inhibition of experimental tobacco carcinogen induced head and neck carcinogenesis

Oral Oncology (2004) 40 611–623

http://intl.elsevierhealth.com/journals/oron/

Inhibition of experimental tobacco carcinogeninduced head and neck carcinogenesis

Joel Schwartz*, Vikki Baker, Eric Larios, Dhimant Desai, Shantu Amin

American Health Foundation, Molecular Pathology Facility, Institute for Cancer Prevention,One Dana Road, Valhalla, NY 10595, USA

Received 24 November 2003; accepted 10 December 2003

Summary Oral cancer models have attempted to demonstrate inhibition of oralcarcinogenesis. These models used synthetic carcinogens, lacked a specificmechanism of activity or used non-physiologic doses for carcinogen or inhibitor.To correct these problems the tobacco and environmental carcinogen, dib-enzo[a,l]pyrene (DB[a,l]P) (0.25%, 0.010 lM/application) was painted on the tongueand/or vitamin E acid succinate (VEas) (0.43 I.U./0.136 (lM/treatment) adminis-tered by gavages to Syrian hamsters (14 animals per group) using physiologic lowdoses, 5X/week. Oral cytology supplied keratinocytes after 1, 10, or 25 weeks oftreatment. Cells were analyzed by flow cytometry/laser scanning cytometry.Initiation (1–6 weeks) was suppressed by reducing DNA damage (oxidation lesions:8-oxo-dG), and repair (comet, fpg, OGG1, NTH1). Reduction in promotion (6–10weeks) was identified by depressed proliferation (cell cycle, bromodeoxyuridineincorporation (BrdU)) and aneuploidy (propidium iodide stain). p53 and apoptosisexpressions were increased (Sub G1, mitochondrion activation: Apo 2.7, andnucleosomal formation: mebstain (TUNEL)). VEas administration reduced dysplasia(10 weeks) and oral cancer formation at 25 (0/7 vs. 5/7 DB[a,l]P) and 30 weeks (3/7vs. 6/7 DB[a,l]P). Inhibition of oral carcinogenesis by VEas involved reversal ofseveral cellular events that contribute towards oral cancer.

�c 2004 Elsevier Ltd. All rights reserved.

KEYWORDS

Head and neckcarcinoma;

Tobacco carcinogen;Dibenzo[a,l]pyrene;Vitamin E acidsuccinate

Introduction

In the United States 22% of the total adult pop-ulation smoke tobacco products.1 To prevent orreduce the number of new oral cancers, a reduc-

* Corresponding author. Tel.: +1-914-789-7125; fax: +1-914-592-6317.

E-mail address: [email protected] (J. Schwartze).

1368-8375/$ - see front matter �c 2004 Elsevier Ltd. All rights reserdoi:10.1016/j.oraloncology.2003.12.012

tion in tobacco use or environmental exposure re-quires insights into mechanisms for malignanttransformation and prevention of oral keratino-cytes.

The hamster oral cancer tumor model permitsthis evaluation assisting clinicians in the treatmentof oral cancer patients. The hamster oral cancermodel has advantages. (1) The technique for car-cinoma induction is simple, and lacks spontaneousoral carcinoma formation. (2) Easy access to

ved.

mail to: [email protected]

612 J. Schwartz et al.

evaluate the development of oral cancer is present.(3) The surface area for analysis of the pouch isenormous. (4) Histopathologic changes, biochemi-cal, molecular similarities to human tissue changeshave been documented. (5) Consistent and analo-gous responses by the human mucosa to chemo-preventives and chemotherapeutic agents arenoted (e.g., tocopherols, retinoids, carotenoids,alkylating and anti-metabolites). (7) A large body ofinformation concerning the staging of oral carci-nogenesis in the hamster tumor model is available.(8) The lack of appendages such as sebaceous glandsor hair follicles facilitates a clear indication for theactivity of a carcinogen on oral keratinocytes.2;3 Amajor disadvantage has been an over reliance ofsynthetic polycyclic aromatic hydrocarbon (PAH),7,12-dimethylbenz[a]anthracene (DMBA) to induceoral cancer. Another problem has been the site ofinduction in the buccal cheek pouch, which is ananatomic site lacking lymphatic drainage and notpresent in humans.2 This has been corrected byinducing severe dysplasia and oral carcinoma in thetongue using tobacco carcinogens. Endophyticgrowth, as occurs in human oral carcinoma of thetongue is rapidly spread by a rich lymphatic drain-age to the mediasteum.2;4

To correct the use of DMBA, another PAH,benzo[a]pyrene (B[a]P), a tobacco derived carcin-ogen, has been applied to the hamster buccalpouch to induce oral carcinoma. The tumor inci-dence and multiplicity is not equal to DMBA. B[a]Papplied in a dose of 32 lM produced oral carcinomain the buccal mucosa of the hamster in 61% of agroup of 20 hamsters. 33.3% were found to havedysplastic lesions after 6 months of treatment. Onetumor per pouch and a volume of about 5–8 mm3

per tumor was noted.5

To our knowledge, DB[a,l]P another environ-mental and tobacco derived carcinogen has notbeen used to generate oral cancer. DiB[a,l]P willinduce lung carcinoma in other rodent models.6;7

DiB[a,l]P carcinogenesis induction has severaladvantages over B[a]P. DB[a,l]P, a fjord regioncontaining structure compared to the bay regioncarcinogen, B[a]P has a higher level of carcinogenicpotency (Fig. 1). A smaller quantity is required forcarcinoma formation producing a higher tumorincidence in more types of experimental animals,for example in mice or rats.6;7

B[a]P and DB[a,l]P are found in low concen-trations (several parts per billion) in smoke con-densate. To parallel human exposure experiencesmall physiologic doses were used in this study.Doses used were also guided by prior applicationconcentrations and toxicity in other rodent mod-els.6–9

VEas has been used in animal studies,10–12 clinicalstudies13;14 or cell culture15–21 assays to inhibit oralcarcinoma cell growth. VEas has not been studied forprevention or inhibition of experimental tobaccooral carcinogensis. This study was conducted tofurther our knowledge for the clinical use of VEas. Tobegin to establish a new oral cancer model usingphysiologic doses of a tobacco carcinogen. To recordthe oral carcinogen potency for DB[a,l]P, and toverify the importance for a sequence of cell stagesduring oral carcinogenesis and inhibition (Fig. 2).

