Effects of Gastrin, Proglumide, and Somatostatin on Growth of Human Colon Cancer

GASTROENTEROLOGY 1988;95:1541-8

Effects of Gastrin, Proglumide, and Somatostatin on Growth of Human Colon Cancer JILL PALMER SMITH and TRAVIS E. SOLOMON

Department of Medicine, Milton S. Hershey Medical Center, The Pennsylvania State University, Hershey, Pennsylvania; Departments of Medicine and Physiology, Kansas University Medical Center, Kansas City, Kansas, and Research Service, Kansas City Veterans' Administration Medical Center, Kansas City, Missouri

The effects of gastrin, proglumide (a gastrin receptor antagonist), and somatostatin on growth of human colon adenocarcinoma cell lines CX1, X56, and HT29 were examined in two experimental models. Nude mice bearing xenografts of colon cancer CXl or X56 were treated for 14-25 days subcutaneously with saline, pentagastrin (0.5 or 1.0 mg/kg), proglumide (250 or 500 mg/kg) , or somatostatin 14 (33, 100, or 300 j.tg/kg) twice daily. Tumor volume, weight, protein, and deoxyribonucleic acid were measured. HT29 cells were grown in vitro and the effects of gastrin 17, proglumide, and somatostatin on growth were evaluated by cell counts or [3H] thymidine incorporation. The larger dose of pentagastrin significantly increased tumor growth in the nude mouse (p < 0.005) and gastrin induced a biphasic effect on deoxyribonucleic acid synthesis in tissue culture with significant increases of up to 39% (p < 0.025). Somatostatin alone significantly inhibited tumor growth in two of the cell lines and also inhibited the gastrin-induced growth. Proglumide had no effect by itself but significantly inhibited gastrin-stimulated growth. These findings suggest that growth of some human colon cancers may be hormone-dependent.

G

rowth of many tissues is regulated by hormones, and growth of tumors arising from these tissues may remain under hormonal control. Thyroidstimulating hormone influences the growth of thyroid cancer (1). Growth of breast cancer is influenced by estrogens. About 50% of all primary breast cancers are estrogen receptor protein-positive, and response to hormonal therapy is correlated with the incidence and quantity of estrogen receptors (2). Cancer of the prostate responds to treatment with estrogens or bilateral orchiectomy (3). Because gastrointestinal hormones regulate many

aspects of cell proliferation and differentiation in the gastrointestinal tract (4), it is possible that growth of some gastrointestinal cancers is also hormonally regulated. Gastrin is the best characterized hormone in terms of its trophic effects on normal gastrointestinal mucosa. Exogenous and endogenous gastrin increase deoxyribonucleic acid (DNA), ribonucleic acid, and protein synthesis in oxyntic gland mucosa, small intestinal mucosa, and colonic mucosa (4,5). These effects are presumed to be receptor-mediated, and both normal rat colonic mucosa and some human colon cancers have been shown to possess gastrin receptors (6). Recently, exogenously administered pentagastrin has been demonstrated to stimulate growth of carcinogen-induced colon cancers in two animal models (7,8). Pentagastrin has also been shown to exert a trophic effect on normal and malignant human colonic epithelial cells maintained in vitro (9). Gastrin 17 was found to be a weak stimulant of in vitro proliferation in another human colon carcinoma cell line (10); however, gastrin 17 in physiologic doses can cause marked stimulation of colon cancer growth when the cells are synchronized with [3H)thymidine (11). Proglumide [( ±) -4-(benzoylamino)-5-(dipropylamino)-5-oxopentanoic acid), a gastrin and cholecystokinin receptor antagonist, has been shown to inhibit the trophic action of pentagastrin on oxyntic gland, duodenal, and colonic mucosa in rats (12). In a recent report, proglumide was found to inhibit growth of a transplantable murine colon carcinoma and to prolong survival in tumor-bearing mice, an effect postulated to be due to blockage of the trophic effects of endogenous gastrin on the tumor by

© 1988 by the American Gastroenterological Association

0016-5085/88/$3.50

1542

SMITH AND SOLOMON

proglumide (13). This effect has not been demonstrated on human colon carcinoma. Somatostatin inhibits the release of several gastrointestinal hormones and their action on target organs (14). Exogenous somatostatin decreases normal and gastrin-stimulated cell proliferation in several gastrointestinal organs (15-19), but its effects on proliferation in normal colonic mucosa or colon carcinoma have not been studied. Somatostatin has been shown to decrease weight and volume of carcinogeninduced acinar pancreatic carcinoma in rats and ductal pancreatic carcinoma in Syrian hamsters (20), suggesting that it might have inhibitory effects on the growth of other types of gastrointestinal cancers. The purpose of the present study was to determine the effects of gastrin, proglumide, and somatostatin on the growth of two transplantable human colon cancer cell lines maintained in nude mice and one human colon cancer cell line maintained in tissue culture.

