British Journal of Cancer (2004) 91, 1624 – 1631
`& 2004 Cancer Research UK All rights reserved 0007 – 0920/04 $30.00
`
`www.bjcancer.com
`
`Dietary induction of NQO1 increases the antitumour activity of
`mitomycin C in human colon tumours in vivo
`
`A Begleiter*,1,2, MK Leith1, JA Thliveris3 and T Digby2
`1Department of Internal Medicine, Manitoba Institute of Cell Biology, CancerCare Manitoba, University of Manitoba, 675 McDermot Avenue, Winnipeg,
`Manitoba, Canada R3E 0V9; 2Department of Pharmacology and Therapeutics, Manitoba Institute of Cell Biology, CancerCare Manitoba, University of
`Manitoba, 675 McDermot Avenue, Winnipeg, Manitoba, Canada R3E 0V9; 3Department of Human Anatomy and Cell Science, University of Manitoba,
`730 William Avenue, Winnipeg, Manitoba, Canada R3E 3J7
`
`The bioreductive antitumour agent, mitomycin C (MMC), requires activation by reductive enzymes like NAD(P)H:quinone
`oxidoreductase 1 (NQO1). We used a novel approach to increase MMC efficacy by selectively inducing NQO1 in tumour cells in
`vivo. CD-1 nude mice were implanted with HCT116 cells, and fed control diet or diet containing 0.3% of the NQO1 inducer,
`1 MMC and dicoumarol, an NQO1 inhibitor.
`dimethyl fumarate (DMF). The mice were then treated with saline, 2.0, 3.5 or 2.0 mg kg
`The DMF diet increased NQO1 activity by 2.5-fold in the tumours, but had no effect in marrow cells. Mice given control diet/
`
`1 MMC had tumours with the same volume as control mice; however, mice given DMF diet/2.0 mg kg1 MMC had
`2.0 mg kg
`1 MMC were similar to those in mice given control diet/
`significantly smaller tumours. Tumour volumes in mice given DMF/2.0 mg kg
`1 MMC. Tumour inhibition was partially reversed in mice given DMF/2.0 mg kg
`1 MMC and dicoumarol. DMF diet/
`3.5 mg kg
`1 MMC treatment did not increase myelosuppression and did not produce any organ toxicity. These results provide strong
`2.0 mg kg
`evidence that dietary inducers of NQO1 can increase the antitumour activity of bioreductive agents like MMC without increasing
`toxicity.
`British Journal of Cancer (2004) 91, 1624 – 1631. doi:10.1038/sj.bjc.6602171 www.bjcancer.com
`Published online 5 October 2004
`& 2004 Cancer Research UK
`
`Keywords: mitomycin C; antitumour activity; NQO1; diet; induction
`
`

`
`

`
`Bioreductive agents are a class of anticancer drugs that must be
`activated in cells by reduction. These agents are preferentially
`active in solid tumours, which represent the majority of cancers
`and the highest mortality. The clinical potential (Workman and
`Stratford, 1993) and mode of action (Rauth et al, 1993; Rockwell
`et al, 1993) of these agents have been actively studied, and the use
`of the prototype drug mitomycin C (MMC) in the clinic was
`recently reviewed (Begleiter, 2000). Bioreductive agents are
`generally activated by one-electron reducing enzymes
`like
`NADPH:cytochrome P450 reductase (RED) (Pan et al, 1984;
`Rockwell et al, 1993) or two-electron reducing enzymes like
`NAD(P)H:quinone oxidoreductase 1 (NQO1; DT-diaphorase)
`(Riley and Workman, 1992; Ross et al, 1993). RED and NQO1
`are the major activators for current agents (Begleiter et al, 1989,
`1992; Hoban et al, 1990; Riley and Workman, 1992; Begleiter and
`Leith, 1993; Ross et al, 1993; Patterson et al, 1995), but their role
`varies with the agent, enzyme, oxygen level (Marshall et al, 1989;
`Hoban et al, 1990; Begleiter et al, 1992) and pH (Siegel et al, 1990;
`Begleiter and Leith, 1993). Reduction of MMC activates alkylating
`
`*Correspondence: Dr A Begleiter, Departments of Internal Medicine and
`Pharmacology and Therapeutics, Manitoba Institute of Cell Biology,
`CancerCare Manitoba, University of Manitoba, 675 McDermot Avenue,
`Winnipeg, Manitoba, Canada R3E 0V9; E-mail: begleit@cc.umanitoba.ca
`Received 11 June 2004; revised 28 July 2004; accepted 29 July 2004;
`published online 5 October 2004
`
`groups that form DNA crosslinks, and this is the most important
`cytotoxic mechanism for this agent (Lown et al, 1976; Pan et al,
`1986; Tomasz et al, 1987; Riley and Workman, 1992). However,
`in the presence of oxygen, redox cycling occurs to form reactive
`oxygen species, which can also contribute to cytotoxic activity
`(Lown et al, 1976; Dusre et al, 1989; Riley and Workman, 1992).