Methods and materials

Chemicals

Dibenz[a,l]pyrene (DB[a,l]P) was provided bythe Synthetic Chemistry Facility of the Institute forCancer Prevention, headed by Dr. Shantu Amin. Dl-alpha tocopherol acid succinate was purchasedfrom Sigma Chemical, St. Louis, MO.

In vivo, oral carcinogensis study

Performed as previously reported,2–4;11;12 femalehamsters (Mesocricetus auratus, Golden SyrianHamsters) were placed into groups of fourteen andtreated according to the group. Vehicle control(0.25% w/v acetone), DB[a,l]P dissolved in acetoneto aid penetration and adhesion of tongue mucosa(0.25%, 0.010 lM/application) or 98 I.U. (110.3 mg),0.43 I.U/0.136 lM/treatment). VEas was adminis-tered by gavages down the throat of hamsters afterthe application of the carcinogen. A human equiv-alent to the dose administered was calculated to be132.8 I.U. This level of concentration is routinelyself-administered by capsules sold in 200–800 I.U.Treatment was 5X/week and continued for 30weeks. Each animal produced an oral cytologysample using a soft “cytology brush” for weeks 1,10, or 25 weeks. Three animals were euthanizedweek 1, 6, 10, 25 from each group. Groups were:

(1) Acetone control; (2) Mineral oil control; (2)DB[a,l]P, (3) DB[a,l]P, and VEas and (4) VEas. Controlgroups were nearly identical in respones; to sim-plify the presentation we show only the acetonecontrol.

Weights

The initial four week-old starting point of 70grams to a maximum of 247 grams was noted. Fig. 3shows the distribution of weights during this studyfor each group.

Figure 1 Sequential stages required for transformation of normal to malignant oral mucosa. A description of thesequence of stages required for malignant transformation for oral keratinocytes and vitamin E acid succinate pre-vention/inhibition of oral carcinogenesis is presented. Using the tobacco and environment carcinogen, diB[a,l]P wasapplied to the mucosa surface of the tongue. Stages presented are early (initiation), intermediate (promotion) or lateevents (tumor formation). Early events are comprised of DNA damage and repair, which was initially depressed, thenincreased. Intermediate, promotion was associated with premalignant lesion formation. Cell growth, increased DNAcontent and apoptosis characterized this stage. Apoptosis was reduced as cancer appeared. Cancer formation was alsotied to nuclear instability. Vitamin E acid succinate administration during carcinogenesis induced by diB[a,l]P de-creased initiation. DNA damage was depressed. DNA repair increased then decreased with a reduction in cell growth asthe intermediate, promotion stage appeared. Maintenance of normal DNA content, and enhanced apoptosis of dam-aged cells also was noted. A reduced numbers of cells exhibiting nuclear instability were noted and fewer tumorsformed.

Figure 2 Structures of benzo[a]pyrene and dib-enzo[a,l]pyrene. Polycyclic aromatic hydrocarbon chemicalsand respective “bay” and “fjord” regions are depicted.

Vitamin E Prevents Head and Neck Carcinoma 613613

Dl-alpha tocopherol acid succinate andhydrolysis determinations of oral cells

(Vitamin E acid succinate, VEas) (Sigma Chemi-cal, St. Louis, MO) 5 · 106 oral cells were analyzed

using a reverse phase high-pressure liquid chro-matography22 to assess the percentage of hydroly-sis for dl-alpha tocopherol acid succinate exposedto PAH.

Oral cytology

Using a “cytobrush” and a harvesting methodrecently described by our laboratory, oral kerati-nocytes from the tongue mucosa were harvested.23

Samples were obtained from all animals in thestudy for each time point. The technique employedremoval of cells from the stratum basilis containingthe proliferative cells. Oral cells were immediatelywashed in phosphate buffered saline (IX) and fixedwith 1% paraformaldehyde. Cells were analyzedusing laser-scanning cytometry, contouring, settingof threshold, background, and pixel intensity.23

Figure 3 A record of the distribution of weights with time. Treatment of animals for thirty-three weeks and changesin the mean weights per group is provided. Mean standard deviations for weights (grams) per group were: Con-trol ± 13.33; diB[a,l]P ± 18.48; diB[a,l]P + VEas ± 22.49, and VEas ± 19.61.


Morphology changes

Micro-photographic analysis of cells evaluatedfor the presence of nuclear cytoplasmic reversalratio, pseudopodia, mitoses, and cell size (400X)was obtained using a Hitachi digital color cameraand a Systemax, Compucyte PC. A gross examina-tion of hamsters was also performed to identify anytumor masses in oral tissues and the head andneck.

Histopathology

Histologic tissue sections stained with hema-toxylin and eosin was obtained from paraffinembedded and 10% buffered formalin fixed tissues.Two pathologists reviewed histopathology sectionsblinded to the treatment groups. Using a scale of0–1, normal oral mucosa; 2–3, mild dysplasia, andgreater than 3, moderate dysplasia with each cri-teria weighted equally. Tissue samples were ob-tained from three animals from each group at eachtime point.

Cell cycle

Oral cytology cells were incubated with theproliferation, S phase marker, bromodeoxyuridine(BrdU, 1 lg/ml)) for 24 h and assessed using fluo-rescent antibody detection to BrdU (OncogeneScience, Cambridge, MA). Compatible computerlanguage, “WinCycle”, developed for us by PhoenixFlow Systems, San Diego, CA and “ModFit”, pur-chased from Verity Software House, Topsham, MEwere also used to analyze data. The normalizedDNA content data obtained from the laser scanning

cytometer (Compucyte, Boston) was interpreted inthe flow cytometer for cell cycle determinations(Beckman Coulter, Epics XL) as previouslydescribed.23

DNA damage

The presence of oxidation markers, 8-oxo-deoxyguanosine (8-oxo-dG) were detected aspreviously described23 using fluorescence andquantified using a laser scanning cytometer at aemission spectra of 488 nm.

DNA repair

The level of activity for DNA repair glycoslases,fpg (formamidopyrimidine), 8-oxo dG glycosylase(OGG1), exonuclease III (NTH1) assays were per-formed using commercial kits developed by Trevi-gan, Gaithersburg, MD. A comet assay, is a singlecell gel electrophoresis assay based upon the abil-ity of denatured, cleaved DNA fragments to mi-grate out of the cell under the influence of anelectric field (comet). Undamaged DNA migratesslower and remains within the confines of thenucleoid when a current is applied.

