Taylor & Francis Taylor & Francis Group

655/ 0368-2532

DRUG METABOLISM REVIEWS

informa

The CYP2W1 enzyme: regulation, properties and activation of prodrugs

Jia Guo, Inger Johansson, Souren Mkrtchian & Magnus Ingelman-Sundberg

To cite this article: Jia Guo, Inger Johansson, Souren Mkrtchian & Magnus Ingelman-Sundberg (2016): The CYP2W1 enzyme: regulation, properties and activation of prodrugs, Drug Metabolism Reviews, DOI: 10.1080/03602532.2016.1188939

To link to this article: http://dx.doi.org/10.1080/03602532.2016.1188939

Published online: 03 Jun 2016.

Submit your article to this journal

Q

View related articles

View Crossmark data

CrossMark

Taylor & Francis Taylor & Francis Group

REVIEW ARTICLE

The CYP2W1 enzyme: regulation, properties and activation of prodrugs

Jia Guo, Inger Johansson, Souren Mkrtchian and Magnus Ingelman-Sundberg Section of Pharmacogenetics, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden

ABSTRACT

CYP2W1 is expressed in the course of development of the gastrointestinal tract, silenced after birth in intestine and colon by epigenetic modifications, but activated following demethylation in colorectal cancer (CRC). The expression levels in CRC positively correlate with the degree of malig- nancy, are higher in metastases and are predictive of colon cancer survival. The CYP2W1 tran- scripts have been detected also in hepatocellular carcinoma, adrenocortical carcinoma, childhood rhabdomyosarcoma and breast cancer; however, here the protein expression remains to be con- firmed. The CYP2W1 enzyme has an inverted orientation in the endoplasmic reticulum membrane, as compared to other cytochrome P450s and its immediate electron donor is unknown. Several lipid ligands have been proposed as endogenous substrates, among which retinol derivatives appear to have the highest affinities. However, the role of CYP2W1 in the endogenous and tumor localized metabolism of retinol derivatives has yet to be clarified. Indolines constitute high affinity exogenous compounds and specific chloromethylindolines have been shown to be acti- vated by CYP2W1 into cytotoxic products in vitro and also in vivo, inhibiting the growth of human colon tumors in a mouse xenograft model. The CRC specific localization of CYP2W1 and its effect- ive prodrug activation makes it a very promising target for future development of cancer therapeutics.

ARTICLE HISTORY

Received 18 April 2016 Revised 9 May 2016 Accepted 9 May 2016 Published online 3 June 2016

KEYWORDS

Indolines; colon cancer; prodrugs; duocarmycins; fetal colon

Introduction

Following the sequence identification of the human genome in 2000, many new P450 genes were identified and cloned. Previous P450s had been cloned to a great extent after purification and deciphering the primary structure of the corresponding enzyme; however, the genome information allowed cloning based on the sequence homology and many new orphan P450s in family 2 were then cloned and partially characterized including CYP2S1, CYP2R1, CYP2U1 and CYP2W1. In the present review, we summarize the history, distribution, polymorphism, membrane topology, function and pos- sible role in cancer therapy of CYP2W1 including latest data that appeared since the latest CYP2W1 review (Fei-Lei Chung et al., 2015).

Initial findings

A partial sequence for CYP2W1 was first detected in the year of 2000 in a cDNA library from the hepatoma cell line HepG2 (GenBank Accession No. AK000366.1). Using this sequence a Celera sequence database of human genome was screened and the full CYP2W1 gene sequence was identified (Karlgren et al., 2006). The

enzymes with the highest identity to CYP2W1 were CYP2D6 (42%) and CYP2S1 (40%). The gene was found to be relatively conserved and CYP2W1 orthologs with 78% and 66% amino acid identity were found in the mouse and in the rat genomes (GenBank Accession Nos. XM_144624 and XM_221971, respectively). Subsequently, the CYP2W1 cDNA was cloned from HepG2 cells (Karlgren et al., 2006). The mRNA distribu- tion was examined using different mRNA panels. An intensive CYP2W1 mRNA signal from HepG2 cells was seen but not from human brain, kidneys, liver, heart, stomach, intestine, colon, muscle, blood, placenta, blad- der, uterus, adrenals, thyroid or lung tissues. By contrast, a similar screening performed using 29 tumors of differ- ent origins, revealed high expression in several colon tumors and tumors from adrenal glands. An antiserum raised in our laboratory against the C-terminal sequence of the protein (852 Ab) could not detect any CYP2W1 protein in human liver, whereas significant CYP2W1 pro- tein expression was only found in colon tumors. Similar analysis in the rat adult and embryonic tissues detected mRNA expression in the fetal colon. Transcripts of the mouse ortholog were also found in fetal mouse tissues (Choudhary et al., 2005). Based on these data, CYP2W1

was later classified as an oncofetal gene with a selective developmental and tumor expression.

Expression profile

The attempts to detect CYP2W1 expression in a variety of adult human and animal tissues were unsuccessful, only trace amounts of mRNA were found in human prostate, colon, placenta, pancreas, ovary, testis, whereas the protein was completely undetectable. However, CYP2W1 mRNA was identified in the fetal rat colon at the relatively high levels (Karlgren et al., 2006) with the expression dropping to background values shortly after the birth.

