is often hyperactivated in cancer and inhibition of this pathway displays promise in developing anticancer treatment with fewer side effects. Here, we systematically review individual enzymes in the 1C pathway and their compartmentalization to mitochondria and cytosol. Based on these insight, we conclude that (1) except the known 1C targets (DHFR, GART, and TYMS), MTHFD2 emerges as good drug target, especially for treating hematopoietic cancers such as CLL, AML, and T-cell lymphoma; (2) SHMT2 and MTHFD1L are potential drug targets; and (3) MTHFD2L and ALDH1L2 should not be considered as drug targets. We highlight MTHFD2 as an excellent therapeutic target and SHMT2 as a complementary target based on structural/biochemical considerations and up-to-date inhibitor development, which underscores the perspectives of their therapeutic potential.

Our computational model of MTHFD2·NAD + ·P i ·Mg 2+ ·P i ·THF system. Note that the Mg 2+ and inorganic phosphate are shown as spheres, THF and NAD + are shown as sticks and colored white and yellow for C atoms, respectively.

…

Figures - uploaded by Philipp Kaldis

Content may be subject to copyright.

Content uploaded by Philipp Kaldis

Content may be subject to copyright.

Oncogene

https://doi.org/10.1038/s41388-021-01695-8

REVIEW ARTICLE

Therapeutic targeting of the mitochondrial one-carbon pathway:

perspectives, pitfalls, and potential

Li Na Zhao 1●Mikael Björklund2,3,4 ●Matias J. Caldez5●Jie Zheng6●Philipp Kaldis 1

Received: 23 October 2020 / Revised: 27 January 2021 / Accepted: 2 February 2021

Abstract

Most of the drugs currently prescribed for cancer treatment are riddled with substantial side effects. In order to develop more

effective and speciﬁc strategies to treat cancer, it is of importance to understand the biology of drug targets, particularly the

newly emerging ones. A comprehensive evaluation of these targets will beneﬁt drug development with increased likelihood

for success in clinical trials. The folate-mediated one-carbon (1C) metabolism pathway has drawn renewed attention as it is

often hyperactivated in cancer and inhibition of this pathway displays promise in developing anticancer treatment with fewer

side effects. Here, we systematically review individual enzymes in the 1C pathway and their compartmentalization to

mitochondria and cytosol. Based on these insight, we conclude that (1) except the known 1C targets (DHFR, GART, and

TYMS), MTHFD2 emerges as good drug target, especially for treating hematopoietic cancers such as CLL, AML, and T-cell

lymphoma; (2) SHMT2 and MTHFD1L are potential drug targets; and (3) MTHFD2L and ALDH1L2 should not be

considered as drug targets. We highlight MTHFD2 as an excellent therapeutic target and SHMT2 as a complementary target

based on structural/biochemical considerations and up-to-date inhibitor development, which underscores the perspectives

of their therapeutic potential.

Introduction

Historically, drugs to treat cancer patients relied on the fact

that cancer cells proliferate faster than most healthy cells in

our body. This strategy is still valid and is being pursued

around the world by academic and pharmaceutical labora-

tories. In the last decade, our interest in metabolism has

been revived by the ﬁnding that cancer cells remodel their

metabolism and that they can choose from more metabolic

pathways than healthy cells [1,2]. This has led to the

hypothesis that cancer cells can exploit many metabolic

pathways to fulﬁll their metabolic needs, which is generally

not the case for healthy cells [3,4]. The latest reviews on

metabolic ﬂexibility exhibited by cancer cells in their pro-

gression, notably growth, invasion, and drug evasion, dis-

cussed this in details [3,4], which highlights diverse factors

responsible for remodeling of cancer metabolism. Although

this creates complications, this may provide an Achilles

heel for cancer cells that could be exploited therapeutically.

Here we are considering the 1C pathway as a major avenue

to target cancer therapeutically. Nevertheless, targeting

metabolic pathways is not necessarily straightforward since

these pathways are also being used by healthy cells.

Therefore, we have to focus on metabolic pathways that are

either redundant or not usually used in healthy cells but

essential in cancer cells. Many different metabolic path-

ways, such as serine and glycine metabolism [5,6], gluta-

mine metabolism [3–7], and altered glycolysis [8–10], are

*Li Na Zhao

lina.zhao@med.lu.se

*Philipp Kaldis

philipp.kaldis@med.lu.se

1Department of Clinical Sciences, Lund University,

Malmö, Sweden

2Zhejiang University–University of Edinburgh (ZJU-UoE)

Institute, Haining, Zhejiang, PR China

32nd Afﬁliated Hospital, Zhejiang University School of Medicine,

Hangzhou, Zhejiang, PR China

4Deanery of Biomedical Sciences, College of Medicine and

Veterinary Medicine, University of Edinburgh, Edinburgh, UK

5Laboratory of Host Defense, The World Premier International

Research Center Initiative (WPI) Immunology Frontier Research

Center (IFReC), Osaka University, Osaka, Japan

6School of Information Science and Technology, Shanghai Tech

University, Shanghai, PR China

1234567890();,:

being pursued therapeutically. In this review, we are

focusing on the one-carbon (1C) pathway, which starts with

folate and provides 1C units (methyl groups) for DNA

methylation as well as for a variety of anabolic pathways

including DNA, RNA, and amino acid synthesis (Fig. 1).

One feature of this pathway is that it is compartmentalized

to both mitochondria and cytosol.

Due to its essential roles in nucleic acid biosynthesis and

in proliferation, folate has become the basis for a class of

drugs (known as antifolates) that antagonise its action.

Antifolate, as an antimetabolite, has been used in the clinic

since 1948 [11]. Antifolate has been further used for appli-

cations like antibiotics and chemotherapeutics [12–14].

However, these antifolate drugs are generally toxic to normal

cells due to: (a) the targets being present in both normal and

cancer cells; and (b) low speciﬁcity as they target multiple

enzymes and block both DNA and RNA synthesis. The side

effects include hair loss, bone marrow toxicity, and cardiac

anomalies, which can be life-threatening [15–17]. Given the

lack of safe therapies for many cancer patients, and the

importance of folate in biosynthesis, we aim to review the

folate-mediated 1C pathway in depth to provide new per-

spectives for cancer treatment with reduced side effects.

Folate, also known as folic acid, is a generic term for

water-soluble vitamin B. It can be reduced to dihydrofolate

(DHF) and then to tetrahydrofolate (THF) by dihydrofolate

reductase (DHRF), which sequentially transfer hydrogens

from NADPH to folic acid in the positions 5–8 (Fig. 2). The

chemical structure of folate consists of a 2-amino-4-

hydroxy-pteridine ring, para-aminobenzoic acid (pABA),

and polyglutamate. The 1C units are covalently linked to

two nitrogen atoms from the pteridine ring and the pABA

moiety at the 5 and 10 positions (Fig. 2). The pteridine ring

can be reduced or oxidized, and different oxidation states of

THF enable it to carry activated 1C units to be readily

interconverted within the cell [18].

The folate-mediated 1C metabolism is ubiquitous in

eukaryotic cells and organs. Unlike bacteria, yeast, and plants,

animals are not able to synthesize folate and have to take it up

from food. Insufﬁcient folate will lead to complications,

notably neural tube defects [19]andfolate-deﬁciency anemia

[20]. Folate is an essential cofactor in the mammalian 1C

metabolic network (Fig. 1), which comprises two inter-

dependent biosynthetic pathways that are compartmentalized

to the cytoplasm and mitochondria. The 1C network supports

multiple vital physiological processes including (i) biosynth-

esis (purine and thymidine) and homocysteine methylation in

the cytoplasm, (ii) protein synthesis (formylmethionyl-trans-

fer RNA; tRNA) and amino acid homeostasis (serine, glycine,

Fig. 1 Compartmentalization of

the mammalian one-carbon

metabolism to mitochondria and

cytosol. THF: tetrahydrofolate;

DHFR: dihydrofolate reductase;

SAM: S-adenosylmethionine;

TYMS: thymidylate synthetase;

MFT: mitochondrial folate

transporter.

Fig. 2 Chemical structure of folate. Folate consists of three distinct

folate moieties: pterin (red; 2-amino-4-hydroxy-pteridine heterocyclic

ring), pABA (orange; p-aminobenzoyl group), and ployglutamate

(black).

L. N. Zhao et al.

and methionine), as well as (iii) catabolism (choline, purines,

and histidine) in mitochondria [21,22]. The fundamental

nature of 1C metabolism compartmentalization is an evolu-

tionary strategy to enable cellular function properly by opti-

mizing spatial organization of metabolic feats [23] and each

compartment harbors enzyme complexes with speciﬁc activ-

ities to support cellular processes. Understanding the speciﬁc

activities of these enzymes in each compartment, highlights

new opportunities for selective intervention of certain

enzymes activities.

1C units (essentially methyl groups) use THF as a carrier.

THF accepts the 1C units derived from histidine, serine, and

formate in the cytosol. Namely, the 5-formimino-THF gen-

erated by histidine degradation is converted to 5,10-methy-

lene-THF by the formimidoyltransferase cyclodeaminase.

Serines usually come from de novo serine synthesis [24]or

direct through dietary intake. SHMT1/2 transfers the Cβfrom

serine to THF. Furthermore, serine, but not glycine, supplies

the 1C pool for cancer-cell proliferation [25].

The typical chemical reactions involving 1C units are

CH2-THF dehydrogenase (D), methenyl-THF (CH+-THF)

cyclohydrolase (C), and CHO-THF synthetase (S). Usually,

the trifunctional methylenetetrahydrofolate dehydrogenase,

cyclohydrolase and formyltetrahydrofolate synthetase 1

(MTHFD1) will carry out the D,C, and Sactivities in the

cytoplasm. The methylenetetrahydrofolate dehydrogenase

2-like protein (MTHFD2L) will carry out the Dand C

activities in mitochondria in normal cells and is expressed at

low levels in embryonic development [26]. During

embryonic development, MTHFD2 carries out the Dand C

activities in mitochondria. It is expressed in both embryonic

and tumor cells but at low levels in normal cells.

MTHFD1L will carry out the Sactivities in mitochondria.

Facilitated by rapid transportation of serine and exchange of

glycine, mitochondria become the primary source of for-

mate as 1C units for cytoplasmic metabolism [18,27].

In normal cells, the cytosol operates at near-neutral pH

(7.0–7.4) in most tissues and the pH is tightly regulated by

ion transport proteins [28,29], while in the mitochondrial

matrix a more alkaline pH in the range from 7.7 to 7.9 has

been observed. Cancer cells are characterized by altered

metabolism and inﬂammation, leading to an increased

intracellular pH (7.3–7.6) and a decreased extracellular pH

(6.8–7.0 versus ~7.4 for normal cells) [30,31]. The altered

pH promotes a favorable microenvironment for cancer-cell

growth, proliferation, and invasion [32]. Since the cytosolic

and mitochondrial 1C pathways are parallel and the key

enzymes share high sequence identity (an adaptation of the

isoforms to the respective cellular environments) with

nearly the same catalytic properties, this signiﬁcantly

complicates drug development in terms of speciﬁcity, but

also provides opportunities as there are ﬁne differences

between the enzymes in the cytosol and the mitochondria.

The mitochondrial 1C pathway is consistently upregu-

lated in cancer cells [33]. The hyperactivated 1C metabo-

lism has been shown to play a signiﬁcant role in cancer-cell

proliferation and malignancy [34]. Speciﬁcally, MTHFD2

was found to be the most commonly upregulated both at

protein and mRNA levels in various types of cancers

[8,35]; MTHFD1L is overexpressed in colorectal cancer

[36], bladder cancer [37], and hepatocellular carcinoma

[38]; SHMT2 overexpression is observed in various cancers

and the knockdown of SHMT2 yields promising results in

suppressing cancer-cell proliferation and invasion [39]. The

hyperactivity of the 1C pathway partially reﬂects the ana-

bolic metabolism facilitated by 1C units generated from

mitochondria and transported to the cytosol, which sustain

tumorigenesis [33]. The 1C units deprivation resulting from

serine restriction or de novo serine synthesis inhibition can

be an effective strategy against tumorigenesis [5].

Key enzymes in the 1C pathway

In order to review the potential of the 1C pathway in cancer

therapy, we need to introduce the enzymes that make up the

pathway. The focus is on the enzymatic activities of these

proteins and their structural arrangement to highlight the

potential success of chemical inhibitors.

There are ﬁve key enzymes [SHMT2 (EC:2.1.2.1),

MTHFD2 (EC:1.5.1.15), MTHFD2L (EC:1.5.1.15), MTHF

D1L (EC:6.3.4.3), and ALDH1L2 (EC:1.5.1.6)] involved in

the mitochondrial 1C pathway (see Fig. 1), and with the

exception of MTHFD2L, all are upregulated in cancer

[34,38,40,41]. To be considered as potential targets for

selective therapeutic intervention, it is important to under-

stand the fundamental characteristics of these genes and the

encoded proteins. It is also important to compare the detailed

catalytic properties of these enzymes with their closely

related cytosolic counterparts. For example, cytosolic

SHMT1 is an isoenzyme of the mitochondrial SHMT2, and

the cytosolic ALDH1L1 is related to the mitochondrial

ALDH1L2. The detailed chemical reactions are depicted in

Fig. 3. Another issue that needs consideration is the use of

cofactors, which can determine the function, activity, and

speciﬁcity of the enzymes. MTHFD1 uses NADP+as

cofactor, while MTHFD2 and MTHFD2L can use either

NAD+or NADP+but with low efﬁciency when using

NADP+. The origin of the cofactor preference of NAD+or

NADP+is suggested to promote “a more thermo-

dynamically favorable pathway to balance the pools of 10-

formyl-THF during development”[42]. Sequence and

structure comparison of MTHFD1 and MTHFD2 indicates

that Ser197 is a key factor in binding 2′-phosphate of

NADP+while Arg233 is important to bind the phosphate.