Fragment length repair enzyme (Flare) assay,uses lyzed cells which are immobilized in a layer oflow melting point agarose. The cells are incubatedwith the specific enzyme, which catalyzes the re-moval of the specific substrate. Slides are im-mersed in an alkali solution to unwind the DNAstrands and are subject to neutral gel electropho-resis. The denatured cleaved DNA fragments mi-grate out of the cell to form a flare under gelelectrophoresis. SYBR Green fluorescence is added


to visualize and detect comets and flares. Analysiswas obtained using laser scanning cytometry and aformula for tail moment, which was used to iden-tify the comets formed is provided below.

(Integral-Fish/Integral) X (X-Fish X) X¼ X posi-tion of the whole event; X Fish¼ X position of theFish spot (comet head).

DNA content

The cells are treated and analyzed as previouslydocumented.23 In brief, propidium iodide (1.0 lg/ml) was used to stain the nucleus. DNA content isrecorded using laser scanning cytometer analysis(Compucyte, Boston, MA).

Apoptosis induction

Cell cycle, sub G1 was used to initially screen thecells for apoptosis, followed by Apopain/Caspase 3:A fluorescent equivalent, Apo 2.7 (Immunotech,Beckman Coulter Company). TUNEL, mebstain as-say was used to assess the level of histone nucle-osome formation as previously described(Trevigan, MD).2;4;20;23

Tumor masses

Weeks 25 and 30 the number and sizes (meandiameter (mm)) of tumor masses observable in thehead and neck regions of the animals were re-corded.

Statistical evaluation

Each assay was evaluated using a mean andstandard deviation for each group and assay for atleast 10,000 gated fluorescent cells (FITC or Phy-coerythrin).

Results

Oral cytology characterization

Initial staining with toluidine blue or Papanico-laou staining indicated greater than 85% of the cellsnucleated and about 1% of oral cells showed mi-cronuclei. Control and VEas treatments producedsimilar results. DiB[a,l]P treatments gradually in-creased a percentage of micronuclei in oral kerat-inocytes with time. Week 1, about 2%, week 6,about 8% and week 10, about 12% of the totalharvested cell population showed micronuclei. VEas

administration during diB[a, l]P exposure reducedmicronuclei formation. Week 1, about 1%, week 6,about 2% and week 10 about 4% of the cells dis-played micronuclei. The trypan blue dye exclusion(0.25%) determination for viability produced85–90% viability (17.0 · 104 cells/ll). DB[a,l]Pproduced 94% nucleated cells; 78% viability, in36.0 · 104 cells/ll. VEas treatment increased thenumber of nucleated cells to 95% and the viabilityto 95% in 44.0 · 104 cells/ll.

Hydrolysis of VEas

2.0% of VEas hydrolyzed from 5 · 106 oral kerat-inocytes exposed in the laboratory to PAH (nmolper 106 cells¼ 0.115, Tocopherol, and 5.393 nmolof Tocopherol acid succinate) and VEas in con-junction with the PAH increased hydrolysis to 4.3%of oral cells (nmol per 106 cells¼ 0.227 Tocoph-erol, 5.273 nmol of Tocopherol acid succinate).VEas administered in vivo in oral cytologic samplesof oral keratinocytes produced 1.8% hydrolysis,(nmol per 106 cells¼ 0.099, Tocopherol and 5.401nmol Tocopherol acid succinate). VEas and PAHapplication in vivo, produced a 5.8% increase inhydrolysis (nmol per 106 cells¼ 0.567 Tocopherol,4.932 nmol of Tocopherol acid succinate).

DNA damage

After 1 week exposure DB[a,l]P produced highlevels of oxidation lesions such as 8-oxo-dG (Table1) but low levels of p53. In the identical cells wehave also recorded an increased mean percentageof cells both p53 and 8-oxo-dG positive comparedto controls. VEas and DB[a,l]P treatments in com-parison to sole DB[a,l]P treatments decreasedoxidation damage and decreased p53 and 8-oxo-dGdouble expressing cells. For example, 86.5%, forDB[a,l]P and 74.9%, for DB[a,l]P and VEas. VEasadministration 1, 10, or 25 weeks reversed theoxidation damage recorded for exposure toDB[a,l]P (Table 1). The single data values suggest areduction in DNA damage is linked to increasedexpressions of p53, weeks 1, 10, and 25 an indica-tor for repair, and cell cycle checkpoint control.

DNA repair

An array of analyses for DNA repair cell responseto different oxidation lesions by different repairenzymes is presented for weeks 1 and 10 (Table 2).

An important observation for week 1 exposure oforal keratinocytes to DB[a,l]P was a generalreduction in DNA repair as identified by global

Table 1 Changes in DNA damage and repair following administration of carcinogen or vitamin E acid succinatea

Groups 8-oxodG p53 Expression(week 1/week 10/week 25) (week 1/week 10/week 25)

Control 5.2/7.8/7.1b 5.7/5.1/6.7b

diB[a,l]P 26.7/33.8/46.5� 1.4/2.8/3.7diB[a,l]P + VEas 5.5/5.4/5.2 39.3/49.9/52.0�

VEas 7.9/7.1/7.8 17.4/16.5/13.4�

a 10,000 cells were routinely gated for analysis.b Mean and standard deviation for total number of samples for each group. Control: ±6.5%; diB[a,l]P: ±1.2%; diB[a,l]P ± VEas:

±8.8%, and VEas: ±7.4%, Control: ±11.0%; diB[a,l]P: ±11.2%; diB[a,l]P + VEas: ±9.8%, VEas: ±7.7%.* Signifcant difference, p > 0:01 compared to control and other experimental group.

Table 2 Determinations of the level of damage (1/10 weeks) following exposure to diB[a,l]P and preventionactivity of VEas in oral keratinocytes

Groups Comet# Fpg# OGG1# NTH1#

Control 31.2/33.8a ;b 23.5/29.0 18.0/15.1 22.5/20.5diB[a,l]P 39.5/20.1� 30.3/11.5� 28.2/12.3 25.8/13.9diB[a,l]P + VEas 29.0/41.3� 29.4/48.5� 22.5/33.0� 27.8/33.4VEas 24.7/28.8 15.6/26.3 7.8/9.9 10.5/9.3

# Comet and flare assays from oral keratinocytes harvested using a “cytology brush” from the Lateral border of the tongue andglycosylase DNA repair enzyme activity is recorded as an inverse to the % of cells, which appear as flares. 10–15,000 cells analyzedfor each sample from each group. Values indicated are percentages of cells exhibiting a comet or flare.

a Cell area exhibiting fluorescent (SyBr Green) and % cells showing fluorescent flares or comets, (e.g., smaller number of flare/comets indicates more repair) from 1 to 10 weeks of treatments.

b Mean± standard deviation. Comet: Control: ±8.8%; diB[a,l]P: ±2.1%; diB[a,l]P + VEas: ±6.3%; VEas: ±3.9%, Fpg: Control: ±11.4%;diB[a,l]P: ±)19.7%; diB[a,l]P + VEas: ±11.4%; VEas: ±10.7%, OGG: Control: +11.6%; diB[a,l]P: ±11.3%; diB[a,l]P + VEas: ±7.3%; VEas:±12.4%, NTH1: Control: ±3.1%; diB[a,l]P: ±10.2%; diB[a,l]P + VEas: ±11.7%; VEas: ±12.7%.