A recent study from our laboratory analyzed in detail the developmental pattern of CYP2W1 expression show- ing appearance of mRNA in the mouse embryonic colon and small intestine (SI) at the gestational day 13 (E13), peaking at E18 and rapidly decreasing in SI to 10% within first 3 days after birth and to a null level in both gastrointestinal (GI) tissues at post-natal day 28. Similar fetal expression of CYP2W1 was seen in human fetuses 18 and 19 weeks of gestation with no expression in adult colon or SI (Choong et al., 2015) (Figure 1). Using our in-house polyclonal antibodies (852 Ab), we could also detect similar CYP2W1 expression in the human embryonic intestine and colon on the protein level (Choong et al., 2015). Similar pattern of protein expres- sion was seen in the mouse embryonic GI tissues.

In contrast to the silenced expression in the normal adult tissues, CYP2W1 is found at rather high levels in the tumors of different origin. First indications of a CYP2W1 cancer specific expression was its detection in the human hepatoma cell line, HepG2 and later in the

colorectal carcinoma cell line CaCo2-TC7 (Gomez et al., 2010; Karlgren et al., 2006). Besides the colon cancer expression CYP2W1 has also been reported to be expressed at the mRNA level in the breast cancer, childhood rhabdomyosarcoma, hepatocellular carcin- oma, adrenocortical carcinoma and oral squamous cell carcinoma and at the protein level in hepatocellular car- cinoma, adenocarcinoma and childhood rhabdomyosar- coma as summarized in Table 1. However, the reported detection of CYP2W1 in tumors and in some cases even in the adult tissues is very controversial due to the use of not sufficiently specific antibodies. Our experience shows that many of the commercial CYP2W1 antibodies are unreliable. Thus, the Protein Atlas produced poly- clonal antibodies distributed through the Sigma-Aldrich (www.proteinatlas.org) show CYP2W1 expression in almost every normal human tissue, which is not sup- ported neither by mRNA nor by protein expression data using different antibodies. Furthermore, CYP2W1 protein has been claimed to be present in, for example MCF-7 cells (Tan et al., 2011) and in the normal adrenal glands (Ronchi et al., 2014); however, using 852 Ab antibody we see no protein expression (Nolé P et al., unpublished observations). Similarly, the data on the expression of CYP2W1 protein in adenocarcinoma, childhood rhabdo- myosarcoma, breast cancer should be taken with certain degree of caution and validated using different antibod- ies. Therefore, there is an urgent need for the develop- ment of CYP2W1 antibodies with substantially higher affinity and avidity.

Apparently, the only reliable method of protein detection is mass spectrometry or the immunoblotting where the band follows exactly the migration of a standard recombinant CYP2W1 spiked into a sample.

Figure 1. CYP2W1 development in human (A) and murine (B) GI tract. Data from Choong et al. (2015). Permission to reproduce from the publisher.

CYP2W1 mRNA, relative expression D

6.

(B)

3

- Colon

CYP2W1 mRNA, relative expression

4-

Small intestine

2

2-

1

0

18w

19w

19y

28y

33y

36y

39y

43y

48y

61y

0

Fetus

Adult

E13

E16

E18

D3

D7

D28

12W

7M

13M

Table 1. Expression of CYP2W1 in cancer as reported in the literature.
Cancer typesDetection methodPrognostic factorReferences
Colon cancermRNA (qPCR); protein (Western blot, 852 Ab)n.a.Karlgren et al. (2006)
Colon cancermRNA (qPCR); protein (Western blot, 852 Ab)n.a.Gomez et al. (2007)
Colon cancerProtein (IHC, 852 Ab)Yes (n= 162)Edler et al. (2009)
Colon cancerProtein (IHC, 852 Ab)Yes (n = 235)Stenstedt et al. (2012)
Colon cancer; lymph node and liver metastasesProtein (IHC, 852 Ab)n.a.Stenstedt et al. (2014)
Hepatocellular carcinomamRNA (qPCR); protein (IHC, unknown Ab)Yes (n = 133)Zhang et al. (2014)
Adrenal glands; adrenocortical carcinomamRNA (qPCR); protein (IHC, Thermo Scientific Ab; 852 Ab)No (n=196)Ronchi et al. (2014)
Childhood rhabdomyosarcomamRNA (qPCR); protein (Western blot, Santa Cruz Ab)n.a.Molina-Ortiz et al. (2014)
Breast cancermRNA (qPCR)n.a.Bandala et al. (2012)
Oral squamous cell carcinomamRNA (qPCR)n.a.Hartanto et al. (2015)

CYP2W1 as a prognostic marker of CRC

In two clinical studies using immunohistochemical stain- ing of the tumor samples from colorectal cancer (CRC) patients using the 852 Ab (162 + 235 combined), data were presented, indicating that only 7% of the tumors do not express CYP2W1, whereas 26% have a weak staining, 37% have a moderate staining and 30% have an intense staining indicating a high CYP2W1 expres- sion. The presence of CYP2W1 in the surrounding nor- mal tissues was negligible or absent (Edler et al., 2009; Stenstedt et al., 2012). Moreover, a follow-up study dem- onstrated even higher frequency of CYP2W1 expression (48%) in the liver metastases from the CRC patients (Stenstedt et al., 2014). This indicates a higher expres- sion in more malignant forms of the colon cancer cells.