`NQO1 catalyses
`two-electron reduction of quinones and
`nitrogen-oxides (Ernster, 1987; Riley and Workman, 1992). Several
`human diaphorases are known (Jaiswal et al, 1990, 1991; Riley and
`Workman, 1992), but NQO1 is most important for activating
`bioreductive agents (Jaiswal, 1991; Riley and Workman, 1992;
`Belinsky and Jaiswal, 1993). NQO1 is a homodimer that uses
`NAD(P)H as an electron donor (Riley and Workman, 1992). The
`enzyme is mainly cytosolic, but 5 – 10% is membrane bound (Riley
`and Workman, 1992). It is ubiquitous in eukaryotes but levels vary
`in different tissues (Schlager and Powis, 1990; Riley and Workman,
`1992; Belinsky and Jaiswal, 1993), with low levels in haematopoetic
`cells (Schlager and Powis, 1990; Siegel et al, 2001). NQO1 activity is
`usually higher in tumour than normal cells, but is lower in vivo
`than in vitro (Schlager and Powis, 1990; Riley and Workman, 1992;
`Belinsky and Jaiswal, 1993; Ross et al, 1994; Collard et al, 1995).
`Expression of NQO1 is transcriptionally regulated (Riley and
`Workman, 1992), and the enzyme is highly inducible by a wide
`variety of inducers (Prestera et al, 1993). The induction pathway is
`unknown but may involve a cytosolic redox signal that alters
`expression and/or interaction of transcriptional factors like Jun,
`
`Experimental Therapeutics
`
`Page 1 of 8
`
`

`
`ExperimentalTherapeutics
`
`1625
`
`Increasing mitomycin C activity by inducing NQO1 in vivo
`A Begleiter et al
`
`Cell lines and clonogenic assays
`
`HCT116 cells were grown in DMEM/Hams F12 media and 10%
`foetal bovine sera. Clonogenic assays were performed as previously
`described (Begleiter and Leith, 1993). Cloning efficiency ranged
`from 51 to 90%. Cells were incubated at 371C for 48 h in the
`presence or absence of 5 mM DMF and then were treated with
`various concentrations of MMC for 1 h. Cells were plated and
`colonies were counted 6 days later. A linear regression analysis of
`each concentration – survival curve was obtained and the D10 was
`derived from the negative reciprocal of
`the regression slope
`(Begleiter et al, 1992). The cytotoxicities were compared statisti-
`cally by a t-test comparing the significance of the differences of the
`slopes of the concentration – survival curves.
`For implantation into CD-1 nude mice, the HCT116 cells were
`counted, washed three times in HBSS, assessed for viability with
`trypan blue and resuspended at a concentration of 5 107 viable
`cells per ml in HBSS; 100 ml of the cell suspension was injected
`subcutaneously into the right flank of each mouse.
`
`NQO1 activity measurements
`
`NQO1 activity was measured as described previously (Doherty
`et al, 1998) using menadione as electron acceptor and is reported
`1 mg protein
`1. Protein was deter-
`as nmol MTT reduced min
`mined with the Bio-Rad DC kit using gamma-globulin as protein
`standard. NQO1 activity was measured in 0.25 M sucrose sonicates
`of HCT116 cells, mouse marrow cells and in sonicates of
`homogenised mouse tissue. Mouse organs were excised and stored
`frozen at 801C in sucrose until analysis. Marrow was obtained by
`flushing the mouse femurs with HBSS and pooled from several
`mice. The marrow cells were layered over NycoPrep 1.077A and
`spun at 600 g for 15 min; the resulting layer of cells was washed and
`frozen at 801C in sucrose until analysis. NQO1 activities in
`HCT116 cells in vitro were compared statistically by t-test. In
`mouse organs, the NQO1 activities were compared statistically by
`Mann – Whitney rank sum test.
`
`Effect of dose and schedule of DMF on induction of NQO1
`in vivo
`
`To determine the optimal concentration of DMF in the diet for
`NQO1 induction, CD-1 mice were fed experimental diet containing
`0, 0.2, 0.3 or 0.4% DMF for 2 weeks. The mice were then killed and
`the liver, forestomach and kidney were removed and stored frozen
`at 801C in 0.25 M sucrose until analysis for NQO1 activity.
