Vitamin D Analog, EB1089, Inhibits Growth Subcutaneous Xenografts of the Human Colon Cancer Cell Line, LoVo, in a Nude Mouse Model Javed Akhter, M.Sc., Xianghai Chen, M.D., Patricia Bowrey, B.Sc., Elaine J. Bolton, Ph.D., David L. Morris, Ph.D. From the University o f New South Wales, Department o f Surgery, St. George Hospital, Sydney, Australia

PURPOSE: In this study, we investigated the effect of the vitamin D 3 analog, EB1089, on the growth of subcutaneous xenografts of the human colon cancer cell line, LoVo, in a nude mouse model. METHODS: BALB/c Nu/Nu nude mice were inoculated subcutaneously with 10 6 LoVo cells. EB1089 dissolved in isopropanol was administered intra- peritoneally and orally on alternate days at doses of 0.1, 0.5, and 2.5 /,g/kg/day. Control animals received isopropanol alone. Tumor volumes estimated using the formula 0.5 • length X (width). 2 The tumor kinetic index was determined by immunohistochemical detection of proliferating cell nu- clear antigen. RESULTS: Significant dose-dependent inhibi- tion of tumor growth was seen. After 20 days of treatment with 0.1 /~g/kg/day EB1089, mean tumor volume in treated mice was 41 to 49 percent less than that in control animals (P < 0.01). Significant inhibition of tumor growth was also seen with 0.5/~g/kg/day EB1089 after 22 days of treatment (51 percent of control P < 0.01). Treatment with 2.5 /,g/kg/day resulted in weight loss that required termination of this group; these mice were subsequently found to be hypercalcemic. The tumor kinetic index was significantly lower in tumors treated with 0.1 /,g/kg/day EB1089 com- pared with that for control tumors (8 vs. 30 percent in controls). CONCLUSION: These findings suggest that the vitamin D 3 analog, EB1089, is a potent antiproliferative agent for some human colon cancers. [Key words: Vitamin D; Proliferation; Colon' carcinoma; Vitamin D 3 analogs]

Akhter J, Chen X, Bowrey P, Bolton EJ, Morris DL. Vitamin D 3 analog, EB1089, inhibits growth of subcutaneous xeno- grafts of the human colon cancer cell line, LoVo, in a nude mouse model. Dis Colon Rectum 1997;40:317-321.

T he antiproliferative act ion of 1,25-(OH) 2 D 3 in

cancer was first descr ibed in 1981.1 Vitamin D 3

has subsequen t ly b e e n s h o w n to inhibi t growth of

osteosarcoma, breast carcinoma, and colonic carci-

n o m a cell lines. 2-4 In addit ion, effects on cellular

o n c o g e n e t ranscr ipt ion and growth factor receptor

express ion have b e e n demons t ra t ed in vitro. 5

The human colonic adenocarcinoma cancer cell line, LoVo, was a gift from CRC Laboratories, Nottingham, United Kingdom. Read at the Royal Australasian College of Surgeons Annual Scien- tific Congress, Melbourne, Australia, June 5 to 10, 1996. Address reprint requests to Dr. Morris: the University of New South Wales, Department of Surgery, The St. George Hospital, Kogarah, New South Wales 2217, Australia.

Studies in vivo have s h o w n regression of xeno-

grafts of colonic t u m o r s a nd m e l a n o m a s w h e n ani-

mals were treated with large doses of 1,25-(OH) 2 D 3,

a l though animals were ma in t a ined on a low calcium

diet to reduce toxicity from hypercalcemia. 6 Clinical

use of v i tamin D 3 metabol i tes as therapeut ic agents

has b e e n l imited because of hyperca lcemia at doses

of more than a few micrograms per day. Several

synthetic v i tamin D 3 analogs have b e e n deve loped ,

which are active in p romot ing cellular differentiat ion

a nd inhibi t ing cell growth but with r educed calcemic

activity. 7 EB1089, a n e w vi tamin D analog, has already

b e e n f ound to be more effective in inhibi t ing the in

vitro growth of the h u m a n breast cancer cell line,

MCF-7, than 1,25-(OH) 2 D3. s

We have previous ly repor ted the antiproliferative

effect of EB1089 o n in vitro growth of the h u m a n

co lon cancer cell line, LoVo. 9 Here we report the first

in vivo exper imen t with EB1089 us ing a h u m a n co lon

cancer cell xenograf l in a n u d e mouse model .