* Significant changes, p > 0:01, comparison between experimental group and control.


(COMET assay) and/or specific FLARE assays forDNA glycosylases (Fpg, OGG1, NTH1) (five treat-ments). Fifty treatments, week 10 produced anincrease in DNA repair. VEas in conjunction withcarcinogen application week 1, also increased re-pair but with continual exposure, week 10, a de-creased DNA repair was recorded (Table 2).Decreased DNA repair parallels decreased oxida-tion lesion formation and increased p53 expres-sions noted above.

Cell cycle-proliferation

Increased cell proliferation was noted after 1week exposure to DB[a,l]P treatments. FollowingDB[a,l]P treatments increased mean percentage ofcells in S phase (control: 38.0%, DB[a,l]P: 46.3%),G1 (control: 37.5%, DB[a,l]P: 56.4%) and G2 wasreduced (control: 36.5%, DB[a,l]P: 3.3%). Follow-ing week 10 and fifty treatments, G1 (control:36.0%, DB[a,l]P: 12%) and G2 (control: 38.8%,DB[a,l]P: 18.0%) showed decreased percentages ofcells. S phase values were increased from weeks 1

to 10. A specific comparison between percentagesof cells in S phase and a diploid or aneuploid DNAcontent was also studied as described below.

VEas treatments in conjunction with DB[a,l]Preduced the percentage of cells found in S phase,week 1 (26.7%) and enhanced the number in G1

(60.4%) or G2 (12.7%) compared to above controls(Fig. 4). Week 10, VEas and DB[a,l]P treatmentsresulted in 19.1% in G1, 58.6% in S and 15.6% in G2.Sole VEas administration reduced the number ofcells in S phase while increasing the number in G1

(Fig. 4).In general, weeks 1 or 10, acetone treatment

controls increased slightly (2%) percentages ofcells in G1 compared to untreated tongue mucosaand decreased slightly the percentage of cells in Sphase (2%).

Sub G1, an apoptosis indicator appeared afterweek 1 (five treatments) and 10 weeks (fiftytreatments) among DB[a,l]P treated cell popula-tions produced respective levels of 1.2%, and 6.1%.After 25 weeks of treatment, when tumor was re-corded DB[a,l]P treatments reduced numbers of

Figure 4 Distribution of cell cycle following diB[a,l]P and/or VEas administration. Oral keratinocytes harvested afterweek 1 or 10 weeks of treatment were analyzed following staining with propidium iodide. diB[a,l]P week 1 increasedpercentages of cells in S phase with a depression of cell number in G2 and mitosis but an increase in G1. Week 10 agreater percentage of cells in S phase was recorded and a reduction in percentages of cells in G1 and G2 and mitosisphases. Vitamin E acid succinate reduced the percentage of cells in S phase but increased G1 while reducing G2 +M.Week 10 this pattern continued with fewer cells found in S phase and fewer cells are found in G1 and G2. (standarddeviation for each mean percentage shown derived from the total number of samples from each group: control,±12.0%; diB[a,l]P, ±21.3%; diB[a,l]P ± VEas ± 14.6%, VEas ± 6.1%).


cells in sub G1 to 3.0%. After week 1, DB[a,l]P andVEas treatments of the oral cells, produced 1.5% subG1, apoptotic cells. After week 10 treatments,14.9% were sub G1 positive and 16.9% were positiveafter 25 weeks. Sole VEas treatments initially pro-duced week 1, a sub G1 percentage of cells, 3.0%,5.7% by week 10, and 7.4%, week 25 positive.

DNA content, aneuploidy

Week 1, five treatments DB[a,l]P increasedslightly DNA content, ploidy changes. A greater

Table 3 Changes in DNA content and cell proliferation fosuccinate

Groups DNA contentDiploid (2N)/aneuploid (3 + 4N)

(week 1a/week 10) (w

Control 98.3/1.7 93DiB[a,l]P 73.4/26.6� 32DiB[a,l]P + VEas 83.5/16.5 83VEas 93.6/6.4 81

a DNA content was ascertained following staining with propidiumcyte). An average sample contained about 10,000 cells and the perecorded above. BrdU incorporation (1 lg/ml) after 24 h was anal

b Mean ± standard deviation for total number of samples fordiB[a,l]P + VEas: ±7.7%VEas: ±8.0%.

c Mean± standard deviation for total number of samples for each±7.6%VEas: ±7.8%.* Significant changes, p > 0:01, comparison between control and e

number of aneuploid cells were noted after 10weeks. This analysis was performed concurrentlywith the expression of BrdU. BrdU incorporationidentifies single strand breaks and provides addi-tional evidence for enhancement or inhibition ofcell proliferation (Table 3). Results show increasednumbers of oral keratinocytes exposed to DB[a,l]Pthrough week 10 possessed abnormal DNA content,aneuploidy, had single strand breaks and increasedcell proliferation. VEas administration in conjunc-tion with DB[a,l]P maintained the percentage ofdiploid cells through week 10. Using flow cyto-metric software and a normalized cell number

llowing administration of carcinogen of vitamin E acid

Cell proliferationBrdU (S Phase)

eek 1a/week 10) (week 1a/week 10)

.9/6.1b 19.5/18.5c

.9/67.1� 20.7/82.5�

.1/16.9 12.9/70.2�

.9/18.1 22.6/62.6�

iodide and analysis using laser scanning cytometer (Compu-rcentage of cells, either diploid (2N) or aneuploid (3N + 4N) isyzed using the identical cell population.each group: DNA Content: Control: ±3.4% diB[a,l]P: ±10.3%

group: BrdU: Control: ±7.3% diB[a,l]P: ±8.7% diB[a,l]P + VEas:

xperimental group.