`To determine the optimal length of DMF feeding for induction
`of NQO1 activity, CD-1 nude mice were implanted subcutaneously
`in the right flanks with 5 106 HCT116 cells and the mice were fed
`custom experimental diet containing no DMF for 18, 15, 11 or 4
`days. The mice were then fed experimental diet containing 0.3%
`DMF for 0, 3, 7 or 14 days, respectively, prior to killing 18 days
`following tumour implantation. The HCT116 xenografts were
`approximately 100 – 250 mm3 at this time. The mice were killed, the
`organs (kidney,
`liver,
`lung, heart and forestomach) and the
`HCT116 xenografts were removed and marrow was obtained. All
`tissues were frozen at 801C prior to measurement of NQO1
`activity.
`
`In vivo combination treatment studies
`
`For the in vivo combination treatment studies, CD-1 nude mice
`were implanted with 5 106 HCT116 cells subcutaneously in the
`right flank of each mouse. The mice were fed a standard irradiated
`mouse diet for the first 7 – 10 days following tumour implantation
`and then randomly switched to the experimental diet containing 0
`or 0.3% DMF for a further 7 – 10 days. At this time, day 0, mice
`with tumours measuring approximately 100 – 250 mm3 were
`
`Nrf, Maf, Fos and Fra with the xenobiotic response element and the
`antioxidant response element (Belinsky and Jaiswal, 1993; Favreau
`and Pickett, 1993; Venugopal and Jaiswal, 1996, 1998; Yao et al,
`1996; Kepa and Ross, 1999; Nguyen et al, 2000). An NF-kB element
`may also be involved in induction (Yao and O’Dwyer, 1995). NQO1
`is induced by a wide variety of dietary and synthetic agents
`including: dithiolethiones like Oltipraz, an antiparasitic agent;
`isothiocyanates like sulforaphane, found in cruciferous vegetables,
`and dietary metabolites
`like dimethyl
`fumarate (DMF), a
`metabolite of fumaric acid which is found in fruits and vegetables
`(Talalay et al, 1988; Prestera et al, 1993). NQO1 is a member of the
`phase II detoxifying enzymes that help to remove xenobiotics from
`cells and are important in early defence against carcinogenesis
`(Prestera et al, 1993). Inducers of NQO1 like Oltipraz have been
`tested as cancer preventive agents in animals and humans (Egner
`et al, 1994; Hecht, 1998; Kensler et al, 1998).
`Activation of MMC by NQO1 has been extensively studied (Riley
`and Workman, 1992; Ross et al, 1993). Cells with elevated NQO1
`levels are more sensitive to MMC and drug activity is decreased by
`the NQO1 inhibitor, dicoumarol (DIC) (Begleiter et al, 1989;
`Marshall et al, 1989; Siegel et al, 1990). NQO1 is a major activating
`enzyme for MMC and other bioreductive agents including, RH1
`and MeDZQ (Workman and Stratford, 1993; Riley and Workman,
`1992; Beall et al, 1995; Winski et al, 1998). Thus, selectively
`increasing the level of NQO1 in tumour cells by gene transfer
`(Rauth et al, 1998) or by selective induction of NQO1 in tumours
`may be useful for enhancing the efficacy of bioreductive agents.
`We showed previously that we could selectively increase NQO1
`activity in tumour cells compared with normal cells and that this
`enhanced MMC antitumour activity in vitro (Doherty et al, 1998;
`Wang et al, 1999). In this study we investigated whether we could
`selectively induce NQO1 activity in human tumours implanted in a
`nude mouse model, and if this would increase the antitumour
`activity of bioreductive agents without increasing the toxicity of
`these agents.
`
`MATERIALS AND METHODS
`
`Materials
`
`The human colon carcinoma HCT116 cell line was obtained from
`ATCC (Manassas, VA, USA) and DMEM/Hams F12 media, foetal
`bovine sera and Hanks balanced salt solution (HBSS) were
`obtained from Invitrogen (Burlington, ON, Canada). The
`HCT116 cells tested negative for mycoplasma contamination.
`DMF, MMC and DIC were from Sigma-Aldrich Canada (Oakville,
`ON, Canada) as were all reagents for the NQO1 activity assay.
`NycoPrep 1.077A was from Cedarlane (Hornby, ON, Canada). Bio-
`Rad DC Protein kit was from Bio-Rad (Canada) Ltd (Mississauga,
`ON, Canada).