The analysis of CYP2W1 expression in colon tumors revealed that CYP2W1 can serve as an independent prognostic factor in multivariate analysis (p=0.04), where high expression (grade 3) correlates with worse outcome (survival rate) (Edler et al., 2009). It was also shown that CYP2W1 expression may serve as an inde- pendent prognostic factor in the subgroup of patients with colon cancer stage III (p =0.003) but not for those with stage II (Stenstedt et al., 2012). These data thus suggest that higher expression of CYP2W1 is of prog- nostic value in colon cancer. In one relatively small study, the CYP2W1 expression has also been reported to be of prognostic value for hepatocellular carcinoma (cf. Table 1). There is, however, a need to confirm the prognostic value of CYP2W1 expression using additional and larger cohorts.

Substrate specificity

Since the discovery of its cancer-specific expression, CYP2W1 has been tested in many studies for the pos- sible tumor specific prodrug activation as well as for its putative endogenous role in normal and transformed tissues. A number of papers have been published dem- onstrating catalytic capability of CYP2W1 toward exogenous compounds. A vast majority of these studies were carried out using recombinant N-terminally

truncated CYP2W1 expressed in Escherichia coli. The tested compounds include benzphetamine, indole, benzopyrene, A4QN (Nishida et al., 2010; Wu et al., 2006; Yoshioka et al., 2006). Most of them are, however, metabolized at very low rates. Some drugs like fluoro- benzothiazoles GW 610 and 5F 203 (Phortress) have emerged as CYP2W1 substrates generating DNA alkylat- ing metabolites. 5F 203 can be also metabolized by CYP1A1, whereas GW 610 seems to be more specific for CYP2W1 when tested for cytotoxicity in selected breast and colon cancer cell lines (Wang & Guengerich, 2012; Tan et al., 2011). Indeed, the ability of GW 610 to affect the viability of colon cancer cells was diminished after knock down of the CYP2W1 protein expression as vali- dated by Western blotting (Tan et al., 2011). However, the relevance of GW 610 in CYP2W1 mediated antitumor activity remains to be verified by in vivo studies (Table 2).

The catalytic activity of CYP2W1 has also been eval- uated in transfected mammalian cells. Thus, HEK293 expressed CYP2W1 was found to metabolize indoline substrates and also activate aflatoxin B1 into cytotoxic products (Gomez et al., 2010). These data in addition to earlier reports on the role of various CYPs in the oxida- tion of indoles (Gomez et al., 2010) were instrumental in developing indoline compounds (chloromethylindoles) with significantly higher affinity to CYP2W1 and having promising prodrug properties as described below.

There have been many efforts to identify potential endogenous CYP2W1 substrates. In particular, CYP2W1 catalytic activity toward lipid molecules, such as free fatty acids and lysophospholipids, has been reported. An interesting approach was taken by using metabo- nomics to identify lipid substrates using extracts from tumors (Xiao & Guengerich, 2012). The authors demon- strated CYP2W1 dependent catalysis of 18:1 lysophos- phatidylcholine oxidation, which was found at approximately 6-fold higher rates than that of the oleic acid (C18:1). In addition, CYP2W1 was also shown to convert arachidonic acid to a number of dihydroxy derivatives albeit at a low conversion rate. In addition, four other fatty acids were also found to be

Table 2. CYP2W1 substrates reported in the literature.
Types of substratesCatalytic systemKm value (µM)References
Procarcinogens
Aflatoxin B1Reconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Wu et al. (2006)
Stably transfected HEK293 cellsn.a.Gomez et al. (2010)
Aromatic aminesReconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Eun et al. (2010)
Polycyclic aromatic hydrocarbon dihydrodiolsReconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Wu et al. (2006)
SterigmatocystinReconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Wu et al. (2006)
Anti-tumor drugs
A4QNReconstituted system with N-truncated CYP2W1 expressed in E. coli30Nishida et al. (2010)
Benzothiazole GW610KM12 cells; HCC2998 cellsn.a.Tan et al. (2011)
Reconstituted system with N-truncated CYP2W1 expressed in E. coli22 ±7Wang & Guengerich (2012)
Fluorobenzothiazole 5F203(Phortress)Reconstituted system with N-truncated CYP2W1 expressed in E. coli31 ± 11 for hydroxyl product; 35 ± 12 for hydroxylamine productWang & Guengerich (2012)
Duocarmycin analogs ICT2705, ICT2706,Stably transfected HEK293 cells Stably transfected SW480 cellsn.a.Sheldrake et al. (2013) Travica et al. (2013)
Duocarmycin analog ICT2726Transient transfected SW480 cells3.46 ± 1.17Stenstedt et al. (2013)
Putative endogenous substrates
Arachidonic acidReconstituted system with microsome from transfected HEK293 cellsn.a.Karlgren et al. (2006)
Reconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Wu et al. (2006)
Reconstituted system with N-truncated CYP2W1 expressed in E. coli64.5 ± 39.1 #KsZhao et al. (2016)
All-trans retinolReconstituted system with N-truncated CYP2W1 expressed in E. coli40.3 ±7.0 nM #KsZhao et al. (2016)
All-trans retinal55.2 ± 14.0 nM #Ks
All-trans retinoic acid5.6±0.6
17 ß-EstradiolReconstituted system with N-truncated CYP2W1 expressed in E. coli30.1 ± 11.2 #KsZhao et al. (2016)
FarnesolReconstituted system with N-truncated CYP2W1 expressed in E. coli23.8 ±9.0 #KsZhao et al. (2016)
GeranylgeraniolReconstituted system with N-truncated CYP2W1 expressed in E. coli25.0 ±0.3 #KsZhao et al. (2016)
Free fatty acid (FFA)Reconstituted system with N-truncated CYP2W1 expressed in E. coli83 ± 17Xiao & Guengerich (2012)
16:0 Hydroxylation
18:1 Hydroxylation101 ± 31
18:1 Epoxidation95 ± 13
Lysophosphatidylcholines (LPC)
16:0 HydroxylationReconstituted system with N-truncated CYP2W1 expressed in E. coli14±5Xiao & Guengerich (2012)
18:1 Hydroxylation38 ± 12
18:1 Epoxidation34±11
LysophospholipidsReconstituted system with N-truncated CYP2W1 expressed in E. colin.a.Xiao & Guengerich (2012)
Others
BenzphetamineReconstituted system with N-truncated CYP2W1 expressed in E. coli380±40Wu et al. (2006)
Indoles, indolinesReconstituted system with N-truncated CYP2W1 expressed in E. coli4.53 ±0.30Yoshioka et al. (2006)
Stably transfected HEK293 cellsn.a.Gomez et al. (2010)