Figure 5 Histopathology of hematoxylin and eosin stained sections. Application of tobacco carcinogen, diB[a,l]Pbeginning week 1 (five treatments) demonstrated a 2–3 sites/tissue section which was composed of heterochromaticcells, hyperplasia, anaplasia, and growth into adjacent connective tissue of the tongue mucosa. Week 6 (thirtytreatments), increased number of sites, 4–5/tissue section were examined for the identical latter changes in cells.Week 10 (fifty treatments), 5–6/tissue section sites of severe dysplasia were observed with extensive proliferation intounderlying connective tissue. Anaplasia, nucelar cytoplasimic reversal, pleomorphism and hyperplasia extendedthrough all layers of the mucosa. In contrast, Vitamin E acid succinate treatments had suppressed many of the latterdescribed characteristics for oral keratinocytes during week 1, 6, or 10. A loss of mucosa extensions into the con-nective tissue was also noted particularly week 10. Dysplastic sites contained few hyperchromatic cells, exhibitinglimited pleomorphism and nuclear cytoplasic reveral. Control treatment was generally nearly identical to Vitamin Eacid succinate treated tongue mucosa (arrows show sites of dysplasia).


separation of diploid from aneuploid cells wereaccomplished.

Percentage of cells in S phase from either aaneuploid or diploid, DNA content (normalized cellnumber of at least 10,000 cells) obtained for thetime point of premalignant lesion formation, week10 samples were:

DB[a,l]P: aneuploid: 29.4%; diploid 60.7%., andDB[a,l]P and VEas: aneuploid: 17.7% diploid:

48.6%. DB[a,l]P treatments increased the numberof cells with abnormal DNA content in S phasecompared to the results observed with VEasadministration. VEas treatments maintained thenumber of diploid cells in S phase. Control valueswere normalized to about 5,000 cells because ofthe difficulty to obtain sufficient cell counts withaneuploidy: VEas treatment and cells in S phase:aneuploid: 11.6%, diploid: 32.2%. Control treat-


ment and cells in S phase: aneuploid: 11.2%, dip-loid: 37.7%. The percentage of cells in S phase foraneuploid and diploid, DNA content were lowerthan the treatment values.

Week 25, the time of tumor formation anddiB[a]P treatments produced a greater percentageof cells which were aneuploid, 72% compared todiploid, 26%. In these populations of cells, DB[a,l]Ptreatments resulted in an increased mean per-centage of cells in S phase: aneuploid: 53.5% anddiploid: 36.0%. DB[a,l]P and VEas treatments pro-duced a reduced percentage of aneuploid cells,17.1% and more diploid cells, 81.5%. Among thesecell populations the percentages of cells in S phasewere: aneuploid, 35.2%, and diploid: 56.4%. VEasinconjunction with DB[a,l]P produced fewer cells inS phase for the aneuploid cell population and morecell in S phase for the diploid cell populations atthe time of tumor formation than the number ofcells recorded for the DB[a,l]P treatment group.

Apoptosis

Confirmatory apoptosis assays were conductedto compare results to sub G1 values. Nuclearchanges, nucleosomal fragmentation and mito-chondrion induction, were analyzed (Table 4).Apoptosis induction as determined by histone-chromatin nucleosomal fragmentation after fivetreatments of carcinogen (week 1) was relativelylow in all the groups (about 2–6%) (Table 4). Afterfifty treatments, week 10, an increase in apoptosiswas clearly recorded in both carcinogen groups butnot in the sole treatment with VEas Week 25, thetime point for tumor formation, histone nucleoso-mal fragmentation was reduced after DB[a,l]Ptreatments (Table 4). VEas in addition to carcino-gen treatment increased apoptosis at the time oftumor presentation (Fig. 6).

Table 4 Changes in apoptosis following administration of

Groups Mebstaina

(week 1/week 10/week 25

Control 2.9/14.2/20.2diB[a,l]P 2.4/29.9/3.1�

diB[a,l]P + VEas 2.7/17.1/20.3VEas 6.7/10.6/10.7

Mean ± standard deviation.Mebstain: Control±11.7% diB[al]Pi:±8.5% diB[a,l]P + VEas: ±10.9% VEApo 2.7: Control: ±9.4% diB[a,l]P: ±1.9% diB[a,l]P + VEas: ±6.9% VEa

a Mebstain identifies histone-chromatin nucleosomal fragmentatib Apo 2.7 assay, identifies the present of activated caspase-3 whi

10,000 cells are gated for analysis and the percentage of cells posc Harvested tongue oral keratinocytes were analyzed after week

* Significant changes, p > 0:01, comparison between control and e

Following DB[a,l]P treatments induction ofapoptosis through caspase-3, Apo 2.7 inductiongenerally followed the same pattern (Table 4),which was relatively reduced levels week 1, in-creased level by week 10, and depressed by week25. DB[a,l]P and VEas treatments maintained thelevel of caspase-3 induction through week 10 toweek 25.

Histopathology for dysplasia and headand neck carcinoma

Histopathology analysis of tissue sections fromthree hamsters per group per time point was ob-tained from tongue mucosa. After week 1 onlysmall discrete areas of mild dysplasia were seen.Six weeks of treatment with DB[a,l]P showedmoderate to severe dysplastic changes (e.g.,pleomorphism, analplasia, hyperplasia, hyper-chromatism, mitotic figures, hyperkeratosis) andweek 10, these latter histopathology changes weremore extensive. VEas reduced the number of thesechanges weeks 1 through 10 (Table 5 and Fig. 5).

After 25 weeks of application by DB[a,l]P tumormasses around the oral cavity were observed andcounted. These 15 tumors averaged 3 mm indiameter. DB[a,l]P induced well differentiated oralcarcinoma (keratin pearl) (Table 6, Fig. 6). VEas inconjunction with DB[a,l]P treatments did not pro-duce a gross tumor mass at this, week 25, timepoint (0/7 animals). Week 30, we observed 6 tumormasses in the VEas and DB[a,l]P group (3/7 animalswith tumor) in comparison to the sole DB[a,l]Ptreatment group which exhibited 18 tumors (6/7animals with tumor), The diameter averaged 0.7mm for the VEas and DB[a,l]P group in comparisonto 1.2 mm average diameter for the DB[a,l]P group.

carcinogen or vitamin E acid succinate

Apo 2.7bc) (week 1/week 10/week 25c)

6.2/15.5/21.31.5/33.0/12.7�

14.7/31.8/56.8�

5.9/3.0/3.3�

as: ±4.6.

s: ±8.6.on using an in situ hybridization fluorescent detection.ch is a product of mitochondrion induction of apoptosis. Aboutitive for mebstain or Apo 2.7 are presented.1 or 10.xperimental groups.

