[CANCER RESEARCH 44, 2794-2798, July 1984]

Cytological Effects of Sulfur and Selenium Furine Analogues on TwoTransplantable Hepatomas and on Normal Renewing Cells in Mice1

Joan B. Melvin,2 Thomas H. Haight, and Elizabeth H. Leduc

Division of Biology and Medicine, Brown University, Providence, Rhode Island 02912 [J. B. M., E. H. L], and Department of Biology, Framingham State College,Framingham, Massachusetts 01710 [T. H. H.]

ABSTRACT

The effects of 50% lethal doses of three purine analogues, 6-methylmercaptopurine riboside (MMPR), 6-thioguanosine (TGR),and 6-selenoguanosine (SeGR), on mitotic activity in a slow-growing (SS1 H) and a fast-growing (BH3) transplantable hepa-

tocellular adenoma in C3H/SIW and BUB mice, respectively,were analyzed statistically. No significant difference in responsewas found between the two benign hepatomas. MMPR aloneeffectively reduced mitotic activity in the tumors and did so asefficiently on the first day of treatment as on subsequent daysof daily i.p. administrations for up to 10 days. TGR alone andSeGR alone were ineffective in reducing the mitotic index significantly below that of controls. When either TGR or SeGR wasinjected simultaneously with MMPR, the effect on tumor mitosisresembled that of MMPR alone.

The reactions of normal cells of the hosts to these agentswere analyzed quantitatively in duodenal epithelium with respectto mitotic activity and to the number of cells present in the crypts.Differences between the two strains of mice were small and, forthe most part, not significant. MMPR produced a slight but notsignificant reduction in duodenal mitotic activity and cell number.TGR alone induced significant decreases in both after 3 and 5days of treatment. SeGR alone had no effect on the duodena.The effects of a combination of SeGR with MMPR on theduodena differed only slightly from MMPR or SeGR alone, butTGR plus MMPR produced greater inhibition of mitosis than dideither administered alone.

Our results suggest that MMPR may be a promising chemo-

therapeutic agent against some types of solid hepatocellulartumors in vivo because it can inhibit mitosis in these tumorseffectively, rapidly, and continuously, while its inhibitory effect onnormal replicating cells of the host intestine occurs more slowlyand only with long-sustained treatment.

INTRODUCTION

Although chemotherapeutic control of some neoplasia, notablyleukemias and lymphomas, has been successful, progress intreating solid tumors has been slower. It has been postulated(17) that the reason for this is that most clinically useful agentsare principally effective against tumors which contain a largeproportion of dividing cells and that the majority of solid tumorsusually have a low proportion of cells in mitosis at any giventime. We have studied this problem by conducting a statisticalanalysis of the effects of 3 purine analogues on mitotic activityin 2 transplantable murine tumors of hepatocyte origin, one which

exhibits rapid growth and one which grows more slowly. Two ofthe agents tested, the sulfur-substituted purine analogues TGR3

and MMPR, have been clinically successful against leukemiasand lymphomas (15, 17). The third is a selenium derivative of apurine riboside, SeGR. The latter has been shown to have atherapeutic index higher than that of TGR when used againstmouse lymphocytic leukemia L5178Y cells in vitro (2) and Sarcoma 180 cells in vivo (18). MMPR was included as a test agentin part because of its demonstrated synergistic effect when it iscombined with TGR (13) or mercaptopurine (20), to determinewhether the same synergism would occur with SeGR.

We have found that, whereas TGR and SeGR did not have asignificant inhibitory effect on the tumors, MMPR rapidly inducedand subsequently maintained a statistically significant reductionin the number of mitotically active cells in both hepatomas. Thissuggests that a high level of cell replication is not always essentialfor successful chemotherapy.