M A T E R I A L S A N D M E T H O D S

C e l l C u l t u r e

Cells were g r ow n as mono laye r s in RPMI 1640

m e d i u m plus 10 percent fetal calf se rum (Gibco, Syd-

ney, Australia) at 37~ in 5 pe rcen t CO 2.

C e l l C o u n t i n g

Cells were trypsinized, diluted, a nd c oun t e d in trip-

licate in a N e u b a u e r hemocytometer . Viability of cells

was ---90 percent , as de t e rmined by t rypan b lue ex-

clusion.

T r e a t m e n t o f T u m o r - B e a r i n g M i c e

Six-week-old to t en -week-o ld male BALB/c N u / Nu

mice (ABHU, Universi ty New South Wales, Australia)

317

318 AKHTER ETAL

were inoculated subcutaneously with 106 LoVo cells into the right flank. Animals received sterile food and water ad libitum. Mice were randomly assigned to the control group (n = 10) or the treatment group (n = 10). Twenty-four hours after tumor inoculation, treat- ment animals received EB1089 daily. This was admin- istered intraperitoneally and orally on alternate days (to minimize the number of intraperitoneal injections) at a dose of 0.1 / ,g/kg/day (in isopropanol) for 22 days. Control animals received isopropanol alone. In the second experiment, mice received 0.1, 0.5, and 2 .5 / ,g /kg /day EB1089 or isopropanol alone. Tumor size was measured using vernier calipers every three days, and volumes were estimated using the formula 0.5 X l eng th • (width)Z. 1~ At the conclusion of the

experiment, mice were injected intraperitoneally with 330 mg/ml of pentobarbitone sodium, and the tumors were excised. Serum was pooled, and calcium con- centrations were measured by the cucresolphthalein complex one method (Boehringer Mannheim Auto- mated Analysis, Hitachi System 717, Boehringer

Mannheim, Australia).

Tumor Kinetic Index Measurement

At the end of the experiment, tumors were har- vested and fixed in 10 percent formalin and then embedded in paraffin. Four-/zm sections were cut and mounted onto 0.01 percent poly-L-lysine-coated slides. After deparaffinization with xylene and rehy- dration with graded ethanol to 0.05 M Tris(hydroxym- ethyl)aminomethane hydrochloride (Tris/HCl; pH 7.6), slides were immersed in 10 mM citrate buffer (pH 6) and incubated in a 100~ water bath for 30 minutes. Endogenous peroxidase activity was blocked using 0.06 percent hydrogen peroxide in methanol at room temperature for ten minutes; after washing with Tris/HC1, nonspecific binding was blocked by incubation for 30 minutes with normal rabbit serum. Excess serum was drained, and slides were incubated with mouse monoclonal antibody against proliferating cell nuclear antigen (PCNA; No- vacastra Laboratories, Newcastle-on-Tyne, United Kingdom) at a concentration of 1:100 for 60 minutes at room temperature. Nonspecific IgG2a (1:25; Silen- ius Laboratories, Hawthorne, Vic, Australia) was sub- stituted for the primary antibody in negative controls. Sections were then incubated with biotinylated anti- mouse immunoglobulins for 30 minutes (1:300; Dako Corp., Carpinteria, CA), followed by avidin-biotin complex for 30 minutes (Vectorstain ABC kit PK4000,

Dis Colon Rectum, March 1997

Vector Laboratories Inc., Burlingame, CA). After each step, sections were washed in Tris/HC1. The reaction product was visualized using a substrate solution con- taining 0.05 percent 3,3-diaminobenzidine tetrahydro- chloride (Dako Corp.), hydrogen peroxide, and imi- dazole in Tris/HC1 (pH 7.6) with a reaction time of 7 minutes. After washing, slides were counterstained lightly with hematoxylin, dehydrated, and mounted in Depex mounting medium (Sigma, St. Louis, MO).