Figure 6 Presentations of oral squamous cell carcinoma histology. Oral squamous cell carcinoma formation followingtreatment with diB[a,l]P after 25 weeks. (a) Histologic sections following diB[a,l]P treatment showed well differen-tiated squamous cell carcinoma with bizarre mitoses, hyperchromatism, nuclear cytoplasmic reversal, pleomorphism,extending into the adjacent connective tissue. (b) Additional administration with vitamin E acid succinate did notdemonstrate these histologic changes but blunted extensions into the connective tissue were noted compared to (C)mineral oil treatment.

Table 5 Histopathology for dysplasia followingexposure to diB[al]P and/or vitamin E acid succinatea

Groups Number: milddysplasia

Number: moder-ate/severe

Control 0 0VEas 0 0DiB[a,l]P 8 5DiB[a,l]P + VEas 2 0

a Grading: 1 ¼ slight hyperplasia is noted, 2 ¼ mild dys-plasia: hyperchromatism, pleomorphism, anaplasia, hyper-plasia, hyperkeratosis, 3 ¼ moderate/severe dysplasia, asabove, more cells, bizarre mitoses, 1–2/high power field.

Table 6 Head and neck tumors observed

Groups Number of tumors

Week 25 Week 30

Control 0/4 0/4VEas 0/4 0/4DiB[a,l]P 5/7 6/7a

DiB[a,l]P + VEas 0/7 3/7b

a diB[a,l]P treatment: Total number of tumors (week 25)was 15. Total number of tumors (week 30) was 18.

b diB[a,l]P + VEas treatments: Total numer of tumors (week25) was 0. Total number of tumors (week 30) was 6.


Discussion

Treatment with the environmental and tobaccocarcinogen DB[a,l]P initiated and promoted a nor-

mal oral keratinocyte to transform to a malignantcell as assessed by a series of cell events and oralcarcinoma formation. Vitamin E acid succinateadministration suppressed the formation of theselesions by inhibiting oral carcinogenesis. Basic helix-


loop-helix and Per-Arnt-Sim (bHLH-PAS) genes ex-press aryl hydrocarbon receptor (Ahr) and arylhydrocarbon nuclear translocator (Ahnt) pro-teins.24–29 DiB[21a,l]P is a ligand for Ahr, whichforms a heterodimer complex with Ahrnt.28;29 Both,VEas and tocopherol transfer protein, TTP andDB[a,l]P ligand-Ahr-Arnt complexes activate thepregname X (PXR) receptor by a xenobiotic respon-sive element sequence (XRE).30 Both of the fore-mentioned complexes will activate the transcriptionof phase I, cytochrome P450 (CYP450) and phase II,glutathione-S-transderase (GSTs) genes.30 Vitamin Ehas been shown to increase GST p detoxification ofPAH, and increased glutathione reduced (GSHred)levels in hamster oral carcinoma cells followingexposure to PAH.2 Activation of the gene expressingcarboxylesterase from the pregnane X complex, hasbeen reported for both complexes. Hydrolysis ofVEas to alpha tocopherol and acid succinate is per-formed by carboxylesterase. An increased hydrolysisof VEas was noted in this study following exposure toPAH. Increased formation of diol epoxide throughthe activity of carboxylesterase and epoxidehydrolase has been documented.23;31 It is suggested;a decrease in carboxyesterase activity or concen-tration will result in less diol-epoxide product. Diolepoxide contributes to the presence of DNA bulkyand oxidation adducts.32 A reduction in DNA adductsformations has been previously reported followingalpha tocopherol treatment and exposure to variouscarcinogenic agents.33 In this study 8-oxo-dGexpression was suppressed.

dl-Alpha tocopherol is a well-known anti-oxi-dant, reducing agent, which is capable of sup-pressing oxidation lesion formation during 1 weekexposure to DB[a,l]P. Oxidative metabolism ofDB[a,l]P is associated with enhanced cytoP450

activities. CYP450s can degrade side chains oftocopherol and tocotrienol to change redox char-acteristics by initial omega-oxidation and laterbeta-oxidation.34 Vitamin E also inhibits lipid per-oxidation by trapping peroxyl radicals reducinglipid hydroperoxides products.34 Hydrophobicity ofthe tocopherol molecule is another contributoryfactor regulating mutated or oxidized proteins. In arecent laboratory study incubation of mutated p53with vitamin E acid succinate increased stabilityand binding of VEas to exposed hydrophobic sites(data not shown). We suggest dipole electro-chemicals interactions in conjunction with reduc-ing agent capacity suppresses DNA oxidation lesionadduct formation as identified in this study.

During week 1 exposure to DB[a,l]P increases ofDNA repair with VEas administration during oralcarcinogenesis is anticipated to be an additionalproduct of reducing agent capacity of dl-alpha

tocopherol. A reduction in DNA repair activity fol-lowing continual exposure to DB[a,l]P through week10, is expected to be a product of peroxidation.Vitamin E can be oxidized to peroxyl radicals andepoxytocopherones, which hydrolyze easily to to-copherolquinones under acid conditions. Vitamin Ecan be regenerated under reducing conditions usingenzymatic catalysis or reducing capacities by otherantioxidants (e.g., vitamin C).34 A pro-oxidanteffect could contribute to a reduction in DNA gly-cosylases effectiveness by competitively decreasingrecognition of oxidation sites, or through a reduc-tion in enzyme level following lowered transcriptionactivity. Reported studies have generally found anincrease in DNA repair for alpha tocopherol butthese studies have not evaluated repair responsesover extended periods screening patients orincreasing lengths of exposure in cell cultures. Noneof the published studies we reviewed used dl-alphatocopherol acid succinate and no other study hasexplored the relationship of DNA repair to exposureto tobacco carcinogen during carcinogenesis.20;21

In previous studies, VEas and other forms inhib-ited cell proliferation of premalignant and malig-nant oral cells.2;15;16 Vitamin E induces a cellaccumulation in G1 of the cell cycle, with increasedexpressions for p53 protein, an inducer of G1 cellaccumulation.35 VEas will also regulate transcrip-tion of the p53 gene enhancing protection frommutation formation during oral carcinogenesis.2