MATERIALS AND METHODS

Chemotherapeutic Agents. MMPR was purchased from SigmaChemical Co., St. Louis, MO, and TGR was obtained from NutritionalBiochemicals, Cleveland, OH. SeGR was synthesized and kindly provided by Dr. S-H. Chu of Brown University. The nucleosides, rather than

the bases, were used to take advantage of their higher solubility. Theagents were dissolved with careful heating in sterile 0.9% NaCI solution(saline) just before use, such that each dose was injected in 1.0 ml ofthe saline solution. The drugs were administered singly and also incombinations of MMPR with either TGR or SeGR. When single drugswere given, one-half of the dose was injected between 8:30 and 9 a.m.,

and the other half between 4 and 4:30 p.m. When combinations of drugswere used, a full dose of MMPR was given in the morning, and a fulldose of either TGR or SeGR was given in the afternoon on the first day.Subsequently, half-doses of each drug were administered simultaneously

every morning and every aftenoon. All injections were given i.p.The doses of MMPR, TGR, and SeGR were selected on the basis of

previous LDM determinations in hybrids of other strains of mice and 2toxicity experiments in our BUB strain in which published protocols (4)were used. With the BUB mice, we used normal non-tumor-bearing mice

of both sexes and tested 3 or 4 doses of each individual drug and 5levels of each drug combination. One-half of each dose of each drug or

combination of drugs was injected i.p. twice daily, 8:30 a.m. and 4:30p.m., into 5 mice. In the first toxicity test, the mice were treated for 4consecutive days. In the second toxicity test, the mice were treated for5 consecutive days with TGR or M + T and for 10 consecutive dayswith SeGR, MMPR, or M + S. Survival was monitored for 30 days, andthe results were plotted on graphs to determine the appropriate LDi0 toLDgo levels. Convenient dosages just above the LD.»were selected forsubsequent analysis of their effects on hepatocellular adenomas and

1This research was supported in part by NIH Grants PHS RR07085 and PHS

CA 13943.2To whom requests for reprints should be addressed, at Division of Biology and

Medicine, Box G, Brown University, Providence, Rl 02912.Received April 11, 1983; accepted March 30,1984.

'The abbreviations used are: TGR, 6-thioguanosine; MMPR, 6-methylmercap-

topurine riboside; SeGR, 6-selenoguanosine; M -t- T, 6-methylmercaptopurineriboside in combination with 6-thioguanosine; M + S, 6-methylmercaptopurineriboside in combination with 6-selenoguanosine; LD»,50% lethal dose (dose lethalto 50% of animals; other doses are defined similarly).

2794 CANCER RESEARCH VOL. 44

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Purine Analogue Inhibition of Mitosis

duodenal epithelium. The regimens in the toxicity tests and the subsequent experiments were the same. The final doses used were MMPR(80 mg/kg/day) and TGR or SeGR (20 mg/kg/day).

Animals and Tumors. Slow-growing tumor SS1H in C3H/StW mice,transplant generations 119 to 156, and fast-growing tumor BH3 in BUB

strain mice, transplant generations 237 to 278, were used. The formerwere transplanted routinely every 4 weeks, and the latter were transplanted after approximately 10 to 12 days. The tumors were benignhepatocellular adenomas (7), and both consisted of 2 populations of cells(7). The tumors were transplanted into the axillary region by s.c. implantation of approximately 0.05 ml of minced tissue in 0.9% saline solution.The mice were kept in individual plastic cages from the time of implantation of the tumor. Laboratory chow and water were available ad libitum.Room temperatures were 72 ±2°F(22 ±1°C),and light periods were

on a 12-hr cycle. Small numbers of randomly selected and treated animals

were handled at any one time, so that the entire experiment extendedthrough all seasons of the year.

Protocols. Treatment of the tumor-bearing mice began as soon as

the implanted tumors became clearly palpable. Animals receiving SeGRand MMPR, alone or in combination, were treated twice daily for up to10 days, and tissues were sampled after 1, 3, 5, and 9 or 10 days oftreatment. After 10 days and sometimes 9 days of consecutive treatment,some but not all of the recipients appeared moribund. Those whichreceived TGR alone or in combination with MMPR were treated twicedaily for up to 5 days, and tissues were sampled after 1, 3, and 5 daysof treatment. Mice rarely survived more than 5 days of TGR administration. Control animals for each treatment and time period received 0.9%saline solution alone twice daily until they were killed. Males and femaleswere randomly assigned to each treatment group. Five animals of eachstrain were killed at each time point with each drug or combination ofdrugs. In 11 cases, only 3 or 4 tumors of 5 could be used for mitoticcounts because some tumors were too small or too necrotic. Fiveduodena could not be counted because of preparation artifacts. In somecases, 6 to 10 animals were used at one time point, especially with thenewest drug, SeGR, in order to corroborate mitotic counts.