`Female CD-1 and CD-1 nude mice (6 – 8 weeks of age) were
`obtained from Charles River Canada (Montreal, QC, Canada) and
`were maintained according to institutional regulations. The mice
`were
`fed standard irradiated rodent
`chow except during
`the experimental diet period. The experimental diet for the mice
`was a custom powdered autoclavable, semipurified diet with
`antioxidant free corn oil and vitamin K instead of menadione
`from ICN Biochemical Division (Aurora, OH, USA). Diets with
`different concentrations of DMF in the diet were made by the
`addition of finely ground DMF to the custom diet followed by
`autoclaving.
`Microvettes CB 300 with potassium EDTA, used for blood
`collection, were obtained from Sarstedt (St Leonard, QC, Canada).
`Zap-oglobin II lytic reagent and Isoton II were obtained from
`Beckman Coulter
`(Mississauga, ON, Canada). Ketalean and
`Rompum were obtained from Central Animal Care, University of
`Manitoba (Winnipeg, MB, Canada).
`
`& 2004 Cancer Research UK
`
`British Journal of Cancer (2004) 91(8), 1624 – 1631
`
`Page 2 of 8
`
`

`
`1
`1 protein
`1 mg
`94.073.0 nmol min
`cells
`control
`in
`to
`1
`1 mg protein
`194.075.3 nmol min
`in DMF
`treated
`cells
`(Po0.001) (Figure 1). Cells that had been pretreated with DMF
`and then were treated with MMC were more sensitive to MMC than
`cells that did not receive DMF (Figure 1). The D10 for cells treated
`with MMC alone was 2.1270.13 mM while that for cells treated with
`DMF and MMC was 1.5570.43 mM, and this difference was
`statistically significant (Po0.001).
`
`Effect of dose and schedule of DMF on induction of NQO1
`in HCT116 cells in vivo
`
`CD-1 mice were fed a diet containing 0, 0.2, 0.3 or 0.4% DMF for 14
`days. NQO1 activities in the kidney, liver and forestomach were
`increased in mice fed a diet containing DMF compared to mice
`receiving a control diet; however, enzyme activities in mice fed diet
`containing 0.3 or 0.4% DMF were not significantly different.
`CD-1 nude mice implanted with HCT116 cells
`received
`experimental diet containing 0.3% DMF for 0, 3, 7 or 14 days.
`NQO1 activity increased in the implanted tumours and reached a
`plateau level at approximately 7 days on the DMF diet (Figure 2).
`In contrast, enzyme levels in forestomach, kidney, lung and liver
`tissues continued to increase to 14 days on the DMF diet, while
`NQO1 activity in heart tissue remained unchanged. NQO1 activity
`
`* P < 0.001
`
`*
`
`Control
`
`DMF
`
`200
`
`100
`
`0
`
`(nmol min−1 mg protein−1)
` NQO1 activity
`
`1
`
`0.1
`
`0.01
`
`Surviving cell fraction
`
`MMC
`MMC + DMF
`
`0.001
`
`0
`
`2
`MMC (♯M)
`
`4
`
`Figure 1 Effect of DMF on cytotoxic activity of MMC in HCT116 cells in
`vitro. Cells were incubated at 371C for 48 h in the absence or presence of
`5 mM DMF. NQO1 activity was measured in some of the cells, and the
`remaining cells were treated with MMC for 1 h. Cells were plated and
`surviving cell
`fraction was determined by clonogenic assay. The points
`represent the mean surviving cell fraction7s.e. of 5 – 15 determinations.
`The lines are linear regression lines. Inset:
`level of NQO1 activity in cells
`incubated in the absence or presence of DMF. The bars represent the
`mean NQO1 activity7s.e. of 8 determinations. The means were
`compared by a t-test evaluating the significance of the difference of the
`NQO1 activity in control and DMF incubated cells.
`
`Increasing mitomycin C activity by inducing NQO1 in vivo
`A Begleiter et al
`
`1626
`
`randomly assigned to treatment groups. The mice were weighed
`and received a single tail-vein injection of saline (control diet and
`1 MMC (control diet and 0.3% DMF
`0.3% DMF diet), 2.0 mg kg
`1 MMC (control diet only). In addition, some
`diet) or 3.5 mg kg
`mice on the 0.3% DMF diet received a single i.p injection of
`1 DIC 1 h prior to the 2.0 mg kg
`1 MMC injection. Some
`34 mg kg
`mice with tumours were killed at day 0 and organs, marrow and
`xenografts were removed for measurement of NQO1 activity. On
`day 1, the diet for all the mice was switched to the standard
`irradiated diet. On days 0, 3, 7, 10, 15 and 18, tumour diameters in
`three dimensions were measured using digital calipers, and
`tumour volume was calculated by the formula (l w d 0.5236)
`(Rockwell et al, 1972). Mouse body weights were also recorded on
`these same days. Differences in tumour volumes were analysed
`statistically by a t-test comparing the slopes of the regression lines
`for plots of tumour volume vs days after MMC treatment in mice
`receiving different treatments.