Vitamin E acid succinate maintained diploid,DNA content during carcinogenesis. Aneuploidy,increased DNA content, is a result of a mistaken-segregation of homologous chromosomes, mitoticspindles, and cleavage failure.36–39 There is astrong linkage between aneuploidy and apoptosisinduction and elimination of abnormal cells.36–39

PAH carcinogens such as DB[a,l]P have also beenrecognized inducers of aneuploidy.38;39 Someinvestigators have also stated aneuploidy contrib-utes significantly to malignant tumor formation.38

In this study an increased DNA content, aneu-ploidy, and cell proliferation (S phase) correlatedwith premalignant lesions and malignant tumorappearances. Surviving aneuploid cells apparentlyhave a stronger inhibitory effect upon apoptosisinduction contributing to tumor formation. Incontrast, with additional VEas treatment a generalreduction in the total number of aneuploid cells(total diploid and aneuploid) and cell proliferation(S phase) were noted. Induction of apoptosisreducing the total number of aneuploid cells andtumor formation was also recorded. These resultsfor the first time provide evidence that anueploidyto a lesser degree is linked to early DNA damageand repair but to a greater and more significant


level tied to premalignant and malignant oralcancer formation.

Apoptosis induction by vitamin E acid succinatehas been demonstrated in oral and non-oral cellpopulations.19;20 We have reviewed 381 articlesextending back to our original studies beginning inthe late 1980s describing induction of caspase-3dependent or independent apoptosis processes invarious malignant cell and animal derived cellpopulations.

A review of the literature indicated 253 articlesalpha tocopherol or other tocopherol forms re-duced mutations or chromosome aberrations inlaboratory cells, animal models and in a few clini-cal investigations. Premalignant head and necklesions with a loss of heterozygosity (LOH) marker9q21 were observed with complete reversal tohistological normality following treatment withvitamin E, alpha interferon and 13 cis retinoicacid.40 Vitamin E and other chemopreventionagents were shown to have insignificant capacity tomodulate already established cytogenetic changes.Moreover, cytogenetic aberrations do not appearto be good markers for screening early DNA damageand repair events because extensive chromosomedamage noted in premalignant or malignant cells isa product of continual exposure to carcinogenand gradual depression in DNA repair. A suddencompromise in DNA repair systems would resultin severe aneuploidy and loss in viability. Grad-ual appearance of cytogenetic abnormalitiesis compatible with minor aneuploidy, andviability.

Published hepatocyte and Ames analysis studiesfollowing B[a]P and vitamin E treatments indicateda depression in mutation and chromatid sisterbreaks but there, are obvious differences betweenthese assays and animal studies.41 These include,types of cells, times of exposures, concentrationsof B[a]P, differences between metaphase chro-matid technique and spread, micronuclei presen-tation and prokaryote and eukaryote growthpatterns.

In this study, VEas administration was shown forthe first time to suppress low dose tobacco car-cinogen initiation (DNA damage/repair, cell pro-liferation) and promotion (cell proliferation, DNAcontent and apoptosis) oral carcinogenesis. Tumorincidence and numbers of tumors confirmed inhi-bition but not total prevention with increasingtime. DNA repair was depressed with continualexposure but cell proliferation was also depressed.Fewer cells had abnormal DNA content, and moredamaged cells were eliminated by apoptosisinduction. This sequential series of events pro-duced fewer and smaller tumors.

This initial study of DB[a,l]P oral carcinogenesisrequires substantiation using a dose responsecomparison. A larger study with increased numbersof animals to assess incidence and multiplicity oftumor formation is also required. Toxicity was notobserved but six months was needed to observetumor formation. A higher carcinogen dose withouttoxicity but reduction in tumor induction time, andcost, are our future goals.

References

1. Centers for Disease Control and Prevention (CDC). Cigarettesmoking-attributable morbidity-United States, 2000. MMWRMorb Mortal Wkly Resp 2003;52:842–4.

2. Schwartz JL. Hamster oral cancer model. In: Teicher B,editor. Models in Tumor Biology. New York: Humana Press;2000. Chapter 38.

3. Salley JJ. Experimental carcinogenesis in the cheek pouchof the Syrian hamster. J Dent Res 1954;33:253–62.

4. Schwartz JL, Gu X, Kittles RA, Baptiste A, Shklar G.Experimental oral carcinoma of the tongue and buccalmucosa: possible biologic markers linked to cancers attwo anatomic sites. Eur J Cancer Oral Oncol 2000;36:225–35.

5. Rao AR. Modifying influences of betel quid ingredients onB[a]P-induced carcinogenesis in the buccal pouch of ham-ster. Int J Cancer 1984;33:581–6.

6. Prahalad AK, Ross JA, Nelson GB, Roop BC, King LC, NesnowS, et al. Dibenzo[a,l]pyrene-induced DNA adduction tumor-igenicity, and Ki-ras oncogene mutations in strains A/Jmouse lung. Carcinogenesis 1997;18:1955–63.

7. Cavalieri EL, Higginbothan S, RamaKrishna NV, DevanesanPD, Todorovic R, Rogan EG, et al. Comparativedose-response tumorigenicity studies of dibenzo[alpha,l]pyrene versus 7,12 diemethylbenz[alpha]anthracene, benzo[alpha]pyrene and two dibenzo[alpha,l]pyrene dihdrodiolsin mouse skin and rat mammary gland. Carcinogenesis1991;12:1939–44.

8. Higginbotham S, RamaKrishna NV, Johansson SL, Rogan EG,Cavalieri EL. Tumor-initiating activity and carcinogeneicityof dibenzo[a,l]pyrene versus 7,12 dimethylbenz[a]anthra-cene and benzo[a]pyrene at low doses in mouse skin.Carcinogenesis 1993;14:875–8.

9. La Voie EJ, He ZM, Meegalla RL, Weyand EH. Exceptionaltumor-initiating activity of 4-fluorobenzo[j]-fluoranthene onmouse skin: comparison with benz[j]-fluoranthene, 10-fluoranthene, benzo[a]pyrene, dibenzo[a,l]pyrene and7,12 dimethylbenz[a]anthracene. Cancer Lett 1993;70:7–14.

10. Schwartz JL. Molecular and biochemical control of tumorgrowth following treatment with carotenoids or tocophe-rols. In: Prasad K, Santamaria RM, Williams RM, editors.Nutrition and Cancer Prevention and Treatments. Totowa,NJ: Humana Press; 1995. p. 287–316.