hydroxylated and/or epoxidated. In general, all of these reactions occur at low rate, demonstrate Km in the micromolar range and need to be verified under physio- logical conditions.

By employing a high throughput screening method, as assessed by the extent of high spin conversion, Zhao et al. were able to identify retinoids as very high affinity ligands for CYP2W1 with a binding constant at nanomo- lar levels (Zhao et al., 2016). Heterologously expressed and truncated CYP2W1 was under in vitro conditions able to convert all-trans retinoic acid (atRA) into 4-

hydroxy atRA, all-trans retinol to 4-OH all-trans retinol and also oxidize retinal. However, the CYP2W1 mediated metabolism of arachidonic acid as well as the conver- sion of 17ß-estradiol to 2-hydroxy-(17฿)-estradiol, and farnesol to a monohydroxylated product were found to be much less efficient. The apparent affinity for catalysis of the retinoids by CYP2W1 was, however, much lower than for the binding. Taking into account the low level of intracellular retinoids in a variety of cancer cells, but most specifically in colon tumors, the authors suggest that the expression of CYP2W1 probably plays an

important role in the metabolism of retinoids in colorec- tal tumors. Nevertheless, the true in vivo metabolism of retinoids by CYP2W1 in tumor or embryonic tissues still remains to be established.

Membrane topology and post-translational modification

Detection of CYP2W1 in transfected cells and cancer tis- sues consistently revealed two immunoreactive bands (52 and 54 kDa), which prompted a hypothesis of a pos- sible post-translational modification of the enzyme. In silico analysis indicated Asn177 as a potential site for N-glycosylation. Indeed, treatment of membrane fractions from transfected cells or tumors with the deglycosylating enzymes, EndoH and PNGase caused a shift in the migration of 54 kDa band to 52 kDa indica- tive of glycosylation of CYP2W1 being responsible for the occurrence of the 54 kDa immunoreactive band. Mass spectrometry analysis confirmed glycosylation of Asn177 and mutation of this residue led to the dis- appearance of the 54 kDa band (Gomez et al., 2010). The Asn177 residue was found to be important for CYP2W1 catalytic activity (Guo et al., 2016) but still the role of the glycosylation for CYP2W1 function remains to be established.

The discovery of a sugar moiety on CYP2W1 is puz- zling as the classical membrane topology of CYPs (Figure 2) precludes their interaction with glycosyltrans- ferases residing in the ER luminal space. However, (i) results from protease protection and (ii) results obtained based on expression of CYP2W1 as a fusion protein with a redox-sensitive luciferase that acquires full activity only upon oxidative folding characteristic for the ER lumen, demonstrated that CYP2W1 is indeed atypically localized (Figure 2) (Gomez et al., 2010; Guo et al., 2016). A subfraction of CYP proteins can be inversely incorpo- rated into the ER to some minor extent; however, the predominant luminal orientation of CYP2W1 is truly unique. Interestingly, it was previously documented that a topological inversion of a small fraction of some CYPs in the ER may direct the protein to the plasma mem- brane similar to many membrane proteins that are matured in the ER (Neve & Ingelman-Sundberg, 2000, 2010). In line with this cell surface biotinylation assays showed approximately 10% of CYP2W1 being localized at the plasma membrane (Gomez et al., 2010).

The luminal topology of CYP2W1 (Figure 2) presents a certain conundrum: the identity of the electron donors remains an important question since the CYP2W1 local- ization toward the luminal side of the ER makes interac- tions with the P450 reductase (POR) electron transport

Figure 2. The ER membrane topology of CYP2W1. All other microsomal CYPs described are oriented toward the cytosolic part of the ER membrane with the exception for CYP2W1. This enzyme is to a certain extent glycosylated at Asn177 and may receive elec- trons from luminally generated NADPH, probably by hexose-6 phosphate dehydrogenase (H6PD) via an unidentified yet redox partner.