The percentage of nuclei stained with PCNA in each tissue section was calculated using an image analysis program (Donaldson Imaging Pty. Ltd, Syd- ney, Australia) on a Bioscan Optimas Image Analyser (Bioscan Inc., Washington, DC) as we reported pre- viously. 11 Maximally, ten staining fields were exam- ined from each tumor section. The total nuclear area

and nuclear area with positive PCNA staining were m a p p e d using gray value threshold settings, and the percentage of nuclear area stained with PCNA was calculated for each field. For each tumor section, the mean of the six highest percentages was recorded as the maximum proliferative index (MPI). n All sections were evaluated without knowledge of treatment

group results and by the same observer.

Carcinoembryonic Antigen (CEA) Measurement

Carcinoembryonic antigen was measured using an immunoradiometric assay kit (CIS Bio International, Cedex, France).

S ta t i s t i ca l Methods

Data were analyzed by one-way analysis of vari- ance (ANOVA).

RESULTS

Suppression o f I n V ivo G r o w t h o f a

Human Colonic Solid Tumor X e n o g r a f t

by EB 1089 EB1089 (0.1 /xg/kg/day) was administered daily to

mice inoculated with LoVo colon cancer cells, and tumor size was measured for 20 days (Fig. 1). Signif- icant inhibition of tumor growth was observed after 14 days of treatment with EB1089 (P < 0.05). After 20 days, tumors in EB1089-treated mice were 49 +- 10.7 percent smaller than tumors in control animals (P < 0.01 ANOVA). Serum calcium concentration and body weight in the treated group did not differ significantly from those of control mice.

Vol. 40, No. 3 VITAMIN D 3 ANALOG AND COLON CANCER 319

2000

~ 1750 lu

1500

~ 1250

E E 1000

E 750

500 E

250

OL. 5

Control

10 15 20 25

Days Postinjection

Figure 1. Inhibition of LoVo tumor growth in nude mice by the vitamin D 3 analog, EB1089. Treatments are as de- scribed under MATERIALS AND METHODS. Results are mean _+ standard error of the mean (SEM; n = 10). EB1089 caused significant inhibition of tumor growth compared with controls, P < 0.01 (analysis of variance).

1400

1200

+l 1000

~" 800

.t.. r 600

'~ 400 o

~ 200

o

�9 Control -" 2.5 ug/kg/day p<O.O1 �9 0.5 ug/kg/day p<O.O1 I

k t -

F I

r

m I I I

10 15 20 215

(P < 0.01, ANOVA). Inhibition was evident by day 14

and continued for the entire experiment. Maximum

inhibition was achieved with 0.1 /xg/kg/day; by day

22, tumor volumes were 41 percent of those in con-

trois. Mice that received 2 .5/xg/kg/day EB1089 lost 20

percent of their initial body weight after 12 days

treatment. Therefore, these mice were killed. Mea-

surement of pooled serum calcium levels from this

group indicated that these mice had become hyper- calcemic (Table 1). There was no difference in either

body weight or serum calcium levels be tween con-

trois and mice in the two lower EB1089 dose groups.

Maximum Proliferative I n d e x

A tumor kinetic index, referred to as the MPI, was

then determined from tumor tissue by immunohisto-

chemical detection of PCNA (Table 2). In the first

experiment, the MPI was significantly less in tumors

from mice treated with EB1089 (0.1 / ,g /kg/day) than that from control mice. In the second experiment, in

which mice were treated with one of three different

doses of EB1089, the MPI tended to reflect the relative

degree of inhibition. The MPI was approximately 20 percent in the 0.1 / ,g /kg/day group and approxi-

mately 26 percent in the 0.5 /~g/kg/day group com-

pared with approximately 27 percent in controls. M- though terminated early, there was also a lower MPI

in the 2.5 /~g/kg/day EB1089 group (18 percen0.

Serum Carcinoembryonic Antigen

We wanted to establish whether measurement of serum CEA would provide a good indication of inhi-

bition of tumor growth caused by EB1089. Serum CEA

was lower in the EB 1089-treated animals, especially in

the 2 .5 / ,g /kg /day group that were killed early (Table

1). It was not possible to apply statistical tests to this

Days Postinjectlon

Figure 2. Dose-response study of inhibition of LoVo tu- mor growth in nude mice by vitamin D a analog, EB1089. Treatments are as described under MATERIALS AND METHODS. Results are the mean + standard error of the mean (SEM; n = 7 EB1089 groups; n = 10 control).