11. Schwartz JL, Shklar G, Trickler D. p53 in the anticancermechanism of vitamin E. Eur J Cancer Oral Oncol1993;29B:313–8.

12. Shklar G, Schwartz JL. Vitamin E inhibits experimentalcarcinogenesis and tumour angiogenesis. Oral Oncol, Eur JCancer 1996;32B:114–9.

13. Garewal HS. Beta-carotene and vitamin E in oral cancerprevention. J Cell Biochem Suppl 1993;17:262–9.


14. Brenner SC, Winn RJ, Lippman SM, et al. Regression of oralleukoplakia with alpha tocopherol: a community clinicaloncology program, Chemopreventative study. J Natl CancerInst 1993;85:44–8.

15. Shklar G, Schwartz JL. Tumor necrosis factor in experimen-tal cancer: regression with alpha tocopherol, beta carotene,and algae extract. Eur J Cancer Clin Oncol 1988;24:839–50.

16. Schwartz JL, Tanaka V, Khandekar V, Herman TS, Teicher B.Beta carotene and/or vitamin E as modulators of alkylatingagents in SCC-25 human squamous carcinoma cells. CancerChemoth Pharm 1991;29:207–13.

17. Sigounas G, Anagnostou A, Steiner M. Dl-alpha-tocopherolinduces apoptosis in erythroleukemia prostate and breastcells. Nutr Cancer 1997;28:30–5.

18. Schwartz JL, Antoniades DZ, Zhao S. Molecular and bio-chemical reprogramming of carcinogenesis thru the activityof antioxidants and prooxidants. NY Acad Sci1992;686:262–79.

19. Shklar G, Schwartz JL. Effects of vitamin E on oralcarcinogenesis and oral cancer. In: Pacher L, Fuchs J,editors New York: Marcel Dekker; 1993. p. 497–511.

20. Kline K, Yu W, Sanders BG. Vitamin E mechanisms of actionas tumor cell growth inhibitors. J Nutr 2001;131:143–163S.

21. Papas A, Vos E. Vitamin E: Cancer and apoptosis. Am J ClinNutr 2001;73:1113–4.

22. Zhang JG, Fariss MW. Thenoyltrifluoracetone, a potentinhibitor of carboxylesterase activity. Biochem Pharmacol2002;63:751–4.

23. Schwartz JL, Muscat JE, Baker V, Larios E, Stepenson GD,Guo W, et al. Oral cytology assessment by flow cytomery ofDNA adducts, aneuploidy, proliferation, and apoptosisshows differences between smokers and non-smokers. OralOncol 2003;39:842–54.

24. Leonard SW, Terasawa Y, Farese Jr RV, Traber MG. Incor-poration of deuterated RRR-or all-rac-alpha-tocopherol inplasma and tissues of alpha-tocopherol transfer protein—nullmice. Am J Clin Nutr 2002;75:555–60.

25. Kaempf-Rotzoll DE, Traber MG, Arai H. Vitamin E andtransfer proteins. Curr Opin Lipidol 2003;14:249–54.

26. Azzi A, Ricciarelli R, Zingg JM. Non-antioxidant molecularfunctions of alpha-tocopherol (vitamin E). FEBS Lett2002;519:8–10.

27. Kalikin LM, Bugeaud EM, Palmbos PL, Lyons Jr RH, Petty EM.Genomic characterization of human SEC14L1 splice variantswithin a17q25 candidate suppressor gene region and iden-tification of an unrelated embedded expressed sequence.Mamm Genome 2001;12:925–9.

28. Swanson HI. DNA binding and protein interactions of theAHR/ARNT heterodimer that facilitate gene activation.Chem Biol Interact 2002;141:63–76.

29. Sogawa K, Nakano R, Kobayashi A, Kiluchi Y, Ohe N,Matsushita N, et al. Possible function of Ah receptor nucleartranslocator (Arnt) homodimer in transcriptional regulation.Proc Natl Acad Sci USA 1995;92:136–1940.

30. Rushmore TH, Kong TAN. Pharacogenomics, regulation andsignaling pathways of Phase I and II drug metabolizingenzymes. Cur Drug Metabol 2002;3:481–90.

31. Xu G, Zhang W, Ma MK, McLeod HL. Human carboxylesterase2 is commonly expressed in tumor tissue and is correlatedwith activation of irinotecan. Clin Cancer Res 2002;8:2605–11.

32. Robichova S, Slamenova D. Study of N-nitrosomorpholine-induced DNA strand breaks in Caco-2 cells by the classicaland modified comet assay: influence of vitamins E and C.Nutr Cancer 2001;39:267–72.

33. Brigelius-Flohe R. Vitamin E and drug metabolism. BiochemBiophys Res Commun 2003;305:737–40.

34. Packer L, Fuchs J. Vitamin E in health and disease. NewYork: Marcel Dekker; 1993.

35. Patel V, Jakus J, Harris CM, Ensley JF, Robbins KC, YeudallWA. Altered expression and activity of Gl/S cyclins andcyclin-dependent kinase characterize squamous cellcarcinoma of the head and neck. Int J Cancer 1997;73:551–5.

36. Sen S. Aneuploidy and cancer. Curr Opin Oncol2000;12:82–8.

37. Brinkley BR. Managing the centrosome numbers game: fromchaos to stability in cancer cell division. Trends Cell Biol2001;11:18–21.

38. Duesberg P, Rasnick D. Aneuploidy the somatic mutationthat makes cancer a species of its own. Cell Motil Cytoskel2000;47:81–107.

39. Duesberg P, Li R, Rasnick D, Rausch C, Willer A, et al.Aneuploidy precedes and segregates with chemical carcino-genesis. Cancer Genet Cytogenet 2000;119:83–93.

40. Mao L, El-Naggar AK, Papdimitrakopoulou V, Shin DM, ShinHC, Zhou X, et al. Phenotype and genotype of advancedpremalignant head and neck lesions after chemopreventivetherapy. J Natl Cancer Inst 1998;90:1545–51.

41. Lykkesfedt J, Christen S, Wallock LM, Chang HH, Jacob RA,Ames BN. Ascorbate is depleted by smoking and repleted bymoderate supplementation: a study in male smokers andnonsmokers with matched dietary antioxidant intakes. Am JClin Nutr 2000;71:530–6.

inhibition of experimental tobacco carcinogen induced head and neck carcinogenesis

Documents