Materials and Methods Materials and Drugs Pentagastrin was a gift from Professor John Morley, ICI, Macclesfield, England. Proglumide was a gift from Barrows Research Group, Valley Stream, NY. Synthetic somatostatin 14 was donated by Professor E. Wunsch, Munich, F.R.G. The human colon adenocarcinoma CXl was obtained from the National Cancer Institute as tissue fragments frozen on dry ice. A nude mouse bearing human colon adenocarcinoma X56 was provided by Dr. Henry A. Azar, University of South Florida, Tampa. HT29 human colon cancer, first characterized in the laboratory of Dr. Jorgen Fogh at Sloan Kettering, was provided by Dr. John Turner, University of Missouri. Histology of the CXl and X56 tumors were examined by hematoxylin and eosin staining of tissue obtained from xenografts in nude mice, and both had the characteristics of moderately differentiated adenocarcinomas. Male, 5-6-wk-old athymic nude mice were obtained from the National Cancer Institute. Mice were kept in filter top cages in a sterile environment with autoclaved bedding, food, and water. All procedures involving mice were done by sterile technique under a laminar flow hood. HT29 tumor cells were grown in 25-cm 2 Corning tissue culture flasks containing McCoy's medium supplemented with 10% fetal calf serum, L-glutamine (2 mM), 1% (vol/vol) MEM essential vitamins, 2 1Lg/ml gentamicin, 0.5 1Lg/ml fungizone, and 1 1Lg/ml hydrocortisone, at 37°C in humidified air containing 7% CO 2, [3H)Thymidine with a specific activity of 80.1 Ci/mmol was obtained from New England Nuclear Products, Boston, Mass. For determination of [3H)thymidine incorporation, cells were harvested onto Mash II glass fiber filter paper (grade 934 AH) from Whittaker M.A. Bioproducts, Walkersville, Md.

GASTROENTEROLOGY Vol. 95, No.6

Procedure In the in vivo studies, single tumor cell suspensions were made using tumors removed from donor nude mice bearing CXl or X56 colon cancers. The tumor was removed from the donor mouse and cut into small fragments with sterile scissors in saline (0.154 M NaCl). The tumor fragments were mechanically dissociated by passage through sterile needles of decreasing caliber until passage through a 22-gauge needle was performed without difficulty. The tumor cell suspension was allowed to stand for 10 min in a sterile test tube so that any remaining large fragments would settle. Cells in the supernatant were stained with trypan blue and viability was always ~95%. In the first experiment, each mouse received a subcutaneous injection containing 1 x 10 6 viable CXl or X56 tumor cells on day 1. In a second experimental series, each mouse received 5 x 10 6 viable tumor cells. On the third day after injecting the tumor cells, mice with palpable tumors were selected for study. Palpable tumors were present on day 4 in 81 % of the mice receiving 1 x 106 CXl cells and 89% receiving X56; after receiving 5 x 10 6 cells in experiment 2, all of the mice injected with CXl and X56 had tumors. Peptides and drugs were diluted to the required concentration in saline containing 1% (wt/vol) bovine serum albumin. Injections were made in a 1:4 dilution of hydrolyzed gelatin to prolong absorption. Because of precipitation, proglumide could not be suspended in gelatin and was injected in saline. Starting on day 4, mice in experiment 1 were injected subcutaneously twice daily with 0.1 ml of one of the following: 1% bovine serum albumin in saline (control), 500 1Lg/kg pentagastrin, 250 mg/kg proglumide, 100 ILg/kg somatostatin (CXl mice only), or pentagastrin and proglumide in combination (CXl mice only). Mice in experiment 2 were injected with 0.2 ml of one of the following: 1% bovine serum albumin in saline; 1000 1Lg/kg pentagastrin, 500 mg/kg proglumide, or 33, 100, or 300 1Lg/kg somatostatin. Tumors were measured every 4 days and tumor volumes were calculated as length x (width)2 x 0.5 as per the method described by Osieka et al. (21). Mice were reweighed weekly and dosages were recalculated. The final injections were given 12-18 h before the mice were killed; mice were fed ad libitum until they were killed. Mice in experiment 1 were killed on day 24 and tumors were removed, weighed, and analyzed for protein and DNA content. Mice bearing the CXl tumor in experiment 2 were killed on day 18 and mice bearing the X56 tumor were killed on day 28. Tumors were minced and homogenized in ice-cold deionized water for 20-30 s using a Polytron tissue grinder (Brinkmann Instruments, Westbury, N.Y.). Protein content was determined by the procedure of Lowry et al. (22) using bovine serum albumin as standard. Deoxyribonucleic acid content was determined by the procedure of Burton (23) using calf thymus DNA as standard. In another series of experiments, the effects of peptides on HT29 human colon cancer cell line growth were determined with cell counts and [3H)thymidine incorporation. Deoxyribonucleic acid synthesis was measured by exposing 50,000 HT29 tumor cells per well to [3H)thymidine (0.4 1LCi/well) for 48 h using 96-well tissue culture plates. Each