The tumors and duodena from all animals and the livers from 2 of theanimals at each time point were fixed in Bouin's fluid, sectioned in

paraplast at 6 ^m, and stained with hematoxylin and eosin for lightmicroscopy. In sections of tumors, the numbers of mitotic figures per1000 cells were counted at a magnification of x400. The mitotic indexwas expressed as a percentage of cells in mitosis. The duodenal epithelium was selected for study of a normal replicating tissue because of itshigh predictive value (14) and because of the information available in theliterature on crypt mitotic activity (11 ). The length of the crypts (/¿m),thetotal number of cells, and the number of mitotic figures in a section ofeach crypt were recorded at a magnification of x200 for 25 crypts/animal. The crypts chosen were those cut in perfect longitudinal sections(1). On the basis of observed variations in cell number in relation to thelengths of the crypts, it appeared that, in some instances, the cells

became broader. Therefore, average cell width was calculated by dividingthe lengths (um) of the crypt by the numbers of cells lining one side of acrypt. Two samples of each of the tumors, duodena, and host liverswere also prepared for transmisison electron microscopy.

Statistical Analyses. The mitotic indices in the tumors and crypts oftreated animals exhibited wide individual variations. Similarly, cryptsvaried extensively with respect to their length, numbers of cells, and thecalculated cell widths. Therefore, before the data on mitosis were subjected to analyses of variance to test for all possible interactions or tothe Student's f test, they were transformed using either arc-sins orsquare roots (19). Student's f test or the Mann-Whitney U test was used

in some instances to compare the effect on a specific day of treatmentwith the controls for that day. The analysis of mitotic activity in theduodenal crypts was carried out on the number of cells in mitosis in 25crypts as compared to total number of cells, and then the averagenumbers of mitoses per group were converted into percentages toexpress the mitotic indices. Means of the effects of treatments on cryptcell number were plotted against the 95% confidence interval of thecontrols. For the effects on cell width in the crypts, the Kruskal-Wallis

nonparametric analysis, which does not depend on the normal distribution, was performed.

RESULTS

Mitosis in Tumors. Analysis of variance was used to comparethe 2 tumors with respect to their responses to the various drugs(Table 1, Chart 1). Because there were many zero counts amongthe treated tumors, the arc-sin transformation (19) was used. A

nested analysis of variance (21) on these transformed data firstanalyzed the overall treatments and later broke the data downto analyze days within treatments. Analysis of overall treatmentsshowed that there were no significant differences (p > 0.05) inreaction between the fast- and slow-growing hepatomas. There

fore, results with the 2 tumors will be described and discussedtogether.

Control mitotic indices ranged from 3.0 to 5.7% in hepatomaBH3 and 2.7 to 6.1% in hepatoma SS1H.

MMPR was the most effective drug for reducing tumor mitosis.At all time points, it diminished the mitotic indices to well under1% (p < 0.001). The effect 1 day after initiation of treatmentwas not significantly different (p > 0.10) from the reaction onthe third, fifth, or tenth day of treatment.