NADP+

NADPH

e

Cytosol

ER membrane

CYP2C8

POR

H6PD

e

N177

CYP2W1

NADP+

NADPH

NADP+

Hypothetical luminal reductase

ER lumen

chain impossible. This was confirmed using cells with a knocked down expression of POR and inhibitors of POR and NADH cytochrome b5 reductase (Guo et al., 2016). The intracellular CYP2W1 mediated metabolism of chlor- omethylindoline substrates in, for example intact cells indicates a functional electron transport chain with a specific electron donor for CYP2W1. NADPH can be pro- duced in the ER lumen and there are indications that CYP2W1 might be a terminal recipient of electrons from such luminally generated NADPH (Guo et al., 2016). Thus, the enigmatic redox partner(s) of CYP2W1 should be searched among various ER luminal oxidoreductases.

CYP2W1 regulation

The oncofetal pattern of CYP2W1 expression may sug- gest a potential role of epigenetics in the regulation of gene expression. First evidence of such an epigeneti- cally based regulatory mechanism was obtained with the discovery of re-expression of CYP2W1 in cell lines using the demethylating agent AzaC (Karlgren et al., 2006). Later a CpG island in the exon 1-intron1 junction of the CYP2W1 gene has been identified (Gomez et al., 2007). The extent of methylation here was found to inversely correlate with the expression levels of CYP2W1 in the tumors and cell lines. A similar regulatory mech- anism was also found during the development in mouse and human colon and intestine (Choong et al., 2015). Thus, a developmental shift from hypomethylation to hypermethylation in the exon 1/intron 1 CpG island of human CYP2W1 correlated with the silencing of the gene in adult age, consistent with a similar hypermethy- lated state of the CYP2W1 gene in the normal colon ver- sus hypomethylation in the colon tumor tissues. The epigenetic regulation in the murine Cyp2w1 gene is apparently more complex as correlation between the expression and methylation was found only in four out of six scanned CpG sites outside the canonical CpG islands (Choong et al., 2015).

Not much is known about the transcriptional regula- tion of CYP2W1 gene expression. A diverse set of bio- logically active compounds, such as 5-Aza-2’- deoxycytidine (AzaC), the tyrosine kinase inhibitor imati- nib, linoleic acid and its two conjugates significantly increased the CYP2W1 mRNA levels in the HCC2998 human colon cancer cells constitutively expressing CYP2W1 (Choong et al., 2015). However, ligands of PPARy, CAR PXR or AhR that efficiently regulate many CYPs did not induce CYP2W1.

The induction by linoleic acid and its derivatives lends also more support to the hypothesis that CYP2W1 preferentially binds to the ligands of lipid origin, such as fatty acids, lysophospholipids and retinoids.

Interestingly, an earlier study also showed that the expression of CYP2W1 mRNA was upregulated during differentiation of human keratinocytes in vitro (Du et al., 2006b) and substances such as retinoic acid and aryl hydrocarbon receptor ligands that modulate the differ- entiation also affected the expression of CYP2W1 (Du et al., 2006a). However, the AhR agonist TCDD was not effective in the colon carcinoma HCC2998 cells.

Polymorphism

The members of CYP2 family, for example CYP2C9, CYP2C19, CYP2D6, are highly polymorphic, which affects the interindividual susceptibility to the drugs metabo- lized by these enzymes. Several single nucleotide poly- morphisms (SNPs) have been described in CYP2W1. A study covering all nine exons and exon-intron junctions of CYP2W1 gene in a Japanese population reported six nonsynonymous SNPs present in the exons 1, 4 and 9. Haplotype analysis indicates six allelic variants (CYP2W1*1 -* 6) (Hanzawa et al., 2008). Among them, CYP2W1*6 displays the most significant frequency with the amino acid substitution Pro488Leu. Ten novel SNPs have been described recently in Chinese Uygur and Han populations with seven of them in the exons. Four of them had amino acid substitutions that might affect CYP2W1 function according to computational predic- tion. The allelic frequencies are different among Chinese Han, Uygur and Caucasians, for example 2.9%, 5.2%, 9.1% for CYP2W1*2 and 0%, 0%, 33.1% for CYP2W1*3, respectively (Hanzawa et al., 2008; Qi et al., 2015).

Taking into account the CRC-specific expression of CYP2W1 it is reasonable to assume that SNPs affecting its function or altering the expression might contribute to the etiology and/or prognosis of CRC. Gervasini et al. attempted to find possible association between the can- cer risk and expression of certain allelic forms of CYP2W1. It was concluded that the Ala181Thr amino acid change originating from the CYP2W1*2 allelic vari- ant that was suggested to affect the catalytic function of CYP2W1, is associated with the lower risk for CRC (Gervasini et al., 2010). However, a thorough study based on a 10-fold larger cohort of patients could not find the differences in the CYP2W1 allele or genotype frequencies between the control subjects and CRC patients and therefore the conclusion of a lower risk of CRC for the carriers of CYP2W1*2 allele could not be cor- roborated. Moreover, the hypothetical impact of Ala181Thr (CYP2W1*2) substitution on the catalytic activ- ity of the enzyme was also not supported as the func- tional analysis of all of the three CYP2W1 variants (CYP2W1*1, CYP2W1*2 and CYP2W1*6) using a high affinity substrate, ICT2726 revealed comparable kinetic

parameters and ICT2706 revealed identical propensity to produce cytotoxic metabolites (Stenstedt et al., 2013). In the light of the recent finding concerning many novel rare mutations in CYP genes including CYP2W1 (Fujikura et al., 2015), it is likely that the previously unknown germline as well as somatic rare mutations in the CYP2W1 may indeed cause functional alterations. It can be suggested that many mutations in the CYP2W1 gene occur during the colon cell transformation and it would be interesting to identify the frequency and localization of those in the different cancer tumors.