December 1988

HORMONAL EFFECTS ON HUMAN COLON CANCER GROWTH

1543

EXPERIMENT 2

EXPERIMENT 1

1.00 .45

.90

.40

.80

* III

.OIl

*p<.OO5 t p <.05

..... .35

.OIl .....

III

0

III

10

::::J

III

2!

-'

0

>

II:

0 2!

III

.OIl

.....

N

III

=I-

a

.25

o o

i:::) ....

.20

z

CI

t=e

0

f

::::J

~

.15 .10 .05

II:

0 2

::::J ~

III

.OIl

.....

III

:a.

.40

o o o

.OIl

10 III

.,.,

i

!

:::)

.50

-..

III

a

....

CI

0

a:

A-

:a.

~

e

~ (/)

0

:I

en

a: 0

III

o

III

:a.

0 0

t= e

~

~

(/)

.....

III

z

~ (/)

e

III

.OIl

0

e

:I 0

:a.

., 0 0

Z

t=e ~

(/)

0

~

e

:I 0

... 0

(/)

*

a: ~

z

.10

0 0

0

li)

A-

.....

(/)

.20

=I-

~

z

III

.OIl

~

0

...

.30

~

e .OIlIII CI ..... e III

CJ

>

III

e

en

~

-' 0

.60

~

a:

z

::::J

en

~

0

2

o

!

....

0 0

III

E

.30

III

E

.70

III

CJ

9 6 6 8 6 N= N= 6 6 6 7 Figure 1. Effects of peptide treatment on volume of CXl colon adenocarcinoma xenografts grown for 24 days (experiment 1, left panel) or 18 days (experiment 2, right panel) in nude mice. Mice iIi experiment 1 received 1 x 10 6 viable tumor cells subcutaneously, whereas mice in experiment 2 received 5 x 106 cells. The indicated doses of pentagastrin, proglumide, or somatostatin were injected subcutaneously twice per day beginning on day 4. Values shown are the mean ± SE for tumor volume (cm 3 ) on the day the mice were killed; number of animals in each treatment group is shown beneath the bars. Levels of statistical significance between control and treated groups are indicated on the figure. Pentagastrin significantly increased tumor volume in experiment 2, whereas somatostatin significantly decreased volume in both experiments. Proglumide had no significant effect in either experiment. EXPERIMENT 2

EXPERIMENT 1

1$00

300 -r-

III

.p<.025

!

-,

III

.OIl

~

.....- ~

iii

==

II:

150

0 2!

.OIl .....

~

z

-III

_L-

::::J

-' «

.....

-'-

J:

illi

:a.

0 0

100

10

Z

ii:

i 50

....

0

a: ~

Z

N=

~

(/)

e CI e ~

r--

III

e:I .... :::)

A-

A-

9

6

III

.OIl .....

III

E

CI

0

III

-

_L...

Z

0

CJ

-'-

UI

0

a:

10 N

8

III

!

~

C!J

...0

II:

0

~

-'

e

:I 0

(/)

rf 6

rf

z u::

....

0

100

a: ~

z

N=

:a.

o o

...

III

.OIl

.....

III

:a.

~

e

o

!

...

-.

e ~ en

--

~

o e

~

o(/)

z

r-t" 1-'-

i

~ (/)

e CI e

o

~

e

r--

--

~

z

0

III

6

6

CJ

!