The Ser to Arg233 switch from MTHFD1 to MTHFD2

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

could enable the cofactor preference from NADP+to

NAD+. The mitochondrial 1C metabolism, both for produ-

cing and then transferring 1C units to the cytosol, and for

generating glycine and NADPH, is usually studied as a

whole since all chemical processes are folate-mediated.

Folates, in general, share one common structure comprising

three moieties (see Fig. 2): the pteridine ring, which can be

reduced or oxidized; the pABA linker together with the

pteridine ring coordinates the 1C units; and the poly-

glutamate tail, which enhances substrate binding and enables

its retention within mitochondria. Below we follow the ﬂow

of folate through the 1C pathway to review SHMT2,

MTHFD2, MTHFD2L, MTHFD1L, and ALDH1L2.

SHMT2: serine hydroxymethyltransferase 2

The Shmt2 gene contains 11 exons and 10 introns [43].

Alternative splicing results in three conﬁrmed transcript var-

iants. They are the mitochondrial Shmt2, a cytosolic Shmt2

(Shmt2α), which lacks exon 1 that encodes the mitochondrial

targeting peptide, and little is known about the third isoform

which is missing residues 199–208 (199PKTGLIDYNQ208).

SHMT2 catalyzes the ﬁrst pivotal reaction in the mito-

chondrial 1C metabolism pathway, by reversibly interconvert-

ingserinetoglycineandbytransferringthehydroxymethyl

group (Cβof serine) to THF thereby generating 5,10-methy-

lenetetrahydrofolate (5,10-CH2-THF) [44]. It functions as a

Fig. 3 Reaction scheme of key

enzymes in the 1C pathway.

aDHFR catalyzes folic acid

reduction to DHF; bDHF is then

futher processed to THF by

DHFR; cSHMT converts serine

into glycine; dTYMS converts

dUMP to dTMP; and eMTHFD2

carries out the

methylenetetrahydrofolate

dehydrogease and cyclohydrolase

activities.

L. N. Zhao et al.

tetramer in a pyridoxal-5′-phosphate (PLP)-dependent manner,

and PLP, the active form of vitamin B6, is responsible for the

oligomeric state of human SHMT2 and SHMT1 in mito-

chondria and cytosol, respectively [45]. However, SHMT1 can

exist in solution as a tetramer while SHMT2 undergoes a

dimer-to-tetramer transition in a PLP-dependent manner [45].

PLP binds covalently to Lys-280, the succinylation at Lys-280

prevents PLP binding, and desuccinylation by SIRT5 restores

the activity and promotes cell proliferation [46].

SHMT2 generates the major source of activated 1C units

for the de novo purine and pyrimidine synthesis in mito-

chondria. It is important for the translation of the mito-

chondrial respiratory complex by providing methyl donors

to produce the taurinomethyluridine base at the wobble

position of selected mitochondrial tRNAs to maintain the

tRNA pool [47,48]. The loss of SHMT2 leads to impaired

expression of respiratory chain enzymes [47].

The cytosolic isoforms (both SHMT1 and SHMT2α)can

translocate to the nucleus in a small ubiquitin-like modiﬁer

(SUMO)-dependent manner. They assemble the metabolic

complex at sites of DNA replication to sustain de novo thy-

midylate synthesis and to promote cell proliferation during S

and G2/M phases [49,50]. Particularly, Ubc13-mediated

ubiquitination is required for SHMT1 nuclear export and its

increased stability within the nucleus, whereas Ubc9-

mediated modiﬁcation with SUMO2/3 will result in its

nuclear degradation [51]. Both ubiquitin and SUMO mod-

iﬁcation compete for the same lysine residue and thus

determine the localization of SHMT1 in the nucleus [51].

SHMT2αis functionally redundant with SHMT1 in the

nuclear de novo thymidylate synthesis [49].

SHMT1 levels are cell-cycle-regulated since it undergoes

SUMO modiﬁcation (Ub-SHMT1), which leads to its

degradation and to nuclear translocation in Sphase [51].

SHMT2 is expressed at a constant level throughout the cell

cycle to maintain intracellular glycine levels [51]. However,

elevated expression of SHMT2 is observed during tumor cell

proliferation and is reﬂected by increased glycine accumu-

lation, which is also associated with poor prognosis and high

mortality in various cancers [41,52,53]. The disruption of

the mouse Shmt2 gene confers fetal liver-speciﬁc down-

regulation of the 1C pathway resulting in mitochondrial

respiration defects and growth retardation [54]. Since

SHMT2 generates the mitochondrial 10-formyl THF, which

is a cofactor for mitochondrial methionyl-tRNA for-

myltransferase to generate formylmethionyl-tRNAs (fMet-

tRNA), it is important to maintain the fMet-tRNA pool for

the translation of mitochondrially encoded proteins [54,55].

Because SHMT2 is important for embryonic development

and expression of mitochondrial-encoded genes [54,55],

targeting SHMT2 in normal cells is not desirable. However,

overexpression of SHMT2 in cancer cells promotes cell

proliferation [33]. For example, overexpression of SHMT2 is

associated with poor prognosis in intrahepatic cholangio-

carcinoma, and poor survival in patients with kidney cancer

[56,57]. Furthermore, downregulation of SHMT2 suppresses

tumorigenesis in human hepatocellular carcinoma [58].

Therapeutic targeting would require reduction of the SHMT2

activity by a weak inhibitor. There is no potent or speciﬁc

SHMT2 inhibitor available and the published inhibitors are

general inhibitors toward serine hydroxymethyltransferases,

which have not been tested in clinical trials so far [59]. These

general inhibitors include: (i) triazine antifolate (NSC127755),

which irreversibly inhibits SHMT2 and is toxic [60]; (ii)

Leucovorin (5‐formyl‐THF) inhibits both SHMT1 and

SHMT2 at low micromolar concentration and is easily con-

verted into other folic-acid derivatives, which hampers its

clinical use [61]; and (iii) Pyrazolopyran derivatives, which

preferentially inhibit SHMT1 [62].

MTHFD2: NAD+-dependent methylenetetrahydrofolate

dehydrogenase cyclohydrolase

The Mthfd2 gene contains 8 exons and 7 introns. Alter-

native splicing results in two isomers and four potential

isoforms. The identiﬁed two isomers are the mitochondrial

Mthfd2 and the cytosolic Mthfd2, which lacks exon 1

encoding the targeting peptide for its efﬁcient translocation

into mitochondria.

MTHFD2 has emerged as a novel therapeutic target since it

is only expressed during embryonic development [40]andis

almost undetectable in adults. Importantly, MTHFD2 is

overexpressed in cancer cells, notably in acute myeloid leu-

kemia (AML) [63], pancreatic cancer [34], adenocarcinomas

of the colorectum and lung [64], colorectal cancer [65,66],

glioma [67], and renal cell carcinoma [68]. Overexpression is

associated with poor overall survival of patients suffering

from renal cell carcinoma [40,69]. The overexpression of

MTHFD2 supplies cancer cells with building blocks for pyr-

imidine and purine biosynthesis during rapid proliferation and

is broadly required for the growth of cancer cells [40]. Tar-

geting MTHFD2 will unlikely cause major side effects due to

its low expression in most adult tissue [35,66,68–72]. Inhi-

bition of MTHFD2 should lead to NADPH and redox

homeostasis imbalances, and thus should suppress tumori-

genic proliferation and growth, as well as enhance cancer cell

death under hypoxia [66]. In addition, MTHFD2 is not folate

sensitive [73], if the drug targeting MTHFD2 is a folate

antagonist, then the potential side effects can be further miti-

gated by taking a folic-acid supplement [22]. The knockdown

of MTHFD2 causes a decrease in the expression of the cell-

cycle genes such as Ccna2,Mcm7,andSkp2, indicating that

this may interfere with cell-cycle progression [74].

The structure of MTHFD2 has been co-crystallized in

complex with NAD+, inorganic phosphate (Pi), and the inhi-

bitor LY345899, which provides a reliable template (5TC4.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

pdb) for starting structure-based drug discovery [75]. Subse-

quently, MTHFD2 in complex with the inhibitor DS44960156

(6JIB.pdb), Compound 1 (6JID.pdb), and Compound 18

(6KG2.pdb) as well as cofactors (NAD+and Pi) were solved at

2.25 Å, and these structures bear similarity with the ﬁrst

identiﬁed structure. In all four structures, the absolutely

required Mg2+is missing, and the last three structures are

deposited in the Protein Data Bank as dimers. Since MTHFD2

functions as dimer, we have reconstituted the active site of the

MTHFD2 in a homodimer complex with inorganic phosphate,

NAD+, and substrate using homology modeling, and per-

formed a preliminary QM/MM study (Fig. 4; unpublished

data). Our results indicate that without Mg2+, the inorganic

phosphate is drifting away, suggesting that Mg2+plays an

important role in stabilizing the complex. Monomeric human

bifunctional MTHFD2 consists of D and C domains respon-

sible for the Dand Cactivities (see Fig. 1). The D/C domain of

MTHFD2 is composed of two α/βstrands that assemble to

form a cleft. The walls of the cleft are lined with highly con-

served residues. NAD+and Piare bound along one wall, while

the substrate is bound in a deep and rather hydrophobic cleft at

the interface between the two domains. MTHFD2 functions as

a homodimer, and homodimerization occurs by antiparallel

interaction of two NAD+-binding domains. However, the

complete MTHFD2·NAD+·Mg2+·Pi·THF complex is hard to

crystalize, especially in the active state at which the reaction

takes place. This may be because the complex is unstable or

otherwise short-lived. Therefore, we have reconstructed the

MTHFD2·NAD+·Pi·THF complex using homology modeling,

and our simulation indicates that Mg2+plays an important role

in stabilizing the inorganic phosphate. More importantly, pos-

sibly two phosphates may be involved in the binding of Mg2+

(see Fig. 4) or two magnesium ions, which are further coor-

dinated by water molecules. Currently there is no experimental

conﬁrmation for our predictions of the active site. Unlike

magesium ions which are smaller with a low electron density,

two phosphate ions may be easily detected by current experi-

mental technology. Hence, we propose the presence of two

magnesium ions that coordinate the phosphate. Our previous

work in the simulation of the system (see Fig. 5)withonlyone

phosphate group indicates that the Piis drifting away from the

NAD+, which further supports the necessity of having two

magnesium ions. Another important observation from our

simulation indicates that the ﬂexible loop from residue 281 and

285 is important in governing the ligand recruitment and

release. Currently we are progressing to validate our predictions

by quantum mechanics studies as the accuracy of the structures

is key to harness the therapeutic potential of MTHFD2.

MTHFD2 inhibitors have gained attention as an attrac-

tive option as cancer drugs [35,52,64]. The latest progress

includes: (i) carolacton: a natural product, which inhibits

both MTHFD1 and MTHFD2 at the low nM range, and

Fig. 4 Our computational model of

MTHFD2·NAD+·Pi·Mg2+·Pi·THF

system. Note that the Mg2+and

inorganic phosphate are shown as

spheres, THF and NAD+are shown

as sticks and colored white and

yellow for C atoms, respectively.

Fig. 5 Our computational

model of MTHFD2·NAD+

·P ·THF system without Mg+.

Note that the inorganic

phosphate is shown as sphere,

THF and NAD+are shown as

sticks and colored white and

blue for C atoms, respectively.

L. N. Zhao et al.

decreases the growth of cancer cell lines at mM con-

centrations. However, carolacton lacks speciﬁcity toward

MTHFD2 [76]; (2) LY345899: a folate analogue, is capable

of suppressing tumor growth in mouse xenograft models of

colorectal cancer but it inhibits both MTHFD2 (IC50:

663 nM) and MTHFD1 (IC50: 96 nM) at the same time, and

how much inhibition of the tumor growth is due to the

contribution of MTHFD2 inhibition is unclear. Therefore

LY345899 lacks speciﬁcity [66]; (3) DS44960156: a tri-

cyclic coumarin scaffold [77], shows selectivity for

MTHFD2 over MTHFD1 but with low potency (IC50:

1.6 μM); and (4) DS18561882, optimized based on

DS44960156, with >250-fold improved potency and

90-fold improved selectivity for MTHFD2 (6.3 nM) over

MTHFD1 (0.57 μΜ)[78]. DS18561882 seems a decent

drug candidate, however, considering that the preclinical

evaluation procedure and the experimental data revealed

limitations and deﬁciencies as well as a need for further

clinical studies. Overall, more efforts are still needed for the

development of potent and speciﬁc MTHFD2 inhibitors.

The NAD+-dependent activities of MTHFD2 follows an

ordered kinetic mechanism, in which NAD+is recruited

ﬁrst followed by THF, and during the product release, 2,4-

CH+-THF is released before NADH [79]. The kcat value for

MTHFD2 in the presence of NAD+for CH2-H4PteGlu1and

CH2-H4PteGlu5is 12.4 s−1and 15.4 s−1, respectively, while

in the presence of NADP+,thekcat value is 1.5 s−1and 6.4 s−1

[80]. Based on this data, MTHFD2 displays a preference to

use NAD+as cofactor.

MTHFD2 has been described mainly as a mitochondrial

folate metabolic enzyme. However, recent studies suggest

that MTHFD2 has non-enzymatic (or uncharacterized)

functions in the nucleus where MTHFD2 may localize at

sites of DNA replication [81]. In addition, unlike other

metabolic enzymes, MTHFD2 possesses a remarkable short

half-life, which resembles enzymes in RNA metabolism,

cell cycle, or signaling proteins [82]. Furthermore, in nor-

mal tissue, MTHFD2 was reported to be transcribed but not

translated [83]. Understanding these non-enzymatic func-

tions as well as why MTHFD2 is turned over rapidly, is

important for optimal targeting of MTHFD2 in cancer cells.