CYP2W1 as a target for colon cancer therapy

The unique CRC specific expression of CYP2W1 in colon cancer tissues suggests that in addition to its role as a prognostic marker, CYP2W1 can also serve as a target in the prodrug based cancer therapy whereby an inert pro- drug can be converted by CYP2W1 into cytotoxic metabolites, harmful only for cancer cells expressing CYP2W1. This approach was successfully demonstrated in a series of recently published studies. The search for potent anti-cancer drugs led to a synthesis of a number of analogs of duocarmycins, natural antitumor antibiot- ics with DNA alkylating propensity. As the parent com- pounds were found excessively toxic in animal studies the chloromethylindoline duocarmycin analogs were designed with the cytotoxic activity dependent on metabolic activation by CYP1A1 (Sutherland et al., 2013). Because the substrate recognition site of CYP2W1 shares the highest sequence identity with CYP1A1 (Karlgren et al., 2006), it was hypothesized that certain chlorome- thylindolines might also represent CYP2W1 substrates. Indeed, chemosensitivity screens using a panel of cell lines including HEK293 and SW480 stably expressing CYP2W1 (SW480-CYP2W1) identified several chlorome- thylindolines as substrates for these enzymes, where two compounds, designated ICT2705 and ICT2706 dem- onstrated very high affinity for activation by CYP2W1 (EC50=0.5 and 0.3 uM correspondingly). The LC/MS ana- lysis of ICT2706 metabolism in the CYP2W1 expressing cells identified a specific metabolic product (MW 363.2, m/z 364.2), consistent with a hydroxylated derivative of ICT2706 (Sheldrake et al., 2013) (Figure 3).

As predicted, the mechanism of the cytotoxic action of CYP2W1 generated chloromethylindoline metabolites was reminiscent of the one of their parent drugs, duo- carmycins, for example DNA alkylation, which was shown by phosphorylation of H2A.X histone, a positive indicator of DNA damage. It also appeared that the cytotoxic metabolite of chloromethylindolines can be transferred from one tumor cell to others in close

proximity, thus indicating a bystander propagating effect (Sheldrake et al., 2013; Travica et al., 2013).

The antitumor potential of these drugs was evaluated also in vivo in a xenograft model. ICT2706 was adminis- tered intraperitoneally to the mice bearing human colon cancer xenografts generated by subcutaneous injection of SW480-CYP2W1 cells. This resulted in the abolished tumor growth, whereas CYP2W1 negative xenografts displayed unrestrained growth. In addition, LC/MS ana- lysis of ICT2706 in tumors and plasma from dosed mice showed a higher concentration in the tumor, indicating preferential accumulation of the drug in the target tis- sue. Furthermore, no signs of toxicity were seen in mice subjected to ICT2706 multiple dosing (Travica et al., 2013).

CYP2W1 is expressed at high levels in 30% of primary CRC and 50% of liver metastases. Therefore, there is an obvious need to identify patients who would benefit from CYP2W1-targeted therapy. A candidate nontoxic chloromethylindoline CYP2W1 substrate, ICT2726 has been developed that can be used for assessing the activity of CYP2W1 without induction of toxicity and therefore constitutes a potential biomarker signaling the presence of a functional enzyme in the colon tumors.

Conclusions

CYP2W1 appears to be an enzyme with a very restricted temporal and tissue expression. Its high conservation between mammalian species and the absence of any major functionally important polymorphisms might indi- cate an important physiological function. However, the precise endogenous role remains to be elucidated, although one can speculate about its involvement in the development of the GI tract. A potential endogen- ous substrate in this case might be a hypothetical lipid ligand, such as retinoids showing high binding affinity to CYP2W1. It is also tempting to speculate that such function might also be important for the survival and expansion of colon tumors.

CYP2W1 attracts major attention due to its uniquely specific localization in colon tumors and therefore as having a potential as a prodrug target for the treatment of highly malignant CRC and metastases. One confound- ing factor here is that it has not been shown yet that the colon cancer cell resident CYP2W1 is catalytically active, although the CaCo-2TC7, colorectal adenocarcin- oma cells were found to express an active CYP2W1 (Kozyra et al., unpublished observations). Furthermore, the lack of knowledge on the luminal electron donor also makes it difficult to predict the sensitivity of differ- ent colon cancer types to the treatment with CYP2W1 activated prodrugs, since the efficacy of such treatment

Downloaded by [University of Nebraska, Lincoln] at 07:23 04 June 2016

Figure 3. Bioactivation of duocarmycins by CYP2W1 to potent cytotoxic endproducts. A. Whereas the chloromethylindolines ICT2705 and ICT2706 are bioactivated to cytotoxic products causing alkylation of DNA adenine, ICT2726 constitutes a high affinity substrate for CYP2W1 where the metabolite is inert. B. Structures of the chloromethylindolines Figure from Sheldrake et al., (2013). http://pubs.acs.org/doi/full/10.1021/jm4000209.