Dose dependence of the inhibition by EB1089 was then tested. Mice received 0.1, 0.5, or 2 .5 /xg/kg/day

EB1089 for the duration of the experiment, starting 24 hours after tumor cell inoculation (Fig. 2). There was significant inhibition of tumor growth with both 0.1 /xg/kg/day (P < 0.001, ANOVA) and 0.5 /xg/kg/day

Table 1. Measurement of Serum CEA and Calcium Levels in

Mice Treated with Different Doses of EB1089

Tumor CEA, Serum Dose Weight CEA, /~g/g of Calcium

(Mean, g) /~g/ml Tumor Tissue (mmol/I)

Control 9.8 33.16 3.38 2.23 0.1 /~g/kg/day 5.1 14.98 2.93 2.36 0.5/~g/kg/day 6.4 23.34 3.66 2.31 2.5/~g/kg/day* 0.6 2.29 3.82 3.11

CEA = carcinoembryonic antigen. * Observation from day 12 because of premature ter-

mination of experiment.

320 AKHTER ETAL Dis Colon Rectum, March 1997

Tab le 2. PCNATumor Kinetic Index (%MPI _ Standard Error of the Mean) of Tumors from Mice Treated with Different

Doses of EB1089

Doses PCNA (%MPI)

Experiment 1 Control (n = 10) Treatment (n = 10) 0.1 /~g/kg/day

Experiment 2 Control (n = 10) Treatment 0.1 / ,g/kg/day (n = 7) 0.5/~g/kg/day (n = 7) 2.5/~g/kg/day (n = 7)

30.4 _+ 4.5

8.03 _+ 6.0*

26.94 + 1.4

20.42 _+ 1.7" 25.75 _+ 1.4

18.1 _+ 1.4"

PCNA = proliferating cell nuclear antigen; MPI = max- imum proliferative index.

* P < 0.05 vs. control.

data because it was derived from pooled serum.

When CEA levels were expressed per gram of tumor

tissue, there was no difference between the control

and treated groups. Therefore, EB1089 inhibits tumor

growth not CEA secretion, and CEA could, therefore,

be used as a marker of response to EB1089.

D I S C U S S I O N

This is the first report of the use of the vitamin D 3

analog EB1089 in an in vivo xenograft model of colon

cancer. Vitamin D 3 or its analogs have been reported

to inhibit growth of tumors in other systems. The

active form of the parent compound, 1,25-dihy-

droxTvitamin D3, has been shown to inhibit growth of

human colon cancer (COLO 206F) xenografts in nude

mice, although it was necessary to maintain mice on a

low-calcium diet to minimize toxicity. 6 Intratumor or

oral administration of the vitamin D 3 analog, 22-oxa-

1,25-(OH) 2 D3, produced approximately 50 percent

inhibition of growth of the human breast cancer cell

line, MX-1, in nude mice. 12 Administration of the

vitamin D 3 analog, 1,25-dihydroxy-16-ene-23-yne-

cholecalciferol, to nude mice slowed the growth of PC-3 prostate cancer xenografts. ~3 In the N-methyl-

nitrosourea-induced rat mammary tumor model, 0.5

/ ,g /kg/day EB1089 caused a significant inhibition of

tumor growth in the absence of hypercalcemia. 8 The

same concentration of 1,25-dihydroxyvitamin D 3 had

no effect on tumor growth. As reported here, higher doses of EB1089 also inhibited tumor growth but

raised serum calcium levels as well. 8 We have previously reported that growth of the

human colorectal cancer cell line, LoVo, was pro-

foundly inhibited in vitro by the vitamin D 3 analog,

EB1089. 9 It has been reported that inhibition of pro-

liferation by vitamin D 3 and its analogs is related to the levels of receptor expressed on the cells. 13' 14

Vitamin D 3 receptors have been demonstrated in a

number of colon cancer cell lines and primary tu- mors, 15-a7 including the LoVo cell line.18 Although we

do not know if growth of all cells that express the

vitamin D 3 receptor would be inhibited by EB1089, it

is probable that receptor expression is required for

response. Since our earlier report of the effects of

EB1089 on the growth of several human colon cancer

cell lines in vitro, 9 we have studied many more cell

lines (Akhter J, Chen X, Bowrey P, et al., unpublished

data), and it is clear that not all cell lines respond to

EB1089. Measurement of Vitamin D 3 receptor expres-

sion may, therefore, enable detection of the tumors

most likely to respond to treatment with EB1089.