TGR showed a significant difference (p < 0.01) in mitoticactivity between treated and control tumors on each day whenits effect was analyzed by the Student's f test. On the other

hand, the analysis of variance showed no significant differencebetween treated and control tumors when all possible interac-

Table 1Mitotic indices in C3H mice

% of mitosis

MitoticindexSS1Hhepatoma8Duodenal

crypt"Day1351013510Control2.7

±0.8"5.2

±1.36.1+1.83.9±1.64.3

±0.344.8+0.484.2±0.454.1±0.25MMPR0.2±0.1C0.3

±0.3C0.7-t-0.3C0.1±0.04°4.2

±0.54.0±0.23.4±0.263.0±0.34SeGR3.8

±1.41.7±0.82.8

+0.52.4±0.94.2

±0.323.6±0.233.7±0.354.6±0.30M

+S0.5±0.1C0.2±0.1C0.7

±0.6C0.1±0.1C4.4

±0.072.8±0.364.0±0.542.6±0.51TGR7.3

±1.51.7±0.33.2±1.12.8

±0.391.7±0.16e2.5±0.33"M

+T0.5

±0.3001.8

±0.451.3±0.47*0.7

±0.27*

8 SS1H hepatoma mitotic index in slow-growing SS1H tumor after drug treatments.6 Mean ±S.E.c Probabilityof significantdifferences of treatments from the controls and the other treatments (p < 0.001). Differencesbetween days of treatment were not significant." Duodenalcrypt mitotic index determined in 25 crypts per control and drug-treated C3H mice.8Significant differences from control values (p < 0.05).

JULY 1984 2795

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

J. B. Melvin et al.

9

87eeg55

sOi

«*

3tu J(JCUJ

21I

T-T.

i•-IJIIxL_iJ,yLLADAY

135X) 135 10 135 10 13S1O 135135CONTROL

MMPR SeGR M&S TGR M&T

Chart 1. Mitotic index of hepatomas ¡srepresented graphically for the BH3hepatoma in BUB mice in controls and after drug treatments. M + S and M + Trepresent combined treatments with MMPR and SeGR and with MMPR and TGR,respectively. To simplify the chart, only one-half of the S.E. (oars) is depicted. Day10 values for TGR and M + T are absent because mice receiving daily treatmentswith TGR (20 mg/kg) never survived for 10 days. A chart for the SS1H hepatomais similar to this one and therefore is not shown, but the numerical data forhepatoma SS1H are presented in Table 1.

tions were considered. This difference in analyzed results wasdeemed likely to be due to the wide individual variations. TheMann-Whitney U test indicated that on Day 1, TGR produced asignificant (p < 0.01) increase in mitosis; on Day 3, there wasno significant difference between treated and control tumorsand, on Day 5, there was a significant inhibition of tumor mitosis(p < 0.05).

SeGR reduced the number of mitotic figures in hepatoma BH3only on Day 10, and this was not a significant reduction. TheMann-Whitney U test of combined data for the 2 tumors for eachday compared with combined controls for that day failed todemonstrate any significant action by this drug in reducing tumormitosis.

M + T, the combination of MMPR and TGR, produced thesame inhibition as did MMPR alone on Day 1. Subsequently,however, after 3 and 5 days of M + T treatment, no mitoticfigures were found.

M + S, the combination of MMPR and SeGR, produced thesame inhibition of tumor mitosis as did MMPR alone at all timepoints.

Mitosis in Duodenal Crypts. As in the tumors above, themitotic counts of the duodenal crypts were not distributed normally and, therefore, the data were transformed into squareroots (19) for an analysis of variance (Table 1, Chart 2). Nosignificant difference in reaction to the drugs was found betweenthe 2 strains of mice, so they will be described and discussedtogether.

Control mitotic indices varied from 3.2 to 4.0% in the cryptsof BUB mice and from 4.1 to 4.8% in those of C3H mice.

MMPR-treated animals exhibited ranges of 3.0 to 3.8% and3.0 to 4.2% in BUB and C3H mice, respectively. In spite of agradual reduction in the number of mitotic figures with prolongedMMPR administration, the differences between the control andtreated mice could not be substantiated as significant at the 95%confidence level. The differences between successive days oftreatment also were not significant (p > 0.05).

TGR was more effective in reducing mitotic activity in duodenal

crypts than was MMPR. One day after initiation of TGR treatment, the crypt mitotic indices of 2.8% in both strains were notsignificantly different from the Day 1 controls (p > 0.05), but theindices were 2.0 and 1.7% on Day 3 and 2.2 and 2.5% on Day5. The difference between the 2 latter time points was notsignificant (p > 0.05), but the difference of each from theirrespective Day 3 and Day 5 controls was highly significant (p <0.001).