(A)

A

CI

Nº-Adenine

A

N

R

HO

N

1

ZI

4

R

N

deactivated duocarmycin bioprecursor

H

chloromethylindoline-DNA adduct is lethal to the cancer cell

CYP1A1, CYP2W1

sequence-selective alkylation

CI

A

spirocyclization

A

HO

N

R

N

R

2

ZI

3

N

H

seco-duocarmycin

cyclopropane ring necessary for biological potency

(B)

F

CI

F

CI

CI

CI

N

NH

N

NH

N

NH

O

NH

NH

O

ICT2705

ICT2706

ICT2726

would be dependent not only on the level of CYP2W1 expression but also on the level of expression of the putative electron donor.

Bioactivation of prodrugs by different CYPs is not a novel approach in cancer therapy; however, it is com- monly based on the targeted delivery of CYP genes to the tumor tissue, the approach prone with many setbacks, whereas confinement of CYP2W1 to the colon cancer cells circumvents the need for gene delivery as well as the unwanted effects resulting from the hepatic prodrug acti- vation. Future chemotherapeutics based on the ICT2706 structure or similar compounds being activated by CYP2W1 could be trialed in an adjuvant setting of colon cancer to further prevent tumor growth or metastases or in a neoadjuvant setting to assist tumor shrinkage or abla- tion before surgery. Therefore, such a chloroindoline based colon cancer treatment targeting CYP2W1 is very promising and is expected to find soon its way to the clinic. In addition, growing evidence indicates expression of CYP2W1 in tumors of different tissue origin and it will

be very interesting to see whether such expression is indeed significant and if the application of specific CYP2W1 prodrugs in these cancers would be possible.

As mentioned above, successful implementation of CYP2W1 as a target in cancer therapy would require more detailed knowledge of its catalytic mechanisms, bringing up yet unresolved issue of the members of CYP2W1 electron transport chain in the ER lumen and perhaps also, the role of glycosylation for its functional activity. Regulation of CYP2W1 is yet another important aspect of future CYP2W1 research, which hitherto has not been intensively targeted, although finding potent inducers would be beneficial for the adjuvant cancer therapy. In addition, discovering the endogenous func- tion and substrate(s) of CYP2W1 in the tumor cells could open up additional possibilities for anticancer therapy by designing analogs to such compounds. Last but not least, the unique colon cancer specific localization of the CYP2W1 enzyme makes it indeed very attractive to put efforts to develop another series of specific prodrugs

which are more soluble and easy to generate as com- pared to the chloroindolines.

Disclosure statement

The research in the laboratory is supported by grants from The Swedish Cancer Foundation and from The Swedish Research Council.

References

Bandala C, Floriano-Sanchez E, Cardenas-Rodriguez N, et al. (2012). RNA expression of cytochrome P450 in Mexican women with breast cancer. Asian Pac J Cancer Prev 13:2647-2653.

Choong E, Guo J, Persson A, et al. (2015). Developmental regulation and induction of cytochrome P450 2W1, an enzyme expressed in colon tumors. PLoS One 10:e0122820. Choudhary D, Jansson I, Stoilov I, et al. (2005). Expression pat- terns of mouse and human CYP orthologs (families 1-4) during development and in different adult tissues. Arch Biochem Biophys 436:50-61.

Du L, Neis MM, Ladd PA, Keeney DS. (2006a). Differentiation- specific factors modulate epidermal CYP1-4 gene expres- sion in human skin in response to retinoic acid and classic aryl hydrocarbon receptor ligands. J Pharmacol Exp Ther 319:1162-1171.

Du L, Neis MM, Ladd PA, et al. (2006b). Effects of the differen- tiated keratinocyte phenotype on expression levels of CYP1-4 family genes in human skin cells. Toxicol Appl Pharmacol 213:135-144.

Edler D, Stenstedt K, Ohrling K, et al. (2009). The expression of the novel CYP2W1 enzyme is an independent prognos- tic factor in colorectal cancer - A pilot study. Eur J Cancer 45:705-712.

Eun CY, Han S, Lim YR, et al. (2010). Bioactivation of aromatic amines by human CYP2W1, an orphan cytochrome P450 enzyme. Toxicol Res 26:171-175.

Fei-Lei Chung F, Wai Mai C, Yuen NGP, Leong CO. (2015). Cytochrome P450 2W1 (CYP2W1) in colorectal cancers. Curr Cancer Drug Targets 16:71-78.

Fujikura K, Ingelman-Sundberg M, Lauschke VM. (2015). Genetic variation in the human cytochrome P450 super- gene family. Pharmacogenet Genomics 25:584-594.

Gervasini G, DE Murillo SG, Ladero JM, Agundez JA. (2010). CYP2W1 variant alleles in Caucasians and association of the CYP2W1 G541A (Ala181Thr) polymorphism with increased colorectal cancer risk. Pharmacogenomics 11:919-925.

Gomez A, Karlgren M, Edler D, et al. (2007). Expression of CYP2W1 in colon tumors: Regulation by gene methylation. Pharmacogenomics 8:1315-1325.

Gomez A, Nekvindova J, Travica S, et al. (2010). Colorectal cancer-specific cytochrome P450 2W1: Intracellular localiza- tion, glycosylation, and catalytic activity. Mol Pharmacol 78:1004-1011.