III

:I

A-

III

« 200

.,.,

=I-

:a.

::::J

~

III

.OIl .....

l-

III

.OIl

0 0 0

en

.....

.,o

.....

0

III

.OIl

CIt

(/)

== 300

0 2

~

:I

iii

z t=e

a

:::)

J:

=I-

o

.... CI oa:

400

*p<.005

E

o

10 III

....

r-~

200

(!)

r- ~

500

250

Ii

III

.OIl .....

:I

o(/)

T

A-

6

6

7

6

Figure 2. Tumor weight in milligrams from the same experiments as in Figure 1. Design of the figure is the same as Figure 1. pentagastrin significantly increased the weight of CXl tumors in experiment 2. Somatostatin, 100 /Lg/kg in experiment 1 and 300 /Lglkg in experiment 2, significantly de~reased tumor weight. Proglumide had no effect in either experiment.

GASTROENTEROLOGY Vol. 95, No.6

1544 SMITH AND SOLOMON

EXPERIMENT 2

EXPERIMENT 1

35 -,-

110

-,-

r-- ~

.....

!

IZ

w

IZ

r-- ~ 20

_

0

0

z

iii

.....

15

...:

.....

III ~

0 0

I-

10 ...I

0

a::

5

~

Z

N=

! a:: ~ en c CI c ~

Z

()

0

1&1 D..

9

6

'0

80

...

I-

70

Z

r-- f -

j: C

_.....

~

en

0

~

C

:.

1ft

0

IE: D..

III

III ~

0 0

-,-

0

en

--

1&1

2 ...:III :. ..... ::::I

-_J..

III

...I

CI E

0

.....

0

a::

1ft

D.. N

6

8

*

90

...:

25 _L.-

...:

+p <.025 III

g,

III

100

30

!

z

w

I-

z

0 0

z

iii

50

o o o

40 30

0

.....

III

~

...:

III ~ t'I t'I

0 0

....

Z

Z

0

C

i::::I CI

a:: D..

~ ~

en

.....

III ~

.....

III ~

j:

0 0

en

~

z

~

C

C

0

0

C

C

:.

:.

en

en

0

III

...:

0

t'I

j:

~

en 0

~

c

en

:.

0

CI

en

~

~

c c

a::

z

N=

.....

~

....

10

...:

~

z ii:

20

III

III

0 0

0

60

IE: D..

E

1&1

III

*p<.005

III

1ft

...:

I-

.....

0

Z

()

0

1&1 D..

6

6

6

6

7

6

Figure 3. Effects of peptide treatment on content of protein (milligrams) in CXl colon adenocarcinoma xenografts in nude mice. Data are from the same experiments described in Figure 1, and design of the figure is similar. The pattern of effects of pentagastrin, proglumide, and somatostatin are the same as for tumor weight: pentagastrin significantly increased protein content in experiment 2, somatostatin significantly decreased protein content in both experiments, and proglumide had no significant effect.

well was inoculated with 100 ILl of tumor cells in medium, 100 ILl of medium alone (control), or peptide at various concentrations in the media, and 50 ILl of [3Hlthymidine. Gastrin 17 alone in various concentrations (0.4,4,40,400, 4000 pM), proglumide alone (40 nM), or combinations thereof were added. After 48 h cells were harvested onto filter paper using a Mash II sample Harvester and a solution of 0.154 M NaCI containing 0.25% (wtlvol) trypsin with 0.1% (wt/vol) ethylenediaminetetraacetic acid. Filters were punched out, added to 9 ml of scintillation fluid, and counted on a Beckman liquid scintillation counter (Beckman Instruments, Fullerton, Calif.). Growth was also evaluated with viable cell counts in which 224,000 HT29 tumor cells were plated onto 4-cm 2 tissue culture wells containing medium alone (control), gastrin 17 (4 nM or 400 pM), somatostatin (4 nM or 400 pM), or combinations thereof. Cells were harvested after 4 days with a solution containing 0.25% (wt/vol) trypsin and

0.1 % (wt/vol) ethylenediaminetetraacetic acid, and were stained with trypan blue; viable cell counts were performed with a hemocytometer.

Calculations and Statistical Analysis All data represent total amounts per tumor. The data are presented as the mean ± standard error of treatment groups. Statistical analysis was done using the nonparametric Mann-Whitney test.

Results Mice in control groups and all peptide treatment groups gained ~1 g body wt/wk, and there were no significant differences in body weight gain among the groups (data not shown). There were no deaths in control or peptide-treated groups.