`
`White blood cell (WBC) and platelet counts
`On days 0, 3, 6, 9, 12 and 15, approximately 20 ml of blood was
`collected from the saphenous veins of some of the mice from each
`treatment group (Hem et al, 1998). The blood was collected into a
`microvette CB 300 with potassium EDTA and 10 ml aliquots were
`used for the determination of WBC or platelet counts. The counts
`were determined with a Coulter Z2 particle count and cell analyser
`(Beckman Coulter, Mississauga, ON, Canada). The Coulter Z2
`counter was optimised for mouse white blood cells and platelets
`according to the manufacturer’s protocols and all counts were
`analysed with the Coulter AccuComp software program. WBC were
`counted with a 100 mm aperture tube (set for a lower threshold of
`3 mm) following the addition of Zap-oglobin II lytic reagent to the
`coulter counter vial (50 ml per 10 ml Isoton II). For counting of
`platelets, 10 ml of mouse blood was diluted in 240 ml Isoton II and
`spun at 83 g for 3 min to remove the red blood cells. In total,
`100 ml of supernatant was then diluted in Isoton II and the platelets
`were counted using a 50 or 70 mm aperture tube with a lower
`threshold of 1.4 fl and an upper threshold of 24.4 fl. Differences in
`the WBC and platelet counts in different treatment groups were
`analysed statistically by ANOVA for each time point.
`
`MMC toxicity in vivo
`CD-1 nude mice (n¼ 3) without tumours were fed a 0 or 0.3%
`DMF diet for 7 – 10 days and then were treated with a single tail-
`1 MMC (control
`vein injection of saline (control diet), 2.0 mg kg
`diet and 0.3% DMF diet). After 7 days, the mice were anesthetised
`with Ketalean/Rompum and blood was obtained by cardiac
`puncture. The mice were then euthanised with carbon dioxide
`and organs were removed and fixed in neutral buffered formalin.
`The blood was allowed to clot and the serum was collected
`following centrifugation at 1500 g for 15 min. The serum was
`stored frozen at 801C until analysed. The serum was analysed in a
`Roche Hitachi 917 (Health Sciences Centre, Department of Clinical
`Chemistry, Winnipeg, MB, Canada) for the following parameters:
`Na, K, Cl, blood urea nitrogen, serum creatinine, alkaline
`phophatase, alanine transaminase, aspartate transaminase, gam-
`ma-glutamyl transpeptidase,
`lactate dehydrogenase. The organs
`were sectioned, stained with haematoxalin and eosin and examined
`histologically.
`
`RESULTS
`
`Effect of induction of NQO1 on cytotoxic activity of MMC
`in vitro
`
`When HCT116 cells were incubated in vitro with, or without,
`DMF,
`the
`level
`of NQO1
`activity
`increased
`from
`
`Experimental Therapeutics
`
`British Journal of Cancer (2004) 91(8), 1624 – 1631
`
`& 2004 Cancer Research UK
`
`Page 3 of 8
`
`

`
`Increasing mitomycin C activity by inducing NQO1 in vivo
`A Begleiter et al
`
`1627
`
`Table 1 Effect of DMF diet on NQO1 activity in tumours and normal
`tissues
`
`NQO1 activity (nmol min
`
`1 mg protein
`
`1)
`
`Tissue
`
`Control diet
`
`DMF diet
`
`Ratio
`DMF/control
`
`Tumour
`Marrow
`Kidney
`Liver
`Lung
`Heart
`Forestomach
`
`5975
`1372
`13578
`3972
`3375
`9175
`362728
`
`145710
`1773
`329713
`5472
`4374
`8677
`952754
`
`2.5*
`1.4
`2.4*
`1.4*
`1.3
`0.9
`2.6*
`
`CD-1 nude mice implanted with HCT116 tumours were fed a 0 or 0.3% DMF diet
`for 7 – 10 days. Mice were euthanised and NQO1 activity was measured in different
`tissues. Data are mean7s.e. of 9 – 17 mice. NQO1 activity in mice fed control diet
`and diet containing DMF were compared by Mann – Whitney rank sum test.
`*Po0.05.