MTHFD2L: NADP+-dependent

methylenetetrahydrofolate dehydrogenase

The Mthfd2l gene contains nine exons and eight introns.

Alternative splicing results in ﬁve isomers due to skipping/

retaining of exons 2 and/or 8, which are observed in human

brain and placenta [84]. The ﬁve isomers are the canonical

isoform 4 (full-length; 347 amino acids; skipped exon 2 and

exon 8); isoform 2 (1–277 missing [encoding 70 amino

acids]), which retains exon 2 leading to the translation

start from the ﬁrst AUG codon in exon 1 and an early

termination, isoform 1 (1–58 missing; 289 amino acids;

retains exon 2) with translation from the AUG codon in

exon 3; isoforms 3 (1–58 missing, 282–347 missing; 223

amino acids) and 5 (269–310 missing; 305 amino acids)

both of which are skipping exon 2 and/or exon 8. While the

residues 1–20 are encoding for the targeting peptide for its

efﬁcient translocation into mitochondria, only isoforms 4

and 5 include the targeting peptide and are able to be

imported into the mitochondria. However, isoform 5 is

missing the residues from 269 to 310, which may sig-

niﬁcantly lower its afﬁnity for binding the substrate. Since

retention of exon 2 is rare, the isomers 1 and 2 are expressed

at low levels and currently detailed studies of the isoforms 3

and 5 are missing. However, MTHFD2L is a bifunctional

enzyme with CH2-THF dehydrogenase and 5,10-methenyl-

THF cyclohydrolase activities and is expressed in both

embryonic tissues and adult tissues. It is believed to have

little importance in cancer development and no association

with proliferation has been reported [85].

MTHFD2L can use either NAD+or NADP+as cofac-

tors, and for mono- and poly-glutamylated substrates with

similar catalytic efﬁciencies [26]. The kcat value for

MTHFD2L in the presence of NAD+for CH2-H4PteGlu1

and CH2-H4PteGlu5is 2.7 s−1and 8.8 s−1, respectively [15].

Using the 5,10-CH2-H4PteGlu1substrate under saturating

conditions, the dehydrogenase activity of MTHFD2L in the

presence of NAD+was 3.4-fold higher than the presence of

NADP+[80]. This data may indicate that MTHFD2L pre-

fers NAD+as a substrate.

MTHFD1L: methylenetetrahydrofolate

dehydrogenase (NADP+-dependent) 1-like protein

The current genomic annotation of the Mthfd1l gene shows

that it contains 31 exons and 30 distinct GT-AG introns, which

is different from the early characterization with 28 exons plus

one alternative exon [86]. Alternative splicing results in ﬁve

mRNAs variants and one unspliced form. Five spliced

mRNAs encode two complete isoforms (with one missing

residue 276–978) while the unspliced form is apparently

noncoding. MTHFD1L consists of the transit peptide (1–31)

and two major domains: an inactive methylene-THF dehy-

drogenase and cyclohydrolase N-terminal domain (32–347)

and an active larger 10-CHO-THF synthetase C-terminal

domain (348–977). Its 10-CHO-THF synthetase activity

complements MTHFD2L in normal (in embryonic tissue) and

transformed cells in the last step of the 1C pathway. It is

expressed at all stages of embryogenesis and in adults, parti-

cularly highly expressed in the neural tube, developing brain,

limb, and tail bud, as well as in the craniofacial structures. The

deletion of Mthfd1l gene is strongly associated with birth

defects characterized with an aberrant neural tube and cra-

niorachischisis [87]. Mthfd1l is transcriptionally activated by

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

NRF2, overexpressed in multiple cancer types, and plays an

essential role in cancer-cell growth [38]. However, the func-

tion of MTHFD1L in normal tissues needs to be characterized

before evaluating the therapeutic potential of MTHFD1L in

cancer cells.

ALDH1L2: aldehyde dehydrogenase 1 family

member L2

The Aldh1l2 gene contains 23 exons and 22 introns [88].

Alternative splicing results in three isoformes including the

full-length isomer 1, isomer 2 is missing residues from 204 to

923,andisoform3ismissingresiduesfrom517to923.Since

the region 438–923 is important for aldehyde dehydrogenase

activities, the uncharacterized isomer 2 and isoform 3 may be

catalytically inactive. The loss of ALDH1L2 impairs mito-

chondrial functions, leads to distorted mitochondrial mor-

phology, and is associated with accumulation of acylcarnitine

derivatives and Krebs cycle intermediates [89]. ALDH1L2 is

an integral part of mitochondria and plays an important role in

the energy balance of the cell. Because of these essential

functions, it is unlikely to be a useful therapeutic target.

Other therapeutic targets in the 1C pathway

DHFR (EC:1.5.1.3): dihydrofolate reductase

The Dhfr gene contains six exons and ﬁve introns with strictly

conserved intron and exon boundaries. Alternative splicing

results in two isoforms of which one is the canonical one and

the other missing residues 1–52. DHFR is a key enzyme in

folate metabolism as it generates 5,10-methylene-THF for

the subsequent 1C metabolism [90]. Most importantly, it is

involved in the de novo mitochondrial thymidylate biosynth-

esis pathway and catalyzes de novo glycine and purine

synthesisaswellasDNAprecursorsynthesis[91].

Prior to the 1C mitochondrial pathway, tetrahydrofolic

acid is reduced from dihydrofolic acid (DHF) by DHFR

using NADPH as electron donor. DHFR is overexpressed,

notably in acute lymphoblastic leukemia and is associated

with poor survival [92]. DHFR as a drug target has been

widely studied as antibacterial, antifungal, antimalarial, and

anticancer agent, and it has been intensively studied for its

antifolate properties in cancer chemotherapy [93]. Current

DHFR inhibitors include: (a) pyrimethamine [94]: a folic-

acid antagonist, used as antimalarial drug but now mainly

used with leucovorin to treat the cystoisosporiasis, which is

potent for protozoal DHFRs but weak for mammalian

DHFRs [95]; (b) Proguanil [96]: used to treat and prevent

malaria; (c) trimethoprim [97]: an antibacterial drug com-

monly used in association with sulfonamides (altogether

known as trimethoprim-sulfamethoxazole or co-trimoxazole)

to reinforce its action and prevent development of drug

resistance. Proguanil is a potent inhibitor for bacterial

DHFRs and a weak inhibitor for mammalian DHFRs; and (d)

methotrexate (MTX): an analog of folic acid, used in the

clinical setting as a chemotherapy agent and immune system

suppressant. MTX inhibits DHFR with a high afﬁnity by

forming an inactive ternary complex with DHFR and

NADPH through the binding of one or more glutamate

residues to form polyglutamate derivatives, thus reducing the

amount of THF required for the synthesis of pyrimidine,

purines, and several other amino acids [98,99]. MTX has

been used for leukemia and choriocarcinoma treatment.

However, MTX lacks speciﬁcity and has substantial and

undesirable cytotoxicity [100]. DHFR, as an established drug

target for many Food and Drug Administration-approved

drugs [101], would beneﬁt from further drug development in

order to reduce side effects.

Dhfr was altered in 4.4% of 91 cases from TCGA Pan-

Can 2018 dataset with deep deletion in ovarian and prostate

cancer and signiﬁcant mutations in uterine cancer. Forty-

one examined patient samples exhibit 40 Dhfr mutations

across 14 projects from GDC. We note that S119P/Y,

R77K, and E78G are within DHFR active center, which

may affect DHFR normal function and the drug efﬁcacy.

N13S may affect the Met-20 function loop (residues 9–24)

of DHFR, for the other two function loop: the FG loop

(residues 116–132) and GH loop (residues 142–149), a

distant mutant K122N may affect DHFR function. One

extra factor will compound the difﬁculty in predicting,

which mutation will further affect DHFR functions as a

network of dynamically coupled residues in DHFR collec-

tively promoting DHFR activities was suggested [102].

From the literature, we notice that D152V was reported to

cause megaloblastic anemia and cerebral folate deﬁciency

resulting in severe neurologic disease [103].

From DepMap, we noticed that there are 553 (out of 789)

cancer-cell lines whose proliferation will be affected by the

loss of DHFR function. However, DHFR shows a selec-

tivity of 0.186 with lung cancer, leukemia, skin cancer,

breast cancer, ovarian cancer, and pancreatic cancer will

beneﬁt most from DHFR inhibition.

TYMS (EC:2.1.1.45): thymidylate synthetase (TS)

The Tyms gene contains 12 exons and is separated by 11

distinct introns. Alternative splicing results in three isoforms,

one of which is full length (313 amino acids), another one is

missing residues 152–185 and expressed in normal and can-

cerous cells, while the third one is missing residues 69–151

and is expressed only in cancerous tissues. In the presence of

serine and NADPH, mitochondrial SHMT2 and TYMS,

together with human mitochondrial DHFR, convert deoxyur-

idylate (dUMP) to deoxythymidylate (dTMP) and this

L. N. Zhao et al.

contributes to the de novo mitochondrial thymidylate bio-

synthesis pathway [91]. TS forms an essential step in DNA

synthesis and DNA repair [104]. Its expression level has been

used as a biomarker in predicting the outcome of treatment in

prostate cancer as its overexpression is associated with a poor

survival rate and aggressive tumor features [105].

Over the previous half a century, several inhibitors have

been developed to target TS including: (a) Raltitrexed

(ZD1694; tomudex) [106]: one of the strongest quinazoline

(folate-based) antimetabolite in use with an IC50 of 9 nM; (b)

Pemetrexed [12]: a folate antimetabolite, which inhibits TS,

DHFR, and glycinamide ribonucleotide transformylase; (c)

Nolatrexed (AG337): with an IC50 between 0.39 and 6.6 µM,

water-soluble, and lipophilic nonclassical folate analog with

non-glutamate-containing moieties. The lipophilicity leads to

a short retention time, which reduces the active time in

the cell. Nolatrexed is currently in a Phase 3 clinical

trial (NCT00012324) [107]; and (d) Plevitrexed (ZD9331;

AstraZeneca): a non-polyglutamated quinazoline antifolate

drug with speciﬁcity toward TS, currently is in a Phase 2

clinical trial (NCT00014690); (e) 1843U89 (GW 1843):

benzoquinazoline folate analogue speciﬁcally targeting TS

[108]. Its liposomal formulation GS7904L is currently eval-

uated in the clinic [109]. It is noteworthy that the anti-

metabolite 5-ﬂuorouracil (5-FU), a pyrimidine antagonist,

will be anabolized to 5-ﬂuoro-2′-deoxyuridylate (FdUMP),

and as a nucleotide analogue of dUMP will covalently bind

TS and form a ternary complex with THF to inactive this

enzyme, thereby inhibiting DNA synthesis [110].

Tyms was altered in more than 6% of 182 cases from

TCGA PanCan 2018 dataset with noticeable ampliﬁcation in

esophagus, bladder, uterine, head and neck, lung, sarcoma,

pancreatic, and melanoma cancers. Sixty-one examined

patient samples exhibit involving 62 Tyms mutations across

22 projects from GDC. We note that D49N, D218E, M311L,

V79E, F80L, R78H, and S229R are not distant from active

center, which are likely to affect biochemical activities of

TYMS in terms of kcat or KMvalue and also affect the drug

efﬁcacy if the inhibitor is targeting the active site. In addition,

A191V and P188H are located at the active loop 181–197,

which may deactivate TYMS since this loop’sﬂipping is

important for substrate binding and closing of the active site

[111]. V3L and M190K were observed to stabilize the

inactive loop [112,113]. The other variants of TYMS are

important to evaluate for cancer chemotherapy since some

variants are important in determining 5-FU treatment and

cytotoxicity [114].

From DepMap, we noticed that there are 328 (out of 789)

cancer-cell lines whose proliferation will be affected by the

loss of TYMS function. However, TYMS shows a relatively

high selectivity of 0.766. Lung cancer, leukemia, skin cancer,

breast cancer, and pancreatic cancer are among the top cell

lines that are sensitive toward TYMS loss-of-function.

GARS-AIRS-GART (hmGART): glycinamide

ribonucleotide synthetase (GARS), aminoimidazole

ribonucleotide synthetase (AIRS), and glycinamide

ribonucleotide formyltransferase (GART)

The Gars-Airs-Gart gene (also abbreviated as Gart) encodes

two proteins: full-length trifunctional GARS-AIRS-GART,

composed of three activities: GARS (PurD, EC:6.3.4.13),

AIRS (PurM, EC:6.3.3.1), and GARTfase (PurN,

EC:2.1.2.2.). GARS-AIRS-GART is involved in the second,

third, and ﬁfth steps of de novo purine synthesis, and the

proximity of the three activities in one enzyme complex

and the ﬂexibility of these domains elevate the pathway

output, while the fourth step is catalyzed by phosphor-

ibosylformylglycineamide amidotransferase (FGARAT,

PurL). The other protein resulting from alternative splicing

is the monofunctional GARS protein [115]. The activity of

GART requires 10-formyltetrahydrofolate and has been the

subject of much research interest for antifolate drug devel-

opment [116]. Current inhibitors targeting GART include:

(a) Lometrexol ([6R]-dideazatetrahydrofolate, DDATHF): a

folate analog antimetabolite [117], currently in clinical trial

NCT00024310 (phase 1) and NCT00033722 (phase 2); (b)

Pelitrexol (AG 2037): developed by Pﬁzer and has com-

pleted two phase 2 trials by 2006 in the US [118]; (c)

LY254155 and LY222306: differ from lometrexol in the

replacement of the phenylene moiety by a thiophene and a

furan, respectively, which may have clinical activity; (d)

Pemetrexed disodium hemipentahydrate (LY231514): can

inhibit GART with an IC50 of 65 nM, and inhibit TS and

DHFR at 1.3 and 7.2 nM, respectively, is being tested in

extensive clinical trials (several Phase 3 clinical trials both

active and completed) [119].