SeGR alone had no significant effect on crypt mitosis.M + T produced crypt mitotic activities with means of 3.3 and

1.8% on Day 1, 0.9 and 1.3% on Day 3, and 0.3 and 0.7% onDay 5. The reduction on Day 3 and Day 5 is significant (p <0.05).

M + S induced a significant diminution in crypt mitotic activityonly after 10 days of treatment in one strain of mice.

Crypt Cell Number. Analysisof varianceagain showed nosignificant difference between the 2 strains of mice (Table 2).

Controls exhibited means of 33.2 to 36.2 cells in longitudinalsections of the crypts in BUB mice and 33.3 to 34.1 in C3H mice.

On Day 1, there was no significant difference between themean of any treatment and the mean of its control in either strainof mice, as detemined by plotting means of treatments againstthe 95% confidence level of controls.

MMPR-treated animals showed no significant differences fromtheir respective controls, with the single exception of Day 3 inC3H mice, when the mean of the treated animals (27.5 ±1.3)fell just outside the confidence interval of their controls (33.8 ±2.5).

In TGR-treated mice, the reduction in number of cells wassignificant (p < 0.05) in both strains after Day 3 (24.1 and 22.3as compared to control levels of 36.2 and 33.8 in BUB and C3Hmice, respectively) and after Day 5 (21.5 and 26.1 as comparedto 33.3 and 33.9, respectively). On Days 3 and 5, the cells werewider and shorter than normal and often had pyknotic nuclei.Mitotic figures were aberrant. Cellular debris was present in thedistended lumens of the crypts.

SeGR alone or in combination with MMPR had no effect on

40 •

31M«

K՜

20OIO

CM15s

IQo10I5T

TJTJ|IJL1lL]\il11LLl1l135

10 135 10 135 1O 135 1O 135135CONTROL

MMPR SeGR M&S TGR M&T

Chart 2. Mitotic index of dudodenal crypt cells is shown in control and drug-treated BUB hosts of hepatoma BH3. One-half of the S.E. (bara) is illustrated.There are no Day 10 values for TGR and M + T because these treatments do notpermit survival beyond 5 or 6 days. A similar chart for C3H mice bearinghepatomaSS1H is not shown, but numerical data for the duodenal crypt cells of this strainare presented in Table 1.

2796 CANCER RESEARCH VOL. 44

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Purine Analogue Inhibition of Mitosis

Table 2

Crypt cell numberNo.

ofcellsStrainBUBC3HDay1

351013

510Control34.6

±1.2a

36.2 ±0.433.3 +1.336.2+1.834.1

±0.433.8 ±0.433.9 ±1.433.3

+ 0.8MMPR33.0

±33.7 ±29.9+29.6+31

.5 ±27.5 ±32.8±29.7

±0.2

1.42.60.41.6

1.31.10.9SeGR32.5

±31.6 +35.4±33.5±33.9

±34.6 ±31.9+33.0

±0.4

0.71.41.80.6

0.71.10.5M

+S33.0

±1.331.1 ±0.430.4 ±2.736.1±31

.2 ±31.1 ±33.9+31

.5 ±3.41.8

0.50.51.0TGR32.2

±1.124.1 + 2.221.5 ±4.8o31.1

±2.122.3 ±0.7*26.1 ±2.2"M

+T32.3

+20.9 ±5.7+30.5

+23.2 ±1.0

1.5"0.8°1.9

2.5"

" Mean ±S.E. of the number of cells lining a vertical section of the crypt after drug treatment.6 Significant values falling outside the 95% confidence interval of controls (p < 0.05) occur only with TGR and M + T

treatments.

crypt cell number.In M + T-treated animals, the reduction in total crypt cell

number was significantly (p < 0.05) greater than with eitherMMPR or TGR alone on Days 3 and 5, but it was not significantlydifferent from either drug on Day 1 or from the control value (p> 0.05). Evidence of cellular destruction was similar to that inanimals treated with TGR alone.