`
`Control
`MMC (2.0 mg kg−1)
`MMC (3.5 mg kg−1)
`DMF + MMC (2.0 mg kg−1)
`DMF + MMC (2.0 mg kg−1) + DIC
`
`600
`
`400
`
`200
`
`Tumour volume (mm3)
`
`ExperimentalTherapeutics
`
`0
`
`5
`
`10
`Days post MMC
`
`15
`
`Figure 3 Effect of DMF diet on MMC antitumour activity in HCT116
`xenografts in CD-1 nude mice. Mice were implanted with tumour cells
`subcutaneously in the right flank of each mouse. After 7 – 10 days the mice
`were fed a diet containing 0 or 0.3% DMF for 7 – 10 days. The mice were
`weighed and received a single tail-vein injection of saline (control diet and
`1 MMC (control diet and 0.3% DMF diet) or
`0.3% DMF diet), 2.0 mg kg
`1 MMC (control diet only). In addition, some mice on the 0.3%
`3.5 mg kg
`1 DIC 1 h prior to the
`DMF diet received a single i.p. injection of 34 mg kg
`1 MMC injection. On days 0, 3, 7, 10, 15 and 18, tumour volumes
`2.0 mg kg
`were measured using digital calipers. The points represent the mean
`tumour volume7s.d. of 6 – 14 mice. The lines are linear regression lines.
`
`1
`that observed in mice fed control diet and treated with 3.5 mg kg
`MMC. DMF diet alone did not inhibit the increase in tumour
`volume (data not shown).
`
`Tumour
`Liver
`Forestomach
`Kidney
`Lung
`Marrow
`Heart
`
`500
`
`400
`
`300
`
`200
`
`100
`
`NQO1 activity (as % of control)
`
`0
`
`3
`
`9
`6
`Days on DMF diet
`
`12
`
`Figure 2 Effect of time on DMF diet on NQO1 activity in human
`tumour xenografts and normal tissues in CD-1 nude mice. CD-1 nude mice
`were implanted subcutaneously in the right flank with HCT116 cells and
`the mice received diet containing 0.3% DMF for 0, 3, 7 or 14 days. The
`mice were killed, the organs (kidney, liver, lung, heart, forestomach) and the
`HCT116 xenografts were removed and marrow was obtained. For each
`tissue, NQO1 activity was measured and is reported as the NQO1 activity
`in that tissue as a percent of NQO1 activity in the same tissue in mice fed
`diet containing no DMF. The points represent the means of values obtained
`from three mice.
`
`in bone marrow reached a peak at 3 days on the DMF diet, and
`decreased after that time.
`
`Effect of induction of NQO1 on MMC antitumour activity
`and toxicity in vivo
`
`CD-1 nude mice implanted with HCT116 cells were fed either
`control experimental diet or diet containing 0.3% DMF for 7 – 10
`days. Some mice were euthanised at that time and tumours and
`tissues were excised. The remaining mice were weighed, tumour
`volumes were measured and blood samples were obtained from
`some mice for WBC and platelet counts. The mice were then
`1
`treated with saline or DIC, and then with saline, 2.0 or 3.5 mg kg
`MMC. On various days, tumour volumes were measured and blood
`samples were obtained for WBC and platelet counts. Average
`tumour volumes on the day of MMC treatment ranged from 132 to
`163 mm3 for the different treatment groups and these differences
`were not statistically significant. The level of NQO1 increased by
`2.5-fold in the tumours and to a similar extent in the kidneys and
`forestomach (Po0.001) in mice fed diet containing DMF (Table 1).
`In contrast, NQO1 activity increased by 1.4-fold in liver (Po0.02),
`but did not
`increase significantly in bone marrow,
`lung or
`heart.
`The tumour volumes increased in all the mice, with the tumour
`volume in control mice (mice fed control diet and treated with
`saline) increasing by four-fold 18 days after treatment (Figure 3).
`1 MMC showed
`Mice fed control diet and treated with 2.0 mg kg
`the same rate of increase in tumour volume as control mice. In
`contrast, the rate of tumour growth in mice fed DMF diet and
`1 MMC was approximately 15% lower than
`treated with 2.0 mg kg
`the rate of tumour growth in control mice or in mice fed control
`1 MMC, and this difference was
`diet and treated with 2.0 mg kg
`statistically significant (Po0.01). Mice fed DMF diet that received
`1 MMC showed a partial reversal of the
`DIC prior to 2.0 mg kg
`enhanced antitumour effect observed in mice fed DMF diet and
`1 MMC. The antitumour effect seen in mice
`treated with 2.0 mg kg
`1 MMC was equivalent to
`fed DMF diet and treated with 2.0 mg kg
`
`& 2004 Cancer Research UK
`
`British Journal of Cancer (2004) 91(8), 1624 – 1631
`
`Page 4 of 8
`
`

`
`Table 2 Effect of DMF diet and MMC on blood chemistry
`
`Tissue
`
`Test
`
`MMC
`
`DMF+MMC
`
`Liver
`
`Kidney
`
`Dehydration
`
`Alkaline phosphatase
`Alanine transaminase
`Aspartate transaminase
`Gamma-glutamyl transpeptidase
`Lactate dehydrogenase
`Serum creatinine
`Blood urea nitrogen
`Sodium
`Potassium
`Chloride
`
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`Normal
`
`CD-1 nude mice without tumours were fed a 0 or 0.3% DMF diet for 7 – 10 days and
`1
`then were treated with a single tail-vein injection of saline (control diet), 2.0 mg kg
`MMC (control diet and 0.3% DMF diet). After 7 days, three mice/group were
`anaesthetised and blood was obtained by cardiac puncture. The blood was allowed
`to clot and the serum was collected and analysed.