One hundred and seventy-nine examined patient samples

exhibit displaying 195 Gart mutations across 26 projects

from GDC. G157R will affect manganese and phosphate

binding thus compromising the catalytic activities. E190K/D

will directly affect the ATP binding afﬁnity, and L193F/I will

slightly affect the KMof ATP. Detailed structural studies

have been done for the region 467–794 [120]. For 10-

formyltetrahydrofolate binding regions (896–899 and

947–951), we note that R871C, E948G, D951V, and L899I

will affect the 10-formyltetrahydrofolate binding; for

5’-phosphoribosylglycinamide binding region (818–820 and

977–980), K977N will signiﬁcantly affect biochemical

activities of GART in terms of kcat or KMvalue and also

affect the drug efﬁcacy if the inhibitor is targeting the active

site. D951V mutation disables its ability to raise pKa for the

active site His residue and the frameshift in G924Vfs*23

needs to be studied carefully as it is in the proximity of the

active center. This data indicate that there are mutations in

GART that will affect drug binding and this needs to be

considered in drug development.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

From DepMap, we noticed that there are 184 (out of 789)

cancer-cell lines whose proliferation will be affected by the

loss of GART function. However, GART shows a relatively

high selectivity of 0.459. Lung cancer, leukemia, breast

cancer, colon/colorectal cancer, myeloma, lymphoma, and

gastric cancer are among the top cancer types that should be

sensitive toward GART loss-of-function.

We have described studies of the mutations of each gene

among different cancer cases, since even a single mutation

may have profound implications for cancer development and

treatment. Therefore, it is important to note that usually a set

of various mutations across multiple genes collectively

contributes to cancer pathology. Taken together, this sum-

mary of the genes in the 1C pathway provides a basis for us

to study in-depth the roles of these enzymes in proliferation,

differentiation, and metastasis. In addition, a deeper under-

standing of the 1C pathway is important for the rational

design of drug combinations in the treatment of cancer.

Discussion

Potential pitfalls in targeting folate-mediated

metabolic enzymes

In order to develop a safe and speciﬁcinhibitor,thereare

several key enzymes we need to consider besides the well-

established MTHFD2. First and foremost there is

MTHFD1, as a close analogue of MTHFD2. MTHFD1 is

expressed in normal tissues and most abundantly in the

liver [76]. Inhibition of MTHFD1 would most likely cause

unwanted side effects since (a) MTHFD1 deﬁciency

results in a phenotype that includes severe combined

immunodeﬁciency and megaloblastic anemia [121]; and

(b) MTHFD1 inhibition disrupts the nuclear metabolite

composition and alters gene expression, which impairs

fast-growing cells [122]. Meanwhile, MTHFD2L, sharing

60–65% identity and the highly conserved substrate and

cofactor binding pocket with MTHFD2, is expressed in all

tissues examined, with the highest expression levels in

brain and lung in both humans and rodents [55,57].

Unfortunately, there are no studies whether inhibitors of

MTHFD2 or MTHFD1 affect the activity of MTHFD2L.

Therefore, a selective inhibitor for MTHFD2 without

activity against MTHFD1 and MTHFD2L would be

desirable as a lead for drug discovery to warrant successful

clinical investigations. As discussed previously, the

cofactor (NAD+) pocket and the substrate (CH2-THF)

pocket of MTHFD2 are not speciﬁc enough as NAD(P)

+-dependent enzymes are ubiquitous in metabolism and

other cellular processes [51]. This substrate (folate) pocket

is also present in enzymes of the THF-mediated 1C

metabolic pathway, which has been the basis for the

antifolate drug development [123]. Therefore inhibitors

that target either pocket will likely result in undesired side

effects. Hence, we are proposing to exploit a novel inte-

grated pocket, different from the current cofactor and

substrate clefts, which is formed at the interface between

twoMTHFD2monomerswhentheydimerize(manuscript

in preparation), which could serve as an allosteric site for

the structure-based drug discovery.

Studies have shown that MTHFD2 knockdown only

inhibits tumor growth but is not sufﬁcient to kill cancer cells

[64]. This could be due to incomplete silencing of MTHFD2

or due to compensation by other enzymes. Therefore, it is

important to analyse CRISPR knockout cells as well as

consider additional enzymes of the 1C pathway as combi-

natorial treatments to enhance efﬁciency. This has been done

in the DepMap project and the proliferation of only four

cancer-cell lines is affected by Mthfd2 KO (see above).

SHMT2 is such a potential candidate for combinatorial

treatment. As discussed earlier, SHMT2 is overexpressed in

multiple cancer lines similar as MTHFD2 and the knock-

down of SHMT2 was found to suppress tumorigenesis in

human hepatocellular carcinoma [58]. As a potential ther-

apeutic target in combination with MTHFD2, weak inhibi-

tion of SHMT2 could be desirable to reduce its activity and

to enhance MTHFD2 inhibition. Overall, this would provide

a new perspective in the anticancer strategy targeting the 1C

pathway.

Another avenue for intervention is at the transcriptional

level. MYC binds directly at Mthfd2,Mthfd1l, and Shmt2

promoters [124], and c-myc also targets other genes such as

Phgdh,Psph,Slc19a1,Dhfr,Tyms,Gart,Shmt1,Mthfd1,

Fgps, and Gcsh that are involved in serine, folate, and

glycine metabolism [125,126]. Since myc activation cor-

relates or is fundamentally responsible for the over-

expression of Mthfd2,Mthfd1l, and Shmt2 genes [124],

targeting myc activation may be the last resort but is riddled

with difﬁculties. This is mostly due to the fact that it is

difﬁcult to target transcription factors although for myc,

there has been some progress with ONCO-myc [127].

Better than MTX?

MTX is an “old school”but mainstay drug that has been used

in the clinic since 1947 to treat cancer [128]. There are many

reasons why this drug is used, one main reason is that it was

less toxic than the then-current anticancer regimens but MTX

causes a plethora of side effects including haematopoietic- and

hepatic-toxicity, dizziness, drowsiness, etc [129–131]. It is

clear that MTX exerts its efﬁcacy on DHFR and TS from the

1C pathway; however, off-target(s) of MTX are not well

known especially since it acts as anti-inﬂammatory agent

[132]. Furthermore, it is known that many cancers become

quickly resistant to treatment with MTX, which is a major

L. N. Zhao et al.

issue [133–136]. In summary, although MTX works, it is not a

perfect drug and improvements would be desirable. In com-

parison with MTX, we would expect that inhibitors of

MTHFD2 and/or SHMT2 have the following advantages:

since MTHFD2 is only expressed in cancer cells and not adult

cells, MTHFD2 inhibitors should be potent enough to elim-

inate cancer cells while sparing healthy cells, thereby reducing

the side effects.

Compartmentalization of the 1C pathway

The rationale for compartmentalization of the 1C pathway

has been described previously [22] but there are other

considerations of the compartmentalization. As an evolu-

tionary strategy, 1C metabolism spatially orchestrates into

distinct compartments to enable cellular function properly

(or more accurately) and to decrease interference with each

other [23]. Each compartment conﬁnes/concentrates spe-

ciﬁc enzyme complexes as well as their substrates and

products with speciﬁc microenvironments to optimize

subcellular biochemical processes. This compartmentali-

zation has all the advantages it needs for its function.

However, there are advantages and challenges in drug

development when targeting mitochondrial enzymes like

MTHFD2 and SHMT2. One big challenge to deliver drugs

to the mitochondria given that they have to cross two

membranes (outer and inner mitochondrial membrane) and

the intermembrane space. In addition, the chemical condi-

tions are different in mitochondria compared to the cytosol.

However, diverse delivery mechanisms have been proven

to be successful in delivering prodrugs or organelle-speciﬁc

chemical agents to its place of action [137]. As an effective

strategy to target mitochondria, the compartmentalization

could prolong drug retention in mitochondria, which may

enhance the drug efﬁciency. At the same time, since there

is crosstalk between the cytoplasmic and mitochondrial 1C

pathway, one has to wonder whether targeting a cyto-

plasmic 1C enzyme could simultaneously inhibit the

mitochondrial 1C pathway. One possibility is the mito-

chondrial folate transporter (MFT encoded by the Slc25a32

gene), which imports THR from the cytosol to mitochon-

drial. CRISPR KO of Slc25A32 only affected proliferation

of four cancer-cell lines (DepMap), which could indicate

that inhibiting MFT in combination with either MTHFD2

or SHMT2 could be successful. In summary, compart-

mentalization can be viewed both as a drawback or

opportunity. Since the mitochondrial and cytoplasmic 1C

pathways can compensate for each other, this causes

additional problems for drug development. Nevertheless, if

the nodes where the two pathways interact are being con-

sidered, one could envision taking advantage of this by

affecting both the cytoplasmic and mitochondrial pathway

at the same time.

Open questions

Overall, there are uncharted territories in the mitochondrial

1C metabolism, which should be considered in future drug

development:

(1) For MTHFD2, one aspect that needs clarifying is

whether the transcriptional regulation that leads to

high expression of the mRNA and protein in the

proliferation stage of tumors are the same as in

embryonic development. Enzymes that are expressed

only in proliferating cells but not in adult tissues can

be considered as a potential anti-proliferation target.

(2) A previous study suggested that proliferation and

invasion are mutually exclusive [84]. This should be

considered when designing drugs targeting proliferat-

ing cells, since cell-cycle-arrested cells could become

more prone to invasion leading to metastasis [138].

Thus it is of paramount importance when inhibiting

cell division to investigate whether cancer cells arrest

and invade or whether they undergo cell death. The

latter would be obviously more desirable.

(3) A substantial pitfall in therapeutic inhibition of the

mitochondrial 1C pathway is that the loss of the

mitochondrial activity generating 1C units could be

compensated by cytosolic metabolism [139]. Both

cytosolic and mitochondrial pathways could sustain

tumorigenesis while the mitochondrial pathway is

essential in nutrient-poor conditions [139]. Thus the

interplay of the mitochondrial and cytosolic 1C

pathway needs to be investigated in more detail (as

discussed above).

(4) To what extent “overexpression”matters for metabolic

enzymes in general and particularly for MTHFD2, which

has a catalytic efﬁciency of 12.4 s−1?Does“over-

expression”really elevate catalytic output of MTHFD2?

In addition, the half-life of proteins are generally in the

range of 0.5–35 h in dividing mammalian cells (~24 h

cell cycle) and ~43 h in nondividing cells [140].

Nevertheless, the time for protein synthesis takes

between 20 s to several minutes (average 2 min) [141],

which would indicate that protein synthesis may be more

important than the half-life of a protein. Therefore, a

quantitative understanding of protein levels, turn-over,

and catalytic activity is desirable.

(5) In general, a tumor biomarker should be very speciﬁc,

for example the FLT3 internal tandem duplication

(FLT3/ITD) has been reported to be a biomarker for the

response to MTHFD2 inhibition in AML [124].

However, for other tumor types, a simple and rapid

tests to predict whether a tumor will respond clinically

to inhibition of MTHFD2 would be very useful for

drug development.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

Concluding remarks

In this review, we are not only making an attempt to

investigate comprehensively the key enzymes in the 1C

mitochondrial pathway but also consider their counterparts

in the cytosol, based on conﬁrmatory evidence and their

therapeutic perspectives. By integrating experimental data,

and literature information regarding the key enzymes in

mitochondrial 1C pathway in tumors and microorganisms,

we come to the following conclusions: (1) MTHFD2 is an

excellent therapeutic target; (2) SHMT2 is a good drug

target; (3) MTHFD1L is a potential drug target; and (4)

MTHFD2L and ALDH1L2 should not be considered for

drug development. Meanwhile, based on structural/bio-

chemical and up-to-date inhibitor development, we have

given special attention to MTHFD2 and SHMT2 as a

complementary target, which will deepen our understanding

of their therapeutic potential. In addition, we have provided

a brief review of other important enzymes in the 1C path-

way, which are well-established drug targets, some of which

are widely used in the clinic. Even for the established drugs

more efforts are needed to decrease the side effects. Atten-

tion shall also be given to less developed inhibitors, which

could be potential game changers in the future. Rather than

restricting our attention to gene “overexpression,”we have

focused on the functional importance in normal cells to

identify which should be therapeutically targeted.

Overall, we have discussed some important pitfalls and

possible ways to move forward in targeting the 1C meta-

bolism for cancer therapy. Since the 1C pathway is a

dynamic, interdependent network, the chemical steps cata-

lyzed by the 1C enzymes are coupled to the ﬂow of the

folic-acid derivatives in this entire network. It is important

to systematically study the 1C pathway to capture the

dynamics of these biochemical processes, which will pro-

vide a better understanding of this pathway in terms of

overexpression as well as hyperactivity in cancer cells. This

may eventually open new therapeutic perspectives for

treating various cancers safely and efﬁciently.

Acknowledgements LNZ was supported by an A*STAR International

Fellowship (AIF) from Singapore. This work is supported by the

IngaBritt och Arne Lundbergs Forskningsstiftelse LU2020-0013; PK

is supported by the Faculty of Medicine, Lund University; the Swedish

Foundation for Strategic Research Dnr IRC15-0067, and Swedish

Research Council, Strategic Research Area EXODIAB, Dnr

2009–1039. LNZ would like to thank Prof. Chew Lock Yue (NTU)

and Prof. Ulf Ryde (LU) for comments on the manuscript.

Compliance with ethical standards

Conﬂict of interest The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to

jurisdictional claims in published maps and institutional afﬁliations.

References

1. Pavlova NN, Thompson CB. The emerging hallmarks of cancer

metabolism. Cell Metab. 2016;23:27–47.

2. Vander Heiden MG. Targeting cancer metabolism: a therapeutic

window opens. Nat Rev Drug Discov. 2011;10:671–84.

3. McGuirk S, Audet-Delage Y, St-Pierre J. Metabolic ﬁtness and

plasticity in cancer progression. Trends Cancer. 2020;6:49–61.