There is some knowledge of the effects of vitamin D 3 compounds on cell cycle �9 19 progression. When

used at high but nontoxic concentrations (10 -7 M),

1,25-dihydroxyvitamin D 3 causes a substantial reduc-

tion in the percentage of S phase ceils as well as an

increase in G 2 and M cells. However, at lower con-

centrations, there is still an increase in 0 2 and M cells

but little effect on S phase cells. In our first experi-

ment, the tumor MPI following treatment with EB1089

was significantly lower than in control tumors. When

different doses of EB1089 were studied, MPI reflected

the relative inhibitory activity of each dose, although

we did not achieve the same degree of inhibition of

the MPI with 0 .1 / ,g /kg/day EB1089 as was seen in the

first experiment. The reason for this was not clear; it

may be related to the route or time of administration

of EB1089 on the previous day, because PCNA pre-

dominantly stains cells in the S phase and, at least in

vitro, the decline in percentage of S phase ceils by

1,25-dihydroxTvitamin D 3 is dependent on concentra- tion and time after administration. 19

Serum CEA levels could not be compared statisti-

cally because blood was pooled; although there was a large difference between treated and control mice,

there was no significant difference when corrected for

tumor weight. This indicates that the effect of EB1089

on serum CEA is attributable to its effect on tumor

growth rather than on CEA expression or secretion.

Therefore, CEA levels might provide a convenient

indicator of response to EB1089 in a clinical situation. Although we observed good inhibition of tumor

growth following administration of EB1089 in the absence of hypercalcemia, the highest dose (2.5 /*g/

Vol. 40, No. 3 VITAMIN D 3 ANALOG AND COLON CANCER 321

mg/day) did raise serum calcium levels as reported by

others. 3 Doses were chosen imperically; however, these s ame doses we re tes ted in a N-methyl-ni t ro-

s o u r e a - i n d u c e d rat m a m m a r y tumor. 8 W e d o no t

k n o w w h a t s e rum concen t ra t ions w e r e a c h i e v e d wi th

this ana log , and, certainly, these doses w o u l d no t b e

e x p e c t e d to b e in the phys io log ic range. Our m o d e l

looks at the s u p p r e s s i o n o f g rowth of s u b c u t a n e o u s

xenograf t s ra ther than the e t io logy o f co lon cance r or

the metas ta t ic p rocess . It m a y b e that a l ow ca lc ium

diet wil l pe rmi t use of a h ighe r concen t ra t ion o f

EB1089, w h i c h m a y p r o d u c e grea ter inh ib i t ion of tu-

m o r growth. W e also p l an to s tudy the use of EB1089

in a d v a n c e d and metas ta t ic t u m o r models .

In the p r e sen t s tudy, w e have s h o w n that a syn-

thet ic ana log o f 1,25-(OH) 2 D3, EB1089, p r o d u c e d a

large and signif icant s u p p r e s s i o n o f i n v ivo g r o w t h of

a h u m a n co lon ic t u m o r xenogra f t wi th little effect on

s e r u m ca lc ium levels. These s tudies p r o v i d e further

ev idence that EB1089 m a y b e a useful t he rapeu t i c

agen t for the t r ea tmen t of h u m a n co lon cancer .

ACKNOWLEDGMENT

The au thors t hank Leo Pharmaceu t i ca l Products ,

Ba l l emp , Denmark , for supp ly ing the EB1089.

REFERENCES

1. Abe E, Miyaura C, Sakagami H, et aL Differentiation of mouse myelo id leukemia cells induced by 1,25- dihydroxyvitamin D 3. Proc Natl Acad Sci U S A 1981; 76:4990~4.