Crypt Length. There was no statistically significant (p > 0.05)change in the length or depth of the duodenal crypts in eitherstrain or with any treatment during the 5 or 10 days of theexperiments. This stability was maintained even with TGR orM + T treatment in which there was a significant reduction in thenumber of cells lining the crypts.

The range of crypt lengths (/¿m)of BUB mice was as follows:controls, 123 ±6 to 138 ±6; MMPR, 126 ±6 to 132 ±9; TGR,126 ±6 to 138 ±9; SeGR, 126 ±9 to 138 ±9; M + T, 87 ±9to 135 ±6; and M + S, 126 ±6 to 171 ±9. In the C3H mice,the data were as follows: controls, 120 ±9 to 135 ±3;MMPR,105 ±3 to 144 ±9; TGR, 126 ±3 to 159 ±12; SeGR, 117 ±3 to 129 ±3; M + T, 108 ±24 to 138 ±15; and M + S, 123 ±6 to 132 ±3. Calculations of crypt cell widths based on cryptlength and the number of cells lining one side of the cryptrevealed significant increases in cell width after treatment withTGR or M + T for 1, 3, or 5 days (p < 0.05) and no significantchange with MMPR, SeGR, or M + S.

Ultrastructure. No subcellular alterations were detected in anyof the nonreplicating normal cells of the host liver. Alterations intumor cells in treated hosts were nonspecific changes associatedwith cell degeneration. Similarly, there were changes in theduodenal epithelium which are nonspecific indications of injury.

DISCUSSION

The results of this study show that MMPR was, alone, themost promising drug investigated. It effectively reduced mitoticactivity in both the slow-growing and the fast-growing hepato-

mas, and this was as effective on the first day as after the tenthday of treatment. Moreover, MMPR produced a minimal andnonsignificant reduction in mitotic activity in the duodenal crypts.The action of MMPR on other parameters of the duodenal cryptwas also mild.

It has been a common assumption that chemotherapy is oflittle effect against hepatomas (17), and some investigators (20)have felt that MMPR is only a weak chemotherapeutic drug. Yetagainst both solid tumors in this study, MMPR, which is generallyused to treat only acute leukemias (17), all but eradicated hepa-

toma mitosis 24 hr after the administration of one-half of an LD50dose followed by another half-dose 6 to 8 hr later. This demonstrated that a twice-daily schedule of treatment permitted the

effective retention or recycling of MMPR and/or its active metabolites (15,16) in these hepatomas. It has been shown, in mouseEhrlich ascites carcinoma cells, that the enhancement of othertherapeutic drugs by a single dose of MMPR is maximal after 6to 12 hr and is still effective after 96 hr (16).

TGR was ineffective against these tumors. Furthermore, itwas very damaging to the epithelial lining of the intestine. WhenTGR and MMPR were given together, tumor mitosis was inhibited completely. The cessation of mitosis in the tumors couldhave been due entirely to MMPR, since TGR alone had littleeffect. The crypts showed the same pattern of cell loss as withTGR alone, and crypt damage could have been due to TRGalone. The principal target of TGR that correlates with its cyto-toxicity is DMA (8-10). It has been suggested (6) that the specific

toxicity of TGR for bone marrow cells is due to the fact that theydo not possess an enzyme to deaminate it as do other rapidlydividing cells. One might speculate that the hepatoma cellsstudied here may possess such an enzyme which would accountfor their relative insensitivity to TGR.

The effectiveness of TGR was demonstrated in Sarcoma 180ascites cells and the L1210 and L5178Y mouse lymphoma cellsboth in vitro and when injected i.p. into mice (15). In similarstudies, SeGR was shown to be as effective as was TGR againsts.c. injections of L1210 lymphomas, L5178Y lymphomas, andSarcoma 180 (2, 12, 18, 20), and it was less toxic than wasTGR. On the other hand, SeGR proved to have no toxic effectswhatsoever against the 2 solid tumors of hepatocyte originstudied here and also did not affect mitosis in the intestine. Themechanism of action of SeGR has not yet been elucidatedcompletely. Our results with SeGR are in agreement with Griffin's

conclusions in a recent review (5) that, although initial resultsagainst leukemias and lymphomas were promising, the variousselenium analogues have not been shown to be effective in thetreatment of solid tumors. Selenium in excess is toxic to theorganism (5), but the lack of harmful effects in our animals whichreceived high doses of SeGR may be due to the fact that organicselenium compounds are less toxic, even at high levels, than arethe inorganic compounds (5).