4. Kreuzaler P, Panina Y, Segal J, Yuneva M. Adapt and conquer:

metabolic ﬂexibility in cancer growth, invasion and evasion. Mol

Metab. 2020;33:83–101.

5. Amelio I, Cutruzzolá F, Antonov A, Agostini M, Melino G.

Serine and glycine metabolism in cancer. Trends Biochem Sci.

2014;39:191–8.

6. Locasale JW. Serine, glycine and one-carbon units: cancer

metabolism in full circle. Nat Rev Cancer. 2013;13:572–83.

7. Choi Y-K, Park K-G. Targeting glutamine metabolism for cancer

treatment. Biomol Ther. 2018;26:19–28.

8. Fadaka A, Ajiboye B, Ojo O, Adewale O, Olayide I, Emuo-

whochere R. Biology of glucose metabolization in cancer cells. J

Oncological Sci. 2017;3:45–51.

9. Shiratori R, Furuichi K, Yamaguchi M, Miyazaki N, Aoki H,

Chibana H, et al. Glycolytic suppression dramatically changes

the intracellular metabolic proﬁle of multiple cancer cell lines

in a mitochondrial metabolism-dependent manner. Sci Rep.

2019;9:18699.

10. Lin J, Xia L, Liang J, Han Y, Wang H, Oyang L, et al. The roles

of glucose metabolic reprogramming in chemo- and radio-

resistance. J Exp Clin Cancer Res. 2019;38:218.

11. Farber S, Diamond LK, Mercer RD, Sylvester RF, Wolff JA.

Temporary remissions in acute leukemia in children produced by

folic acid antagonist, 4-aminopteroyl-glutamic acid (ami-

nopterin). N Engl J Med. 1948;238:787–93.

12. Chattopadhyay S, Moran RG, Goldman ID. Pemetrexed: bio-

chemical and cellular pharmacology, mechanisms, and clinical

applications. Mol Cancer Ther. 2007;6:404–17.

13. Estrada A, Wright DL, Anderson AC. Antibacterial antifolates:

from development through resistance to the next generation.

Cold Spring Harb Perspect Med. 2016;6. https://doi.org/10.1101/

cshperspect.a028324.

14. Purcell WT, Ettinger DS. Novel antifolate drugs. Curr Oncol

Rep. 2003;5:114–25.

15. Carver JR, Szalda D, Ky B. Asymptomatic cardiac toxicity in

long-term cancer survivors: deﬁning the population and recom-

mendations for surveillance. Semin Oncol. 2013;40:229–38.

16. Yeh Edward TH, Tong Ann T, Lenihan Daniel J, Wamique

YusufS, Joseph Swafford, Christopher Champion, et al. Car-

diovascular complications of cancer therapy. Circulation.

2004;109:3122–31.

17. Wang Y, Probin V, Zhou D. Cancer therapy-induced residual

bone marrow injury-mechanisms of induction and implication

for therapy. Curr Cancer Ther Rev. 2006;2:271–9.

18. Appling DR. Compartmentation of folate-mediated one-carbon

metabolism in eukaryotes. FASEB J.1991;5(12):2645–51.

19. Toriello HV. Folic acid and neural tube defects. Genet Med.

2005;7:283–4.

20. Kujovich JL. Evaluation of anemia. Obstet Gynecol Clin North

Am. 2016;43:247–64.

21. Fox JT, Stover PJ. Folate-mediated one-carbon metabolism.

Vitam Horm. 2008;79:1–44.

22. Ducker GS, Rabinowitz JD. One-carbon metabolism in health

and disease. Cell Metab. 2017;25:27–42.

23. Agapakis CM, Boyle PM, Silver PA. Natural strategies for the

spatial optimization of metabolism in synthetic biology. Nat

Chem Biol. 2012;8:527–35.

L. N. Zhao et al.

24. Kit S. The biosynthesis of free glycine and serine by tumors.

Cancer Res. 1955;15:715–8.

25. Labuschagne CF, van den Broek NJF, Mackay GM, Vousden

KH, Maddocks ODK. Serine, but not glycine, supports one-

carbon metabolism and proliferation of cancer cells. Cell Rep.

2014;7:1248–58.

26. Shin M, Bryant JD, Momb J, Appling DR. Mitochondrial

MTHFD2L is a dual redox cofactor-speciﬁc methylenetetrahy-

drofolate dehydrogenase/methenyltetrahydrofolate cyclohy-

drolase expressed in both adult and embryonic tissues. J Biol

Chem. 2014;289:15507–17.

27. Cybulski RL, Fisher RR. Intramitochondrial localization and

proposed metabolic signiﬁcance of serine transhydrox-

ymethylase. Biochemistry. 1976;15:3183–7.

28. Balut C, vandeVen M, Despa S, Lambrichts I, Ameloot M,

Steels P, et al. Measurement of cytosolic and mitochondrial pH

in living cells during reversible metabolic inhibition. Kidney Int.

2008;73:226–32.

29. Boron WF. Regulation of intracellular pH. Adv Physiol Educ.

2004;28:160–79.

30. Shirmanova MV, Druzhkova IN, Lukina MM, Matlashov ME,

Belousov VV, Snopova LB, et al. Intracellular pH imaging in

cancer cells in vitro and tumors in vivo using the new genetically

encoded sensor SypHer2. Biochim Biophys Acta (BBA) - Gen

Subj. 2015;1850:1905–11.

31. White KA, Grillo-Hill BK, Barber DL. Cancer cell behaviors

mediated by dysregulated pH dynamics at a glance. J Cell Sci.

2017;130:663–9.

32. Hu Y, Li Y. Effect of low pH treatment on cell cycle and cell

growth. FASEB J. 2018;32:804.49.

33. Vazquez A, Tedeschi PM, Bertino JR. Overexpression of the

mitochondrial folate and glycine–serine pathway: a new determinant

of methotrexate selectivity in tumors. Cancer Res. 2013;73:478–82.

34. Noguchi K, Konno M, Koseki J, Nishida N, Kawamoto K,

Yamada D, et al. The mitochondrial one-carbon metabolic

pathway is associated with patient survival in pancreatic cancer.

Oncol Lett. 2018;16:1827–34.

35. Nilsson R, Jain M, Madhusudhan N, Sheppard NG, Strittmatter

L, Kampf C, et al. Metabolic enzyme expression highlights a key

role for MTHFD2 and the mitochondrial folate pathway in

cancer. Nat Commun. 2014;5:3128.

36. Agarwal S, Behring M, Hale K, Al Diffalha S, Wang K, Manne

U, et al. MTHFD1L, a folate cycle enzyme, is involved in pro-

gression of colorectal cancer. Transl Oncol. 2019;12:1461–7.

37. Eich M-L, Rodriguez Pena MDC, Chandrashekar DS, Chaux A,

Agarwal S, Gordetsky JB, et al. Expression and role of methy-

lenetetrahydrofolate dehydrogenase 1 like (MTHFD1L) in

bladder cancer. Transl Oncol. 2019;12:1416–24.

38. Lee D, Xu IM-J, Chiu DK-C, Lai RK-H, Tse AP-W, Lan Li L, et al.

Folate cycle enzyme MTHFD1L confers metabolic advantages in

hepatocellular carcinoma. J Clin Investig. 2017;127:1856–72.

39. Wu M, Wanggou S, Li X, Liu Q, Xie Y. Overexpression of

mitochondrial serine hydroxyl-methyltransferase 2 is associated

with poor prognosis and promotes cell proliferation and invasion

in gliomas. Onco Targets Ther. 2017;10:3781–8.

40. Tedeschi PM, Vazquez A, Kerrigan JE, Bertino JR. Mitochondrial

methylenetetrahydrofolate dehydrogenase (MTHFD2) over-

expression is associated with tumor cell proliferation and is a

novel target for drug development. Mol Cancer Res. 2015;13:

1361–6.

41. Liu Y, Yin C, Deng M-M, Wang Q, He X-Q, Li M-T, et al. High

expression of SHMT2 is correlated with tumor progression and

predicts poor prognosis in gastrointestinal tumors. Eur Rev Med

Pharm Sci. 2019;23:9379–92.

42. Di Pietro E, Sirois J, Tremblay ML, MacKenzie RE. Mitochon-

drial NAD-dependent methylenetetrahydrofolate dehydrogenase-

methenyltetrahydrofolate cyclohydrolase is essential for embryo-

nic development. Mol Cell Biol. 2002;22:4158–66.

43. Stover PJ, Chen LH, Suh JR, Stover DM, Keyomarsi K, Shane

B. Molecular cloning, characterization, and regulation of the

human mitochondrial serine hydroxymethyltransferase gene. J

Biol Chem. 1997;272:1842–8.

44. Florio R, di Salvo ML, Vivoli M, Contestabile R. Serine

hydroxymethyltransferase: a model enzyme for mechanistic,

structural, and evolutionary studies. Biochim Biophys Acta.

2011;1814:1489–96.

45. Giardina G, Brunotti P, Fiascarelli A, Cicalini A, Costa MGS,

Buckle AM, et al. How pyridoxal 5′-phosphate differentially

regulates human cytosolic and mitochondrial serine hydro-

xymethyltransferase oligomeric state. FEBS J. 2015;282:1225–41.

46. Yang X, Wang Z, Li X, Liu B, Liu M, Liu L, et al. SHMT2

Desuccinylation by SIRT5 drives cancer cell proliferation.

Cancer Res. 2018;78:372–86.

47. Morscher RJ, Ducker GS, Li SH-J, Mayer JA, Gitai Z, Sperl W,

et al. Mitochondrial translation requires folate-dependent tRNA

methylation. Nature. 2018;554:128–32.

48. Lucas S, Chen G, Aras S, Wang J. Serine catabolism is essential

to maintain mitochondrial respiration in mammalian cells. Life

Sci Alliance. 2018;1:e201800036.

49. Anderson DD, Stover PJ. SHMT1 and SHMT2 are functionally

redundant in nuclear de novo thymidylate biosynthesis. PLoS

ONE. 2009;4:e5839.

50. Guiducci G, Paone A, Tramonti A, Giardina G, Rinaldo S,

Bouzidi A, et al. The moonlighting RNA-binding activity of

cytosolic serine hydroxymethyltransferase contributes to control

compartmentalization of serine metabolism. Nucleic Acids Res.

2019;47:4240–54.

51. Anderson DD, Eom JY, Stover PJ. Competition between sumoy-

lation and ubiquitination of serine hydroxymethyltransferase 1

determines its nuclear localization and its accumulation in the

nucleus. J Biol Chem. 2012;287:4790–9.

52. Jain M, Nilsson R, Sharma S, Madhusudhan N, Kitami T, Souza

AL, et al. Metabolite proﬁling identiﬁes a key role for glycine in

rapid cancer cell proliferation. Science. 2012;336:1040–4.

53. Lee GY, Haverty PM, Li L, Kljavin NM, Bourgon R, Lee J, et al.

Comparative oncogenomics identiﬁes PSMB4 and SHMT2 as

potential cancer driver genes. Cancer Res. 2014;74:3114–26.

54. Tani H, Ohnishi S, Shitara H, Mito T, Yamaguchi M, Yone-

kawa H, et al. Mice deﬁcient in the Shmt2 gene have mito-

chondrial respiration defects and are embryonic lethal. Sci Rep.

2018;8:425.

55. Minton DR, Nam M, McLaughlin DJ, Shin J, Bayraktar EC,

Alvarez SW, et al. Serine catabolism by SHMT2 is required for

proper mitochondrial translation initiation and maintenance of

formylmethionyl-tRNAs. Mol Cell. 2018;69:610–21.e5.

56. Ning S, Ma S, Saleh AQ, Guo L, Zhao Z, Chen Y. SHMT2

overexpression predicts poor prognosis in intrahepatic cho-

langiocarcinoma. Gastroenterol Res Pract. 2018;2018:4369253.

57. Wang H, Chong T, Li B-Y, Chen X-S, Zhen W-B. Evaluating

the clinical signiﬁcance of SHMT2 and its co-expressed gene in

human kidney cancer. Biol Res. 2020;53:46.

58. Woo CC, Chen WC, Teo XQ, Radda GK, Teck Hock Lee P.

Downregulating serine hydroxymethyltransferase 2 (SHMT2)

suppresses tumorigenesis in human hepatocellular carcinoma.

Oncotarget. 2016;7:53005–17.

59. NonakaH,NakanishiY,KunoS,OtaT,MochidomeK,Saito

Y, et al. Design strategy for serine hydroxymethyltransferase

probes based on retro-aldol-type reaction. Nat Commun.

2019;10:876.

60. Snell K, Riches D. Effects of a triazine antifolate (NSC 127755)

on serine hydroxymethyltransferase in myeloma cells in culture.

Cancer Lett. 1989;44:217–20.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

61. Stover P, Schirch V. 5-Formyltetrahydrofolate polyglutamates are

slow tight binding inhibitors of serine hydroxymethyltransferase. J

Biol Chem. 1991;266:1543–50.

62. Marani M, Paone A, Fiascarelli A, Macone A, Gargano M, Rinaldo

S, et al. A pyrazolopyran derivative preferentially inhibits the activity

of human cytosolic serine hydroxymethyltransferase and induces cell

death in lung cancer cells. Oncotarget. 2016;7:4570–83.

63. Gu Y, Si J, Xiao X, Tian Y, Yang S. miR-92a inhibits proliferation

and induces apoptosis by regulating methylenetetrahydrofolate

dehydrogenase 2 (MTHFD2) expression in acute myeloid leuke-

mia. Oncol Res. 2017;25:1069–79.

64. Koseki J, Konno M, Asai A, Colvin H, Kawamoto K, Nishida N,

et al. Enzymes of the one-carbon folate metabolism as anticancer

targets predicted by survival rate analysis. Sci Rep. 2018;8:303.