2. Niendor f A, Arps H, Dietel M. Effect of 1,25- dihydroxyvitamin D 3 on human cancer ceils in vitro. J

Steriod Biochem 1987;27:825-8.

3. Colston KW, Berger U, Coombes RC. Possible role for vitamin D in controlling breast cancer proliferation. Lancet 1989;1:188-91.

4. Frampton RJ, Osmond SA, Eisman JA. Inhibition of human cancer cell growth by 1,25-Dihydroxyvitamin D 3 metabolites. Cancer Res 1983;43:4443-7.

5. Minghetti PP, Norman AW. 1,25-(OH)a-vitamin D 3 re- ceptor: gene regulation and genetic circuitry. FASEB J 1988;2:3043-53.

6. EismanJA, Barkla DH, Tutton PJ. Suppression of in vivo

growth of human cancer solid tumor xenograft by 1,25- Dihydroxyvitamin D 3, Cancer Res 1987;47:21-5.

7. Katsuyuki H, Shinomiya H. Novel vitamin D 3 analogs and their potential for inhibiting cancer ceil growth. Drugs Future 1993;18:1057-61.

8. Colston KW, Mackay AG, James SY, Binderup I,

Chander S, Coombes RC. EB1089: A new vitamin D

analogue that inhibits the growth of breast cancer cells

in vivo and in vitro. Biochem Pharmacol 1992;44:

2273-80. 9. Akhter J, Goerdel M, Morris DL. A vitamin D 3 analogue

(EB1089) inhibits in vitro cellular proliferation of hu-

man colon cancer cells. Br J Surg 1996;83:229-30. 10. Euhus D, Hudd C, La Regina M, Johnson F. Tumor

measurement in nude mouse. J Surg Oncol 1986;31:

229-34.

11. Stewart GJ, Connor JL, Lawson JA, Preketes A, King J,

Morris DL. Octreotide reduces the kinetic index, prolif-

erating cell nuclear antigen-maximum proliferative in-

dex, in patients with colorectal cancer. Cancer 1995;76:

572-8.

12. Abe J, Nakano T, Nishii Y, Matsumoto T, Ogata E, Ikeda

KA. N o v e l v i t a m i n D 3 a n a l o g , 2 2 - o x a - 1 , 2 5 -

dihydroxyvitamin D3, inhibits the growth of human

breast cancer in vitro and in vivo without causing hy-

percalcemia. Endocrinology 1991;129:832-7.

13. Schwartz GG, Hill CC, Oeler TA, Becich MJ, Bahnson

RR. 1,25-Dihydroxy-16-ene-23-yne-vitamin D 3 and

prostate cancer cell proliferation in vivo. Urology 1995;

46:365-9. 14. Shabahang M, Buras RR, Davoodi F, Scumaker IM,

Nauta RJ, Evans DR. 1,25-Dihydroxswitamine D 3 recep-

tor as a marker of human colon carcinoma cell line

differentiation and growth inhibition. Cancer Res 1993;

53:3712-8. 15. Lointer P, Meggouh F, Dechelotte P, et al. 1,25-

Dihydroxyvitamin D 3 receptors and human colon ade-

nocarcinoma. BrJ Surg 1991;78:435-9.

16. Frampton RJ, Suva J, Eisman JA, et aL Presence of

1,25-dihydroxyvitamin D 3 receptors in established hu-

man cancer ceil lines in culture. Cancer Res 1982;42:

1116-9.

17. Vandewalle B, Adenis A, Hornez I, Revillion F, Lefebvre

J. 1,25-Dihydroxyvitamin D 3 receptors in normal and

malignant human colorectal tissues. Cancer Lett 1994;

86:67-73.

18. Lointier P, Wargovich MJ, Saez S, Levin B, Wildrick DM,

Boman BM. The role of vitamin D 3 in the proliferation

of a human colon cancer cell line in vitro. Anticancer

Res 1987;7:817-22.

19. EismanJA, Sutherland RL, McMenemy ML, FragonasJC,

M u s g r o v e EA, P a n g GY. E f f e c t s o f 1 ,25-

dihydroxyvitamin D 3 on cell cycle kinetics of T47D

human breast cancer ceils. J Cell Physiol 1989;138: 611-6.