When SeGR was given in conjunction with MMPR (M + S),their combined action on both the tumors and the intestine wasessentially the same as that of MMPR alone. Therefore, thereappeared to be no synergistic activity in the combination ofMMPR and SeGR. The variation was too great for us to pick up

JULY 1984 2797

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

J. B. Melvin et al.

small differences between the effects of MMPR and M + T thatwould indicate an additive effect.

The data presented in this study, like those from many biological systems, frequently did not have normal distributions orsimilar variances. Indeed, one of the effects of continued drugtreatment of the animals was to increase the variances. Forinstance, in some of the most severely injured duodena, therewere a few nearly normal crypts. Within treatments, also, someindividual animals retained nearly normal parameters, while others in the same group exhibited more extensive effects. It is forthese reasons that data which were subjected to parametricanalysis had to be transformed (19) for the analysis of varianceor Student's f test and that, with some data, nonparametric

methods proved to be more efficient.We are not certain why the mitotic indices of the fast (BH3)

and slow (SS1H) hepatomas are so similar, but the work ofEchave Llanos and Nash (3) is relevant. We supplied them withSS1H and a different (SS1K) slow hepatoma which they analyzedfor mitotic circadian rhythms. They found that, although the fasttumor had a significantly higher mitotic index during the darkcycle when mitotic activity was low, the 2 tumors had identicalmitotic indices during the light cycle when mitotic activity washigh. All of our animals were sacrificed during the light cycle.

In both hepatomas, BH3 (Chart 1) and SS1H (Table 1), therewas a daily variation in the mitotic indices of controls the hostsof which received injections of saline alone without any of thedrugs. We are not certain of the basis of this variability, but wesuggest that it may be the result of our randomized counting ofnonrandomly distributed mitotic figures, because it is often evident that the figures occur in clusters. To increase the numberof animals per group would not alter the variability and, althoughit would reduce the S.E. and thereby increase the reliability ofour figures, we decided that it would not be cost effective.Whereas this variability in mitotic indices among untreated tumors might raise doubts about the validity of our interpretationsof the effects of TGR alone or SeGR alone, the inhibitory effectof MMPR alone or in combination with either of the other 2 drugsis so pronounced that its validity cannot be questioned on thebasis of control variability.

The data presented in this paper point to a need for furtherwork in understanding the metabolic pathways of mouse hepatomas to reveal why MMPR was toxic to the tumors and relativelyharmless to proliferating intestinal epithelium. In addition, dosageschedules of MMPR administration against hepatomas in vivomight be developed which would contribute significantly to longersurvival of the hosts.

ACKNOWLEDGMENTS

We greatly appreciate the excellent assistance of Myrtle Lawton and JeanWaage. We thank Dr. Ming Chu, Brown University, for her advice and assistance.

REFERENCES

1. Carriere, R. M. The influence of thyroid and testicular hormones on theepithelium of crypts of Lieberkuhn in the rat's intestine. Anat. Ree., 156: 423-

432,1966.2. Chu, S-H. Potential antitumor agents. Selenoguanosine and related com

pounds. J. Med. Chem., 14: 254-255, 1971.3. Echave Llanos, J. M., and Nash, R. E. Mitotic circadian rhythm in a fast-

growing and a slow-growing hepatoma: mitotic rhythm in hepatomas. J. Nati.Cancer Inst., 44: 581-585,1970.

4. Geran, R. I., Greenberg, N. H., MacDonald, M. M., Schumacher, A. M., andAbbott, B. J. Protocols for screening chemical agents and natural productsagainst animal tumors and other biological systems. Cancer Chemother RepPartili, 3: 1-103, 1972.