Guo J, Thiess S, Johansson I, et al. (2016). Membrane topology and search for potential redox partners of colon cancer- specific cytochrome P450 2W1. FEBS Lett 590:330-339.

Hanzawa Y, Sasaki T, Mizugaki M, et al. (2008). Genetic poly- morphisms and haplotype structures of the human CYP2W1 gene in a Japanese population. Drug Metab Dispos 36:349-352.

Hartanto FK, Karen-Ng LP, Vincent-Chong VK, et al. (2015). KRT13, FAIM2 and CYP2W1 mRNA expression in oral squa- mous cell carcinoma patients with risk habits. Asian Pac J Cancer Prev 16:953-958.

Karlgren M, Gomez A, Stark K, et al. (2006). Tumor-specific expression of the novel cytochrome P450 enzyme, CYP2W1. Biochem Biophys Res Commun 341:451-458.

Molina-Ortiz D, Camacho-Carranza R, Gonzalez-Zamora JF, et al. (2014). Differential expression of cytochrome P450 enzymes in normal and tumor tissues from childhood rhabdomyosarcoma. PLoS One 9:e93261.

Neve EP, Ingelman-Sundberg M. (2000). Molecular basis for the transport of cytochrome P450 2E1 to the plasma mem- brane. J Biol Chem 275:17130-17135.

Neve EP, Ingelman-Sundberg M. (2010). Cytochrome P450 proteins: Retention and distribution from the endoplasmic reticulum. Curr Opin Drug Discov Devel 13:78-85.

Nishida CR, Lee M, DE Montellano PR. (2010). Efficient hypoxic activation of the anticancer agent AQ4N by CYP2S1 and CYP2W1. Mol Pharmacol 78:497-502.

Qi GZ, Wang X, Miao XJ, et al. (2015). Novel single nucleotide polymorphisms (SNPs) of CYP2W1 gene in Chinese Uygur and Han populations. Drug Metab Pharmacokinet 30:449-452.

Ronchi CL, Sbiera S, Volante M, et al. (2014). CYP2W1 is highly expressed in adrenal glands and is positively associated with the response to mitotane in adrenocortical carcinoma. PLoS One 9:e105855.

Sheldrake HM, Travica S, Johansson I, et al. (2013). Re- engineering of the duocarmycin structural architecture enables bioprecursor development targeting CYP1A1 and CYP2W1 for biological activity. J Med Chem 56:6273-6277.

Stenstedt K, Hallstrom M, Johansson I, et al. (2012). The expression of CYP2W1: A prognostic marker in colon can- cer. Anticancer Res 32:3869-3874.

Stenstedt K, Hallstrom M, Ledel F, et al. (2014). The expression of CYP2W1 in colorectal primary tumors, corresponding lymph node metastases and liver metastases. Acta Oncol 53:885-891.

Stenstedt K, Travica S, Guo J, et al. (2013). CYP2W1 poly- morphism: Functional aspects and relation to risk for colo- rectal cancer. Pharmacogenomics 14:1615-1622.

Sutherland M, Gill JH, Loadman PM, et al. (2013). Antitumor activity of a duocarmycin analogue rationalized to be metabolically activated by cytochrome P450 1A1 in human transitional cell carcinoma of the bladder. Mol Cancer Ther 12:27-37.

Tan BS, Tiong KH, Muruhadas A, et al. (2011). CYP2S1 and CYP2W1 mediate 2-(3,4-dimethoxyphenyl)-5-fluoro- benzothiazole (GW-610, NSC 721648) sensitivity in breast and colorectal cancer cells. Mol Cancer Ther 10:1982-1992.

Travica S, Pors K, Loadman PM, et al. (2013). Colon cancer- specific cytochrome P450 2W1 converts duocarmycin ana- logues into potent tumor cytotoxins. Clin Cancer Res 19:2952-2961.

Wang K, Guengerich FP. (2012). Bioactivation of fluorinated 2- aryl-benzothiazole antitumor molecules by human cyto- chrome P450s 1A1 and 2W1 and deactivation by cyto- chrome P450 2S1. Chem Res Toxicol 25:1740-1751.

Wu ZL, Sohl CD, Shimada T, Guengerich FP. (2006). Recombinant enzymes overexpressed in bacteria show broad catalytic specificity of human cytochrome P450 2W1 and limited activity of human cytochrome P450 2S1. Mol Pharmacol 69:2007-2014.

Xiao Y, Guengerich FP. (2012). Metabolomic analysis and iden- tification of a role for the orphan human cytochrome P450

2W1 in selective oxidation of lysophospholipids. J Lipid Res 53:1610-1617.

Yoshioka H, Kasai N, Ikushiro S, et al. (2006). Enzymatic prop- erties of human CYP2W1 expressed in Escherichia coli. Biochem Biophys Res Commun 345:169-174.

Zhang K, Jiang L, He R, et al. (2014). Prognostic value of CYP2W1 expression in patients with human hepatocellular carcinoma. Tumour Biol 35:7669-7673.

Zhao Y, Wan D, Yang J, et al. (2016). Catalytic activities of tumor-specific human cytochrome P450 CYP2W1 towards endogenous substrates. Drug Metab Dispos 44:771-780.