65. Yan Y, Zhang D, Lei T, Zhao C, Han J, Cui J, et al. MicroRNA-

33a-5p suppresses colorectal cancer cell growth by inhibiting

MTHFD2. Clin Exp Pharm Physiol. 2019;46:928–36.

66. Ju H-Q, Lu Y-X, Chen D-L, Zuo Z-X, Liu Z-X, Wu Q-N, et al.

Modulation of redox homeostasis by inhibition of MTHFD2 in

colorectal cancer: mechanisms and therapeutic implications. J

Natl Cancer Inst. 2018. https://doi.org/10.1093/jnci/djy160.

67. Xu T, Zhang K, Shi J, Huang B, Wang X, Qian K, et al.

MicroRNA-940 inhibits glioma progression by blocking mito-

chondrial folate metabolism through targeting of MTHFD2. Am

J Cancer Res. 2019;9:250–69.

68. Nishimura T, Nakata A, Chen X, Nishi K, Meguro-Horike M,

Sasaki S, et al. Cancer stem-like properties and geﬁtinib resistance

are dependent on purine synthetic metabolism mediated by the

mitochondrial enzyme MTHFD2. Oncogene. 2019;38:2464–81.

69. Lin H, Huang B, Wang H, Liu X, Hong Y, Qiu S, et al. MTHFD2

overexpression predicts poor prognosis in renal cell carcinoma and

is associated with cell proliferation and vimentin-modulated

migration and invasion. Cell Physiol Biochem. 2018;51:991–1000.

70. Asai A, Koseki J, Konno M, Nishimura T, Gotoh N, Satoh T,

et al. Drug discovery of anticancer drugs targeting methylene-

tetrahydrofolate dehydrogenase 2. Heliyon. 2018;4:e01021.

71. Mejia NR, MacKenzie RE. NAD-dependent methylenetetrahy-

drofolate dehydrogenase is expressed by immortal cells. J Biol

Chem. 1985;260:14616–20.

72. Mejia NR, MacKenzie RE. NAD-dependent methylenetetrahy-

drofolate dehydrogenase-methenyltetrahydrofolate cyclohydrolase

in transformed cells is a mitochondrial enzyme. Biochem Biophys

Res Commun. 1988;155:1–6.

73. Patel H, Pietro ED, MacKenzie RE. Mammalian ﬁbroblasts

lacking mitochondrial NAD+-dependent methylenetetrahy-

drofolate dehydrogenase-cyclohydrolase are glycine auxotrophs.

J Biol Chem. 2003;278:19436–41.

74. Yu C, Yang L, Cai M, Zhou F, Xiao S, Li Y, et al. Down-

regulation of MTHFD2 inhibits NSCLC progression by sup-

pressing cycle-related genes. J Cell Mol Med. 2020;24:1568–77.

75. Gustafsson R, Jemth A-S, Gustafsson NMS, Färnegårdh K,

Loseva O, Wiita E, et al. Crystal structure of the emerging cancer

target MTHFD2 in complex with a substrate-based inhibitor.

Cancer Res. 2017;77:937–48.

76. Fu C, Sikandar A, Donner J, Zaburannyi N, Herrmann J, Reck

M, et al. The natural product carolacton inhibits folate-

dependent C1 metabolism by targeting FolD/MTHFD. Nat

Commun. 2017;8:1529.

77. Kawai J, Ota M, Ohki H, Toki T, Suzuki M, Shimada T, et al.

Structure-based design and synthesis of an isozyme-selective

MTHFD2 inhibitor with a tricyclic coumarin scaffold. ACS Med

Chem Lett. 2019;10:893–8.

78. Kawai J, Toki T, Ota M, Inoue H, Takata Y, Asahi T, et al.

Discovery of a potent, selective, and orally available MTHFD2

inhibitor (DS18561882) with in vivo antitumor activity. J Med

Chem. 2019. https://doi.org/10.1021/acs.jmedchem.9b01113.

79. Rios-Orlandi EM, MacKenzie RE. The activities of the

NAD-dependent methylenetetrahydrofolate dehydrogenase-

methenyltetrahydrofolate cyclohydrolase from ascites tumor cells

are kinetically independent. J Biol Chem. 1988;263:4662–7.

80. Shin M, Momb J, Appling DR. Human mitochondrial MTHFD2

is a dual redox cofactor-speciﬁc methylenetetrahydrofolate

dehydrogenase/methenyltetrahydrofolate cyclohydrolase. Cancer

Metab. 2017;5:11.

81. Gustafsson Sheppard N, Jarl L, Mahadessian D, Strittmatter L,

Schmidt A, Madhusudan N, et al. The folate-coupled enzyme

MTHFD2 is a nuclear protein and promotes cell proliferation. Sci

Rep. 2015;5:15029.

82. Koufaris C, Nilsson R. Protein interaction and functional data

indicate MTHFD2 involvement in RNA processing and trans-

lation. Cancer Metab. 2018;6:12.

83. Kg P, Re M. NAD+-dependent methylenetetrahydrofolate

dehydrogenase-cyclohydrolase: detection of the mRNA in nor-

mal murine tissues and transcriptional regulation of the gene in

cell lines. Biochim Biophys Acta. 1993;1171:281–7.

84. Bolusani S, Young BA, Cole NA, Tibbetts AS, Momb J, Bryant

JD, et al. Mammalian MTHFD2L encodes a mitochondrial

methylenetetrahydrofolate dehydrogenase isozyme expressed in

adult tissues. J Biol Chem. 2011;286:5166–74.

85. Nilsson R, Nicolaidou V, Koufaris C. Mitochondrial MTHFD

isozymes display distinct expression, regulation, and association

with cancer. Gene. 2019;716:144032.

86. Prasannan P, Pike S, Peng K, Shane B, Appling DR. Human

mitochondrial C1-tetrahydrofolate synthase: gene structure, tissue

distribution of the mRNA, and immunolocalization in Chinese

hamster ovary calls. J Biol Chem. 2003;278:43178–87.

87. Momb J, Lewandowski JP, Bryant JD, Fitch R, Surman DR,

Vokes SA, et al. Deletion of Mthfd1l causes embryonic lethality

and neural tube and craniofacial defects in mice. PNAS.

2013;110:549–54.

88. Krupenko NI, Dubard ME, Strickland KC, Moxley KM, Oleinik

NV, Krupenko SA. ALDH1L2 is the mitochondrial homolog of

10-formyltetrahydrofolate dehydrogenase. J Biol Chem. 2010;

285:23056–63.

89. Sarret C, Ashkavand Z, Paules E, Dorboz I, Pediaditakis P, Sumner

S, et al. Deleterious mutations in ALDH1L2 suggest a novel cause

for neuro-ichthyotic syndrome. NPJ Genom Med. 2019;4:17.

90. Schober AF, Mathis AD, Ingle C, Park JO, Chen L, Rabinowitz

JD, et al. A two-enzyme adaptive unit within bacterial folate

metabolism. Cell Rep. 2019;27:3359–70.e7.

91. Anderson DD, Quintero CM, Stover PJ. Identiﬁcation of a de

novo thymidylate biosynthesis pathway in mammalian mito-

chondria. PNAS. 2011;108:15163–8.

92. Organista-Nava J, Gómez-Gómez Y, Illades-Aguiar B, Rivera-

Ramírez AB, Saavedra-Herrera MV, Leyva-Vázquez MA. Over-

expression of dihydrofolate reductase is a factor of poor survival

in acute lymphoblastic leukemia. Oncol Lett. 2018;15:8405–11.

93. Raimondi MV, Randazzo O, La Franca M, Barone G, Vignoni E,

Rossi D, et al. DHFR inhibitors: reading the past for discovering

novel anticancer agents. Molecules. 2019;24:1140.

94. Raimondi MV, Randazzo O, La Franca M, Barone G, Vignoni E,

Rossi D. et al. DHFR Inhibitors: Reading the Past for Dis-

covering Novel Anticancer Agents. Molecules. 2019;24:1140

95. Schweitzer BI, Dicker AP, Bertino JR. Dihydrofolate reductase

as a therapeutic target. The FASEB J. 1990;4:2441–52.

96. Patel SN, Kain KC. Atovaquone/proguanil for the prophylaxis

and treatment of malaria. Expert Rev Anti Infect Ther.

2005;3:849–61.

97. Wróbel A, Arciszewska K, Maliszewski D, Drozdowska D.

Trimethoprim and other nonclassical antifolates an excellent

template for searching modiﬁcations of dihydrofolate reductase

enzyme inhibitors. J Antibiotics. 2020;73:5–27.

L. N. Zhao et al.

98. Srinivasan B, Skolnick J. Insights into the slow-onset tight-

binding inhibition of Escherichia coli dihydrofolate reductase:

detailed mechanistic characterization of pyrrolo [3,2-f] quina-

zoline-1,3-diamine and its derivatives as novel tight-binding

inhibitors. FEBS J. 2015;282:1922–38.

99. Rajagopalan PTR, Zhang Z, McCourt L, Dwyer M, Benkovic SJ,

Hammes GG. Interaction of dihydrofolate reductase with meth-

otrexate: ensemble and single-molecule kinetics. Proc Natl Acad

Sci USA. 2002;99:13481–6.

100. Kremer JM. Toward a better understanding of methotrexate.

Arthritis Rheum. 2004;50:1370–82.

101. Yuthavong Y, Tarnchompoo B, Vilaivan T, Chitnumsub P,

Kamchonwongpaisan S, Charman SA, et al. Malarial dihy-

drofolate reductase as a paradigm for drug development against a

resistance-compromised target. PNAS. 2012;109:16823–8.

102. Agarwal PK, Billeter SR, Rajagopalan PTR, Benkovic SJ,

Hammes-Schiffer S. Network of coupled promoting motions in

enzyme catalysis. PNAS. 2002;99:2794–9.

103. Cario H, Smith DEC, Blom H, Blau N, Bode H, Holzmann K,

et al. Dihydrofolate reductase deﬁciency due to a homozygous

DHFR mutation causes megaloblastic anemia and cerebral folate

deﬁciency leading to severe neurologic disease. Am J Hum Genet.

2011;88:226–31.

104. Kotoula V, Krikelis D, Karavasilis V, Koletsa T, Eleftheraki AG,

Televantou D, et al. Expression of DNA repair and replication

genes in non-small cell lung cancer (NSCLC): a role for thy-

midylate synthetase (TYMS). BMC Cancer. 2012;12:342.

105. Burdelski C, Strauss C, Tsourlakis MC, Kluth M, Hube-Magg

C, Melling N, et al. Overexpression of thymidylate synthase

(TYMS) is associated with aggressive tumor features and early

PSA recurrence in prostate cancer. Oncotarget. 2015;6:

8377–87.

106. Calvert AH, Alison DL, Harland SJ, Robinson BA, Jackman AL,

Jones TR, et al. A phase I evaluation of the quinazoline antifolate

thymidylate synthase inhibitor, N10-propargyl-5,8-dideazafolic

acid, CB3717. J Clin Oncol. 1986;4:1245–52.

107. Webber S, Bartlett CA, Boritzki TJ, Hillard JA, Howland EF,

Johnston AL, et al. AG337, a novel lipophilic thymidylate syn-

thase inhibitor: in vitro and in vivo preclinical studies. Cancer

Chemother Pharm. 1996;37:509–17.

108. Duch DS, Banks S, Dev IK, Dickerson SH, Ferone R, Heath LS,

et al. Biochemical and cellular pharmacology of 1843U89, a

novel benzoquinazoline inhibitor of thymidylate synthase. Can-

cer Res. 1993;53:810–8.

109. Chu E, Callender MA, Farrell MP, Schmitz JC. Thymidylate

synthase inhibitors as anticancer agents: from bench to bedside.

Cancer Chemother Pharm. 2003;52:80–9.

110. Peters GJ, Backus HHJ, Freemantle S, van Triest B, Codacci-

Pisanelli G, van der Wilt CL, et al. Induction of thymidylate

synthase as a 5-ﬂuorouracil resistance mechanism. Biochim

Biophys Acta. 2002;1587:194–205.

111. Chen D, Jansson A, Sim D, Larsson A, Nordlund P. Structural

analyses of human thymidylate synthase reveal a site that may

control conformational switching between active and inactive

states. J Biol Chem. 2017;292:13449–58.

112. Lovelace LL, Johnson SR, Gibson LM, Bell BJ, Berger SH,

Lebioda L. Variants of human thymidylate synthase with loop

181-197 stabilized in the inactive conformation. Protein Sci.

2009;18:1628–36.

113. Huang X, Gibson LM, Bell BJ, Lovelace LL, Peña MMO,

Berger FG, et al. Replacement of Val3 in human thymidylate

synthase affects its kinetic properties and intracellular stability.

Biochemistry. 2010;49:2475–82.

114. Peters EJ, Kraja AT, Lin SJ, Yen-Revollo JL, Marsh S, Pro-

vince MA, et al. Association of thymidylate synthase variants

with 5-ﬂuorouracil cytotoxicity. Pharmacogenet Genom.

2009;19:399–401.

115. Brodsky G, Barnes T, Bleskan J, Becker L, Cox M, Patterson D.

The human GARS-AIRS-GART gene encodes two proteins

which are differentially expressed during human brain develop-

ment and temporally overexpressed in cerebellum of individuals

with Down syndrome. Hum Mol Genet. 1997;6:2043–50.

116. Deis SM, Doshi A, Hou Z, Matherly LH, Gangjee A, Dann CE.

Structural and enzymatic analysis of tumor-targeted antifolates

that inhibit glycinamide ribonucleotide formyltransferase. Bio-

chemistry. 2016;55:4574–82.

117. Bronder JL, Moran RG. Antifolates targeting purine synthesis

allow entry of tumor cells into S phase regardless of p53 func-

tion. Cancer Res. 2002;62:5236–41.