`
`1 MMC.
`counts as mice fed control diet and treated with 2.0 mg kg
`DMF diet without MMC treatment had no effect on WBC count
`(data not shown).
`No significant decreases in platelet counts were observed in any
`of the groups of mice in the 15 days following MMC treatment
`(Figure 5). DMF diet without MMC treatment had no effect on
`platelet count (data not shown). There were no significant
`differences in body weight in the different treatment groups at
`18 days after MMC treatment (data not shown).
`In another experiment, CD-1 nude mice received either control
`experimental diet or diet containing 0.3% DMF for 7 – 10 days. The
`1 MMC and the
`mice were then treated with saline or 2.0 mg kg
`animals were euthanised 7 days later. Histological analysis of
`tissues from these mice showed no evidence of damage to kidney,
`liver, lung, heart, colon or forestomach tissues in the MMC treated
`mice fed either control or DMF. Furthermore, treatment with
`1 MMC did not produce any significant changes in
`2.0 mg kg
`blood chemistry in mice fed control or DMF diet (Table 2).
`
`DISCUSSION
`
`MMC is an anticancer agent that has proved to be most effective in
`the front line treatment of a small number of solid tumours, such
`as superficial bladder, gastric, pancreatic, anal, oesophageal and
`non-small-cell cancer (Remers, 1979, pp 27 – 32; Doll et al, 1985;
`Begleiter, 2000). It is also used in palliative treatment of advanced
`or resistant cancers, generally in combination regimens (Begleiter,
`2000). There has been continued interest
`in increasing the
`effectiveness of this agent because of its activity in solid tumours
`and its enhanced effectiveness against hypoxic cells that are
`resistant to radiation (Workman and Stratford, 1993).
`Use of MMC in the clinic has been severely restricted by its
`toxicity (Hortobagyi, 1993; Begleiter, 2000). The usual dose-
`limiting toxicity is myelosuppression, which occurs 3 – 4 weeks
`after drug administration, with recovery within 8 weeks (Begleiter,
`2000). Generally, thrombocytopenia or leukopenia is most severe,
`but anemia is also common. Less common, but potentially serious
`or fatal
`toxicities associated with MMC use include severe
`pulmonary toxicity that occurs in 5% of patients (Doll et al,
`1985; Begleiter, 2000), and cancer-associated haemolytic – uremic
`syndrome, which can occur in 4 – 15% of patients (Doll et al, 1985;
`Begleiter, 2000). In addition, some cases of severe congestive heart
`failure have been reported after MMC treatment
`in patients
`previously treated with doxorubicin (Doll et al, 1985).
`MMC requires intracellular activation by reductive enzymes like
`NQO1 and NADPH cytochrome P450 reductase (Rockwell et al,
`
`Increasing mitomycin C activity by inducing NQO1 in vivo
`A Begleiter et al
`
`1628
`
`Control
`MMC (2.0 mg kg−1)
`MMC (3.5 mg kg−1)
`DMF + MMC (2.0 mg kg−1)
`
`0
`
`3
`
`9
`6
`Days post-MMC
`
`12
`
`15
`
`20
`
`15
`
`10
`
`5
`
`0
`
`(10−9 × WBC count)/ l
`
`Figure 4 Effect of DMF diet on MMC toxicity to WBC in CD-1 nude
`mice. CD-1 nude mice were implanted with HCT116 cells subcutaneously
`in the right flank of each mouse. After 7 – 10 days the mice were fed diet
`containing 0 or 0.3% DMF for 7 – 10 days. The mice were weighed and
`1
`received a single tail-vein injection of saline (control diet only), 2.0 mg kg
`1 MMC (control dietMMC (control diet and 0.3% DMF diet) or 3.5 mg kg
`
`only). On various days, approximately 20 ml of blood was collected from
`the saphenous vein of some of the mice in each treatment group. The
`WBC counts were determined with a Coulter Z2 particle counter and cell
`analyser. The points represent the mean WBC count7s.e. of 5 – 16 mice.