5. Griffin, A. C. Role of selenium in the chemoprevention of cancer Adv CancerRes., 29; 419-442, 1979.

6. Henderson, J. F. Biochemical aspects of selective toxicity. In: A. C. Sartorelliand D. G. Johns (eds.), Antineoplastic and Immunosuppressive Agents, Part I,p. 343. New York: Springer-Vertag, 1974.

7. Leduc, E. L., Malick, L. E., and Holden, H. E., Jr. Light microscopic observations of transplantable mouse hepatomas. Cancer Res., 34.: 716-721,1974.

8. LePage, G. A. Basic biochemical effects and mechanism of action of 6-thioguanine. Cancer Res., 23: 1202-1206, 1963.

9. LePage, G. A. New purine analogues. In: P. Buculoss and N. Cascinelli (eds.).Proceedings of the Eleventh Cancer Congress, Florence, Italy, 1974, Abstract,Book 5, pp. 248-252. Amsterdam: Exerpta Medica, 1975.

10. LePage, G. A., and Gottlieb, J. A. Deoxythioguanosine and thioguanine. Clin.Pharmacol. Ther., 9: 966-969, 1973.

11. Upkin, M. Proliferation and differentiation of gastrointestinal cells. Physiol.Rev., 53:891-915. 1973.

12. Mauntner, H. G., Chu, S-H., Jaffe, J. J., and Sartorelli, A. C. The synthesisand antineoplastic properties of selenoguanine, selenocytosine, and relatedcompounds. J. Med. Chem., 6: 36-39, 1963.

13. Nelson, J. A., and Parks, R. E., Jr. Biochemical mechanisms for the synergismbetween 6-thioguanine and 6-<methylmercapto)purine ribonucleoside in Sarcoma 180 cells. Cancer Res., 32: 2034-2041,1972.

14. Owens, A. H. Predicting anticancer drug effects in man from laboratory animalstudies. J. Chronic Dis., 15: 223-228,1962.

15. Paterson, A. R. P., and Tidd, D. M. 6-thiopurines. In: A. C. Sartorelli and D. G.Johns (eds.), Antineoplastic and Immunosuppressive Agents, Part III, pp. 384-403. New York: Springer-Verlag, 1975.

16. Paterson, A. R. P., and Wang, M. C. Mechanism of the growth inhibitionpotentiation arising from combination of 6-mercaptopurine with 6-<methylmer-capto)purine ribonucleoside. Cancer Res., 30; 2379-2387,1970.

17. Pratt, W. B., and Ruddon, R. W. The Anticancer Drugs, pp. 12-19 and 113-122. New York: Oxford University Press, 1979.

18. Ross, A. F., Agarwal, K. C., Chu, S-H., and Parks, R. E., Jr. Studies on thebiochemical actions of 6-selenoguanine and 6-selenoguanosine. Biochem.Pharmacol., 22. 141-154,1973.

19. Schefler, W. C. Statistics for the Biological Sciences, pp. 131-132,214-216,and 221-223. Reading, MA: Addison-Westey Publishing Co., 1979.

20. Scholar, E. M., Brown, P. R., and Parks, R. E., Jr. Synergistic effect of 6-mercaptopurine and 6-methylmercaptopurine ribonucleoside on the levels ofadenine nucleotides of Sarcoma 180 cells. Cancer Res., 32: 259-269, 1972.

21. Snedecor, G. W. Statistical Methods, pp. 291-393. Ames, IA: Iowa StateUniversity Press, 1956.

2798 CANCER RESEARCH VOL. 44

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1984;44:2794-2798. Cancer Res   Joan B. Melvin, Thomas H. Haight and Elizabeth H. Leduc  in MiceTwo Transplantable Hepatomas and on Normal Renewing Cells Cytological Effects of Sulfur and Selenium Purine Analogues on

  Updated version

  http://cancerres.aacrjournals.org/content/44/7/2794

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/44/7/2794To request permission to re-use all or part of this article, use this link

on May 20, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from