118. Robert F, Garrett C, Dinwoodie WR, Sullivan DM, Bishop M,

Amantea M, et al. Results of 2 phase I studies of intravenous (iv)

pelitrexol (AG2037), a glycinamide ribonucleotide for-

myltransferase (GARFT) inhibitor, in patients (pts) with solid

tumors. J Clin Onocol. 2004;22:3075.

119. Shih C, Chen VJ, Gossett LS, Gates SB, MacKellar WC, Habeck

LL. et al. LY231514, a pyrrolo[2,3-d]pyrimidine-based antifolate

that inhibits multiple folate-requiring enzymes. Cancer Res.

1997;57:1116–23.

120. Welin M, Grossmann JG, Flodin S, Nyman T, Stenmark P,

Trésaugues L, et al. Structural studies of tri-functional human

GART. Nucleic Acids Res. 2010;38:7308–19.

121. Burda P, Kuster A, Hjalmarson O, Suormala T, Bürer C, Lutz S,

et al. Characterization and review of MTHFD1 deﬁciency: four

new patients, cellular delineation and response to folic and

folinic acid treatment. J Inherit Metab Dis. 2015;38:863–72.

122. Sdelci S, Rendeiro AF, Rathert P, You W, Lin J-MG, Ringler A,

et al. MTHFD1 interaction with BRD4 links folate metabolism to

transcriptional regulation. Nat Genet. 2019;51:990–8.

123. New Antifolates in Clinical Development. Cancer Network.

https://www.cancernetwork.com/view/new-antifolates-clinical-

development.

124. Pikman Y, Puissant A, Alexe G, Furman A, Chen LM, Frumm

SM, et al. Targeting MTHFD2 in acute myeloid leukemia. J Exp

Med. 2016;213:1285–306.

125. Zeller KI, Jegga AG, Aronow BJ, O’Donnell KA, Dang CV. An

integrated database of genes responsive to the Myc oncogenic

transcription factor: identiﬁcation of direct genomic targets.

Genome Biol. 2003;4:R69.

126. Vazquez A, Markert EK, Oltvai ZN. Serine biosynthesis with

one carbon catabolism and the glycine cleavage system

represents a novel pathway for ATP generation. PLoS ONE.

2011;6:e25881.

127. Dang CV, Reddy EP, Shokat KM, Soucek L. Drugging the

‘undruggable’cancer targets. Nat Rev Cancer. 2017;17:502–8.

128. Sneader W. Drug discovery: a history. UK: John Wiley & Sons;

2005. https://books.google.se/books?id=jglFsz5EJR8C&printsec=

frontcover&source=gbs_ge_summary_r&cad=0#v=onepa

ge&q&f=false.

129. Wang W, Zhou H, Liu L. Side effects of methotrexate therapy

for rheumatoid arthritis: a systematic review. Eur J Med Chem.

2018;158:502–16.

130. Zachariae H. Methotrexate side-effects. Br J Dermatol.

1990;122:127–33.

131. Albrecht K, Müller-Ladner U. Side effects and management of

side effects of methotrexate in rheumatoid arthritis. Clin Exp

Rheumatol. 2010;28:S95–101.

132. Chan ESL, Cronstein BN. Methotrexate—how does it really

work? Nat Rev Rheumatol. 2010;6:175–8.

133. Bertino JR, Göker E, Gorlick R, Li WW, Banerjee D. Resistance

mechanisms to methotrexate in tumors. Stem Cells. 1996;14:5–9.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

134. Nagura E. [Methotrexate and its drug resistance]. Gan Kagaku

Ryoho. 1988;15:2882–7.

135. Wang J, Li G. Mechanisms of methotrexate resistance in

osteosarcoma cell lines and strategies for overcoming this

resistance. Oncol Lett. 2015;9:940–4.

136. Wojtuszkiewicz A, Peters GJ, van Woerden NL, Dubbelman B,

Escherich G, Schmiegelow K, et al. Methotrexate resistance in

relation to treatment outcome in childhood acute lymphoblastic

leukemia. J Hematol Oncol. 2015;8:61.

137. Frantz M-C, Wipf P. Mitochondria as a target in treatment.

Environ Mol Mutagen. 2010;51:462–75.

138. Medwig-Kinney TN, Smith JJ, Palmisano NJ, Tank S, Zhang W,

Matus DQ. A developmental gene regulatory network for C.

elegans anchor cell invasion. Development. 2020;147. https://

doi.org/10.1242/dev.185850.

139. Ducker GS, Chen L, Morscher RJ, Ghergurovich JM, Esposito

M, Teng X, et al. Reversal of cytosolic one-carbon ﬂux com-

pensates for loss of the mitochondrial folate pathway. Cell

Metab. 2016;23:1140–53.

140. Toyama BH, Hetzer MW. Protein homeostasis: live long, won’t

prosper. Nat Rev Mol Cell Biol. 2013;14:55–61.

141. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P.

From RNA to protein. In: Molecular biology of the cell. 4th ed.

New York: Garland Science; 2002. https://www.ncbi.nlm.nih.

gov/books/NBK26829/.

L. N. Zhao et al.

MTHFD2 in healthy and cancer cells: Canonical and non-canonical functions

Article

Full-text available

Mar 2024

Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) is a mitochondrial enzyme of the folate-mediated one-carbon metabolism pathway. MTHFD2 has become a highly attractive therapeutic target due to its consistent upregulation in cancer tissues and its major contribution to tumor progression, although it also performs vital functions in proliferating healthy cells. Here, we review the diversity of canonical and non-canonical functions of this key metabolic enzyme under physiological conditions and in carcinogenesis. We provide an overview of its therapeutic potential and describe its regulatory mechanisms. In addition, we discuss the recently described non-canonical functions of MTHFD2 and the mechanistic basis of its oncogenic function. Finally, we speculate on novel therapeutic approaches that take into account subcellular compartmentalization and outline new research directions that would contribute to a better understanding of the fundamental roles of this metabolic enzyme in health and disease.

Targeting one‐carbon metabolism for cancer immunotherapy

Article

Full-text available

Jan 2024

Background One‐carbon (1C) metabolism is a metabolic network that plays essential roles in biological reactions. In 1C metabolism, a series of nutrients are used to fuel metabolic pathways, including nucleotide metabolism, amino acid metabolism, cellular redox defence and epigenetic maintenance. At present, 1C metabolism is considered the hallmark of cancer. The 1C units obtained from the metabolic pathways increase the proliferation rate of cancer cells. In addition, anticancer drugs, such as methotrexate, which target 1C metabolism, have long been used in the clinic. In terms of immunotherapy, 1C metabolism has been used to explore biomarkers connected with immunotherapy response and immune‐related adverse events in patients. Methods We collected numerous literatures to explain the roles of one‐carbon metabolism in cancer immunotherapy. Results In this review, we focus on the important pathways in 1C metabolism and the function of 1C metabolism enzymes in cancer immunotherapy. Then, we summarise the inhibitors acting on 1C metabolism and their potential application on cancer immunotherapy. Finally, we provide a viewpoint and conclusion regarding the opportunities and challenges of targeting 1C metabolism for cancer immunotherapy in clinical practicability in the future. Conclusion Targeting one‐carbon metabolism is useful for cancer immunotherapy.

Identification and Absorption–Distribution–Metabolism–Excretion–Toxicity Prediction of Potential MTHFD2 Enzyme Inhibitors from Urtica dioica Ethanolic Leaf Extract

Article

Full-text available

Jun 2024

Shifaa O. Alshammari

This study aimed to explore the potential of Urtica dioica (U. dioica) ethanolic leaf extract for cancer treatment by identifying its components, evaluating its effects on cancer cell lines, and analyzing its molecular docking. The objective of this study was to investigate the anticancer properties of U. dioica ethanolic leaf extract and assess its potential as a therapeutic strategy for cancer treatment. This study utilized high-performance liquid chromatography (HPLC) to analyze the chemical composition of U. dioica ethanolic leaf extract. The anticancer effects of the extract were evaluated by assessing cell viability, determining IC50 values, and conducting ADMET analysis after oral administration. U. dioica ethanolic leaf extract was found to contain methyl hexadecanoate as its primary component, along with flavonoids and polyphenols. It effectively reduced cell viability in various tested cancer cell lines, with IC50 values varying for each cell line. The duration of treatment significantly influenced cell viability, with the most significant reduction observed after 48 h. Molecular docking studies suggested that catechin, kaempferol, and quercetin-3-O-rutinoside may have potential as inhibitors of the MTHFD2 enzyme. This study revealed the potential of U. dioica and its compounds in cancer treatment. Ethanolic leaf extract has been shown to have anticancer effects on various cancer cell lines, with catechin and kaempferol showing promise as inhibitors of the MTHFD2 enzyme. Further research is warranted to explore the therapeutic implications of U. dioica in cancer treatment.

MTHFD2-mediated redox homeostasis promotes gastric cancer progression under hypoxic conditions

Article

Full-text available

May 2024

Objectives Cancer cells undergo metabolic reprogramming to adapt to high oxidative stress, but little is known about how metabolic remodeling enables gastric cancer cells to survive stress associated with aberrant reactive oxygen species (ROS) production. Here, we aimed to identify the key metabolic enzymes that protect gastric cancer (GC) cells from oxidative stress. Methods ROS level was detected by DCFH-DA probes. Multiple cell biological studies were performed to identify the underlying mechanisms. Furthermore, cell-based xenograft and patient-derived xenograft (PDX) model were performed to evaluate the role of MTHFD2 in vivo. Results We found that overexpression of MTHFD2, but not MTHFD1, is associated with reduced overall and disease-free survival in gastric cancer. In addition, MTHFD2 knockdown reduces the cellular NADPH/NADP+ ratio, colony formation and mitochondrial function, increases cellular ROS and cleaved PARP levels and induces in cell death under hypoxia, a hallmark of solid cancers and a common inducer of oxidative stress. Moreover, genetic or pharmacological inhibition of MTHFD2 reduces tumor burden in both tumor cell lines and patient-derived xenograft-based models. Discussion our study highlights the crucial role of MTHFD2 in redox regulation and tumor progression, demonstrating the therapeutic potential of targeting MTHFD2.

SHMT2 promotes papillary thyroid cancer metastasis through epigenetic activation of AKT signaling

Article

Full-text available

Jan 2024

Cancer cells alter their metabolism and epigenetics to support cancer progression. However, very few modulators connecting metabolism and epigenetics have been uncovered. Here, we reveal that serine hydroxymethyltransferase-2 (SHMT2) generates S-adenosylmethionine (SAM) to epigenetically repress phosphatase and tensin homolog (PTEN), leading to papillary thyroid cancer (PTC) metastasis depending on activation of AKT signaling. SHMT2 is elevated in PTC, and is associated with poor prognosis. Overexpressed SHMT2 promotes PTC metastasis both in vitro and in vivo. Proteomic enrichment analysis shows that AKT signaling is activated, and is positively associated with SHMT2 in PTC specimens. Blocking AKT activation eliminates the effects of SHMT2 on promoting PTC metastasis. Furthermore, SHMT2 expression is negatively associated with PTEN, a negative AKT regulator, in PTC specimens. Mechanistically, SHMT2 catalyzes serine metabolism and produces activated one-carbon units that can generate SAM for the methylation of CpG islands in PTEN promoter for PTEN suppression and following AKT activation. Importantly, interference with PTEN expression affects SHMT2 function by promoting AKT signaling activation and PTC metastasis. Collectively, our research demonstrates that SHMT2 connects metabolic reprogramming and epigenetics, contributing to the poor progression of PTC.

One-carbon metabolism shapes T cell immunity in cancer

Article

Jun 2024
TRENDS ENDOCRIN MET

Machine Learning-Based Integrated Multiomics Characterization of Colorectal Cancer Reveals Distinctive Metabolic Signatures

Article

May 2024

Targeting SHMTs and MTHFDs in cancer: attractive opportunity for anti-tumor strategy

Article

Full-text available

Feb 2024

One-carbon metabolism is a universal metabolic process that mediates the transfer of one-carbon units for purine and thymidine synthesis. One-carbon metabolism has been found to be dysregulated in various cancer types due to its role in production of purine and pyrimidine nucleotides, epigenetic program, and redox homeostasis. One-carbon metabolism is composed a network of one-carbon metabolic enzymes. Disturbing the expression and enzymatic activity of these one-carbon metabolic enzymes could lead to fluctuations of metabolites in the tumor microenvironment. Serine hydroxymethyltransferases (SHMTs) and methylenetetrahydrofolate dehydrogenases (MTHFDs) are gradually recognized as important one-carbon metabolic enzymes for regulating tumor initiation and development, representing potential therapeutic targets for anti-tumor strategies. In the review, we primarily focused on the role of SHMTs and MTHFDs in cancer. Several inhibitors targeting MTHFDs and SHMTs have exert its potential to decrease tumor burden and inhibit tumor proliferation, highlighting the potential of targeting one-carbon metabolic enzymes for anti-cancer strategies.