`Differences in the WBC counts in different
`treatment groups were
`analysed statistically by ANOVA for each time point.
`
`Control
`MMC (2.0 mg kg−1)
`MMC (3.5 mg kg−1)
`DMF + MMC (2.0 mg kg−1)
`
`0
`
`3
`
`9
`6
`Days post-MMC
`
`12
`
`15
`
`20
`
`15
`
`10
`
`5
`
`0
`
`(10−11 × platelet count)/l
`
`Figure 5 Effect of DMF diet on MMC toxicity to platelets in CD-1 nude
`mice. CD-1 nude mice were implanted with HCT116 cells subcutaneously
`in the right flank of each mouse. After 7 – 10 days the mice were fed a diet
`containing 0 or 0.3% DMF for 7 – 10 days. The mice were weighed and
`1
`received a single tail-vein injection of saline (control diet only), 2.0 mg kg
`1 MMC (control dietMMC (control diet and 0.3% DMF diet) or 3.5 mg kg
`
`only). On various days approximately 20 ml of blood was collected from the
`saphenous vein of some of the mice in each treatment group. The platelet
`counts were determined with a Coulter Z2 particle counter and cell
`analyser. The points represent the mean platelet counts7s.e. of 2 – 14
`mice. Differences in the platelet counts in different treatment groups were
`analysed statistically by ANOVA for each time point.
`
`Mice fed control diet and treated with MMC showed a dose-
`dependent decrease in WBC counts with the nadir occurring 3 days
`after MMC treatment (Figure 4). WBC counts recovered to control
`levels by 6 days after drug treatment. Mice fed DMF diet and
`1 MMC showed a similar decrease in WBC
`treated with 2.0 mg kg
`
`Experimental Therapeutics
`
`British Journal of Cancer (2004) 91(8), 1624 – 1631
`
`& 2004 Cancer Research UK
`
`Page 5 of 8
`
`

`
`Increasing mitomycin C activity by inducing NQO1 in vivo
`A Begleiter et al
`
`1629
`
`ExperimentalTherapeutics
`
`1993; Workman and Stratford, 1993). Cells with elevated NQO1
`levels are generally more sensitive to MMC (Begleiter et al, 1989;
`Marshall et al, 1989; Siegel et al, 1990), and NQO1 can be induced
`by a wide variety of dietary and synthetic agents (Talalay et al,
`1988; Prestera et al, 1993). We have previously shown that NQO1
`activity can be selectively increased in tumour cells compared with
`normal cells, and that this enhances MMC cytotoxicity in human
`and murine cell lines in vitro (Doherty et al, 1998; Wang et al,
`1999). In this study we investigated a novel approach to increasing
`the antitumour activity of MMC by using a dietary component to
`selectively induce NQO1 activity in tumour cells compared with
`normal cells. We used DMF, a metabolite of fumaric acid that is
`found in fruits and vegetables (Talalay et al, 1988), as the inducer
`of NQO1 and HCT116 cells, which have a moderate level of NQO1
`activity, as a tumour model. This approach could be used with any
`bioreductive agent that is activated by NQO1, but we used MMC
`because it is the prototype bioreductive agent and is the only
`bioreductive agent in widespread clinical use.
`Treatment of HCT116 cells in vitro with DMF for 48 h increased
`NQO1 activity in these cells by two-fold (Po0.001), and cells that
`were pretreated with DMF were 1.4-fold more sensitive to MMC
`than cells that were not pretreated (Po0.001). This result was
`similar to our previous findings in murine lymphoma cells
`(Begleiter et al, 1996) and human breast, lung and colon cancer
`cells (Doherty et al, 1998; Wang et al, 1999) using synthetic
`inducers of NQO1.
`To evaluate the clinical potential of this approach to increasing
`the efficacy of MMC, we used a human tumour xenograft mouse
`model to investigate whether we could obtain a similar enhance-
`ment of antitumour activity in vivo. This model also made it
`possible to evaluate any enhanced toxicity that might result from
`the NQO1 induction. In addition, the NQO1 inducer, DMF, was
`added to the diet of the mice to test the feasibility of this route of
`administration. In preliminary studies we found that adding DMF
`to the diet of mice resulted in a dose-dependent increase in NQO1
`activity in various tissues; however, the increase in enzyme activity
`was similar with 0.3 and 0.4% DMF. Thus, 0.3% DMF was used for
`subsequent experiments. We also found that NQO1 activity in
`HCT116 tumours implanted in CD-1 nude mice reached a plateau
`after 7 days on the DMF diet while enzyme activities in other
`tissues including kidney, liver, lung and forestomach showed a
`further increase at 14 days. Interestingly, there was a very small
`increase