Inhibition of purine metabolism promotes the differentiation of the neuroblastoma driven by MYCN

Preprint

Full-text available

Feb 2024

Background: Neuroblastoma is the common extracranial solid tumor in children, with a poor prognosis for high-risk patients. MYCN amplification is the most important molecular biomarker in the high-risk category. For MYCN’s “undruggable” properties--lack of enzymatic pocket for conventional small molecules to bind and inaccessibility for antibodies due to the predominant nucleus localization of MYCN, current therapeutic strategies have not significantly improved. It is known that MYCN can affect a variety of tumor metabolism and is closely related to tumor differentiation. The objective of this study is to investigate the impact of purine metabolism on the maintenance of stemness in MYCN high-risk neuroblastoma and identify potential small molecule inhibitors that can enhance the differentiation of neuroblastoma cells. Methods: Metabolic mass spectrometry data were used to analyze the differences in metabolites between neuroblastoma cell lines with high and low MYCN, and key metabolic enzymes were analyzed in conjunction with public RNA databases. Different neuroblastoma cell lines were treated with the small molecule inhibitor for cell differentiation, proliferation, colony formation, and cell migration analysis, to find inhibitors that are promotive for cell differentiation and validate them in vivo and in vitro. Results: In our study, we found significant changes in nucleotide metabolism in NB cell lines with high MYCN compared to those with low MYCN. The expression of purine metabolic enzymes was positively correlated with MYCN expression, prognosis, and differentiation status in neuroblastoma. The small molecular lometrexol, a phosphoribosylglycinamide formyltransferase (GART) inhibitor, which blocks the de novo pathway of purine metabolism, can promote a pro-differentiation effect on neuroblastoma cells through in vivo and in vitro experiments and inhibit tumorigenesis. Discussion: Our study suggested that lometrexol, the small molecule inhibitor of nucleotide metabolism, may contribute to improving treatment strategies for pediatric neuroblastoma, enhancing patient prognosis, and improving overall quality of life.

Structural Characterization of 5-Substituted Pyrrolo[3,2-d]pyrimidine Antifolate Inhibitors in Complex with Human Serine Hydroxymethyl Transferase 2

Article

Feb 2024
BIOCHEMISTRY-US

Evaluating the clinical significance of SHMT2 and its co-expressed gene in human kidney cancer

Article

Full-text available

Dec 2020
BIOL RES

Background: Kidney cancer is one of the most common cancers in the world. It is necessary to clarify its underlying mechanism and find its prognostic biomarkers. Current studies showed that SHMT2 may be participated in several kinds of cancer. Methods: Our studies investigated the expression of SHMT2 in kidney cancer by Oncomine, Human Protein Atlas database and ULCAN database. Meanwhile, we found its co-expression gene by cBioPortal online tool and validated their relationship in A498 and ACHN cells by cell transfection, western blot and qRT-PCR. Besides these, we also explored their prognostic values via the Kaplan-Meier plotter database in different types of kidney cancer patients. Results: SHMT2 was found to be increased in 7 kidney cancer datasets, compared to normal renal tissues. For the cancer stages, ages and races, there existed significant difference in the expression of SHMT2 among different groups by mining of the UALCAN database. High SHMT2 expression is associated with poor overall survival in patients with kidney cancer. Among all co-expressed genes, NDUFA4L2 and SHMT2 had a high co-expression efficient. SHMT2 overexpression led to the increased expression of NDUFA4L2 at both mRNA and protein levels. Like SHMT2, overexpressed NDUFA4L2 also was associated with worse overall survival in patients with kidney cancer. Conclusion: Based on above results, overexpressed SHMT2 and its co-expressed gene NDUFA4L2 were all correlated with the prognosis in kidney cancer. The present study might be benefit for better understanding the clinical significance of SHMT2 and provided a potential therapeutic target for kidney cancer in future.

MicroRNA-940 inhibits glioma progression by blocking mitochondrial folate metabolism through targeting of MTHFD2

Article

Full-text available

Feb 2019

The aggressiveness and recurrence of glioma are major obstacles for the treatment of this type of tumor. Further understanding of the molecular mechanisms of glioma is necessary to improve the efficacy of therapy. MicroRNAs have been widely studied in many human cancers. Here, we found that miR-940 was one of the primary downregulated miRNAs in clinical samples and glioma cell lines through bioinformatics analysis and qRT-PCR. Upregulating miR-940 expression significantly inhibited the proliferation and invasion and promoted apoptosis of U87 and U118 cells. In addition, experiments in vivo showed that upregulation of miR-940 expression inhibited xenograft growth. Methylenetetrahydrofolate dehydrogenase (MTHFD2), a dual-functional metabolic enzyme, is involved in the one-carbon metabolism of folate in mitochondria. We found MTHFD2 to be overexpressed in glioma tissues and our clinical samples by qRT-PCR and Western blot assays. Through TargetScan prediction and luciferase assays, we found that miR-940 directly targets MTHFD2. Upregulation of miR-940 expression inhibited the expression of MTHFD2 and led to intracellular one-carbon metabolism dysfunction. Furthermore, the antitumor effects of miR-940 could be attenuated by overexpression of MTHFD2. Together, the results of our study suggest that miR-940 may be a new therapeutic target for the treatment of glioma through targeting of MTHFD2.

Glycolytic suppression dramatically changes the intracellular metabolic profile of multiple cancer cell lines in a mitochondrial metabolism-dependent manner

Article

Full-text available

Dec 2019

Most cancer cells rely on glycolysis to generate ATP, even when oxygen is available. However, merely inhibiting the glycolysis is insufficient for the eradication of cancer cells. One main reason for this is that cancer cells have the potential to adapt their metabolism to their environmental conditions. In this study, we investigated how cancer cells modify their intracellular metabolism when glycolysis is suppressed, using PANC-1 pancreatic cancer cells and two other solid tumor cell lines, A549 and HeLa. Our study revealed that glycolytically suppressed cells upregulated mitochondrial function and relied on oxidative phosphorylation (OXPHOS) to obtain the ATP necessary for their survival. Dynamic changes in intracellular metabolic profiles were also observed, reflected by the reduced levels of TCA cycle intermediates and elevated levels of most amino acids. Glutamine and glutamate were important for this metabolic reprogramming, as these were largely consumed by influx into the TCA cycle when the glycolytic pathway was suppressed. During the reprogramming process, activated autophagy was involved in modulating mitochondrial function. We conclude that upon glycolytic suppression in multiple types of tumor cells, intracellular energy metabolism is reprogrammed toward mitochondrial OXPHOS in an autophagy-dependent manner to ensure cellular survival.

A developmental gene regulatory network for C. elegans anchor cell invasion

Article

Full-text available

Dec 2019

Cellular invasion is a key part of development, immunity, and disease. Using the in vivo model of C. elegans anchor cell invasion, we characterize the gene regulatory network that promotes cell invasion. The anchor cell is initially specified in a stochastic cell fate decision mediated by Notch signaling. Previous research has identified four conserved transcription factors, fos-1a (Fos), egl-43 (EVI1/MEL), hlh-2 (E/Daughterless) and nhr-67 (NR2E1/TLX), that mediate anchor cell specification and/or invasive behavior. Connections between these transcription factors and the underlying cell biology that they regulate are poorly understood. Here, using genome editing and RNA interference, we examine transcription factor interactions before and after anchor cell specification. Initially, these transcription factors function independently of one another to regulate LIN-12 (Notch) activity. Following anchor cell specification, egl-43, hlh-2, and nhr-67, function largely parallel to fos-1 in a type I coherent feed-forward loop with positive feedback to promote invasion. Together, these results demonstrate that the same transcription factors can function in cell fate specification and differentiated cell behavior, and that a gene regulatory network can be rapidly assembled to reinforce a post-mitotic, pro-invasive state.

Down‐regulation of MTHFD2 inhibits NSCLC progression by suppressing cycle‐related genes

Article

Full-text available

Nov 2019
J CELL MOL MED

Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) is a bifunctional enzyme located in the mitochondria. It has been reported to be overexpressed in several malignancies. However, the relationship between the expression of MTHFD2 and non‐small cell lung cancer (NSCLC) remains largely unknown. In this study, we found that MTHFD2 was significantly overexpressed in NSCLC tissues and cell lines. Knockdown of MTHFD2 resulted in reduced cell growth and tumorigenicity in vitro and in vivo. Besides, the mRNA and protein expression level of cell cycle genes, such as CCNA2, MCM7 and SKP2, was decreased in MTHFD2 knockdown H1299 cells. Our results indicate that the inhibitory effect of MTHFD2 knockdown on NSCLC may be mediated via suppressing cell cycle‐related genes. These findings delineate the role of MTHFD2 in the development of NSCLC and may have potential applications in the treatment of NSCLC.

MTHFD1L, A Folate Cycle Enzyme, Is Involved in Progression of Colorectal Cancer

Article

Full-text available

Nov 2019

Identification of new molecular targets is needed for the treatment of colorectal cancer (CRC). Methylenetetrahydrofolate dehydrogenase 1 like (MTHFD1L), an enzyme in the folate cycle, is involved in formate generation and therefore in one-carbon metabolism. Here, we examined the expression and the role of MTHFD1L in CRC progression. Bioinformatics analysis of several public databases showed overexpression of MTHFD1L in CRC tissues as compared to normal tissues. Quantitative real-time PCR and Western blotting revealed that expressions of MTHFD1L RNA and protein were higher in CRC tissues compared to their corresponding normal tissues of CRC patients. Immunohistochemical staining demonstrated higher cytoplasmic MTHFD1L reactivity in tumor tissues compared to paired normal tissues. Further, to determine the functional relevance of MTHFD1L, it was knocked down by an siRNA in CRC cells. Silencing of MTHFD1L inhibited CRC cell proliferation, colony formation, invasion, and migration. Thus, to our knowledge for the first time in the literature, we show that MTHFD1L is involved in CRC progression and that blocking of MTHFD1L decreases the growth of colon cancer cells, thus providing an avenue to target this enzyme with small molecule inhibitors.

Effect of Low pH Treatment on Cell Cycle and Cell Growth

Article

Apr 2018

Cancer growth affects the pH of microenvironment that becomes acidified and supports local invasion. Studies have reported that pH is a factor in cancer growth, division, and spread. Although cancer cells produce low extracellular pH (6.5–6.9), they are able to maintain their intracellular pH at favorable ranges (7.2–7.4). Our preliminary results showed that growth of cancer cells changes after low pH treatment. The objective of our research is to study how low pH affects cell morphology and behavior. In my preliminary experiments, HeLa cells were incubated with low pH solution for 10 min and harvested using trypsin into medium with normal pH. After 24 hours in culture, cell number was assessed using Trypan Blue Exclusion Assay. As a result, low pH treated group doubled 1.4 times; whereas control group doubled 2.3 times, suggesting a division latency in low pH treated group on Day 1. Cell morphology on Day 1 was recorded as images and was categorized into round, spindle, triangle, and irregular groups, among which spindle cells were likely to be at G1 phase. Percentage of the spindle cells gradually increased during 16–24 hours in experimental group; however, in control group, percentage of spindle cells decreased, suggesting that low pH had induced cell cycle arrest at G1 phase. In conclusion, low pH treatment affects cell growth potentially through G1 cell cycle arrest. Future work is to confirm cell cycle arrest at G1 phase after low pH treatment and to study the mechanism of low pH effect on cancer cell growth. Support or Funding Information Funding is provided partially by NIH (#NS081629) to YVL. Thanks to the Osteopathic Heritage Foundations, Graduate Assistantship Program at Ohio University Heritage College of Osteopathic Medicine for supporting graduate student YH. This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .

Drug Discovery: A History

Book

Apr 2005

Walter Sneader

Metabolic Fitness and Plasticity in Cancer Progression

Article

Jan 2020

Cancer cells have enhanced metabolic needs due to their rapid proliferation. Moreover, throughout their progression from tumor precursors to metastases, cancer cells face challenging physiological conditions, including hypoxia, low nutrient availability, and exposure to therapeutic drugs. The ability of cancer cells to tailor their metabolic activities to support their energy demand and biosynthetic needs throughout disease progression is key for their survival. Here, we review the metabolic adaptations of cancer cells, from primary tumors to therapy resistant cancers, and the mechanisms underpinning their metabolic plasticity. We also discuss the metabolic coupling that can develop between tumors and the tumor microenvironment. Finally, we consider potential metabolic interventions that could be used in combination with standard therapeutic approaches to improve clinical outcome.

High expression of SHMT2 is correlated with tumor progression and predicts poor prognosis in gastrointestinal tumors

Article

Nov 2019

Objective: Gastrointestinal tumors are malignant tumors with high morbidity. Mitochondrial serine hydroxymethyltransferase 2 (SHMT2) is a key enzyme in the synthesis of serine and glycine, which has prognostic and therapeutic value for many malignant tumors. However, the role of SHMT2 in gastric cancer (GC), esophageal cancer (ESCC), and colorectal cancer (CC) has not been clarified. Patients and methods: The expression of SHMT2 was detected in GC, ESCC, and CC by immunohistochemistry and reverse real time transcription-polymerase chain reaction. The relationships between SHMT2 expression and clinicopathologic characteristics, recurrence-free survival (RFS), and disease-specific survival (DSS) were analyzed by the survival analysis and correlation analysis. Results: The positive expression rate of SHMT2 in GC, ESCC, and CC was 74.1%, 69.2%, and 71.7%, respectively. Patients with high expression of SHMT2 had a worse prognosis. In GC, high SHMT2 expression had positive correlation with lymph node metastasis (p=0.005) and histological grade (p=0.002). In ESCC, high SHMT2 expression had positive correlation with pT classification (p=0.033) and pM classification (p=0.029). In CC, high SHMT2 expression had positive correlation with tumor size (p=0.004), lymph node metastasis (p=0.035), TNM stage (p=0.007), and histological grade (p=0.020). Notably, SHMT2 expression was an independent prognostic factor for RFS and DSS in GC, ESCC, and CC (p<0.05). Conclusions: SHMT2 is upregulated in GC, ESCC, and CC. The high expression of SHMT2 is correlated with gastrointestinal tumors progression, and poor prognosis, which is a potential new target for the diagnosis and treatment of gastrointestinal tumors.

Therapeutic targeting of the mitochondrial one-carbon pathway: perspectives, pitfalls, and potential

Abstract and Figures

Recommended publications

More Than a Metabolic Enzyme: MTHFD2 as a Novel Target for Anticancer Therapy?

ATF4/MYC Regulates MTHFD2 to Promote NSCLC Progression by Mediating Redox Homeostasis

Targeting MTHFD2 to Exploit Cancer-Specific Metabolism and the DNA Damage Response

Therapeutic Targeting of Mitochondrial One-Carbon Metabolism in Cancer