Table 1. X56 Colon Cancer: Effects of Peptide Treatment in Experiment 1 Control Tumor volume (em 3 ) Tumor weight (mg) Protein content (mg) DNA content (mg)

0.197 179 17.3 0.62

± ± ± ±

0.030 35 3.2 0.11

Pentagastrin (500 ILg/kg) 0.291 221 15.6 0.58

± ± ± ±

0.026° 37 4.1 0.11

Proglumide (250 mg/kg) 0.159 192 15.6 0.57

± ± ± ±

0.018 26 2.2 0.10

DNA, deoxyribonucleic acid. Mice received 1 x 106 viable tumor cells subcutaneously on day 1. Groups of mice (n = 6 for control and proglumide. n = 5 for pentagastrin) were injected from day 4 through day 23 with saline. pentagastrin. or proglumide in the indicated doses twice per day and then killed on day 24. a p :s 0.05 compared with control.

HORMONAL EFFECTS ON HUMAN COLON CANCER GROWTH

Decel:b"er 1988

EXPERIMENT 2

EXPERIMENT 1 3.2

1.2

-,

·p<.025

r- f~

.... zw

....z

-'-

.8

*

,.......

-

=I.

.6

...

....

=I.

E

'"

...i e 0 e

i;:,

0

III II.

II: II.

9

6

-' 0

...z

...z

II:

0

N=

... z

... ... ...0 e c

:a

I/)

I/)

...

0

:a

z

-

II:

e 0 e

0

...z

...z

6

6

II:

.4

0 0

N=

110

I/)

..I

6

0

...

.8

r- • --

8

0

0 0 0

1.2

z

--

-' 0 0

=I.

=I.

0 0

I/)

A-

lit

I/)

III Q

I/)

c(

0

I/)

z

....

... ...e 0 ...e

II:

1.6

110

,JiI. .... 110

z

,JiI.

0

:a

0

I/)

(f) (f)

-'

110

tp<.010

=I.

0 0

2.0

*p< .005

....110

:a

0

I/)

lit

0 0

.4

...e 0 ... e

,JiI.

110

c(

j:

-m-

110

....

,JiI.

.2

! ....Z ....W Z

0 0

110

,JiI.

;:,

-a

110

0 0

I/)

III Q

2.4

110

....

Z

0

0

E

2.8

,JiI.

_L...

0

z

,JiI.

-r-

r-- f0-

.5

110

....

110

-,

1.0

1545

110

.... ,JiI.

110

=I.

.., 0 0

z

~

...e 0 ...e

110

:a

0

110

III II.

7

6

6

6

Figure 4. Total content of DNA (milligrams) in CXl human colon adenocarcinoma xenografts from the same experiments as in Figure 1. Pentagastrin significantly increased DNA content in experiment 2, somatostatin significantly decreased DNA content in both experiments, and proglumide had no significant effect.

The results of experiments 1 and 2 in mice bearing eXl tumors are shown in Figures 1-4. The lower dose of pentagastrin in experiment 1 had no statistically significant effect on any of the four measurements of eXl tumor growth, whereas the higher dose (1000 p,g/kg) in experiment 2 significantly increased tumor volume, weight, DNA content, and protein content. Proglumide given alone did not significantly alter tumor growth at either of the two doses administered when compared with control group values. However, combining proglumide (250 mg/kg) with pentagastrin (500 p,g/kg) in experiment 1 resulted in significant (p < 0.05) reductions compared with the pentagastrin group for tumor volume (0.079 ± 0.023 vs. 0.300 ± 0.103 cm 3 ), weight (69 ± 23 vs. 203 ± 44 mg), protein content (10.5 ± 3.4 vs. 27.8 ± 4.0 mg), and DNA content (0.37 ± 0.11 vs. 0.84 ±

0.13 mg). In experiment 1, somatostatin (100 p,g/kg) reduced all measurements of eXl tumor growth significantly. In experiment 2, in which initial tumor load was 5 times larger, 100 p,g/kg of somatostatin significantly decreased final tumor volume and DNA content, whereas 300 p,g/kg of somatostatin reduced all measurements of growth. The results of the two experiments in mice bearing X56 tumors are shown in Tables 1 and 2. The low dose of pentagastrin significantly increased tumor volume in experiment 1, and the high dose significantly increased tumQr protein and DNA content in experiment 2. Neither proglumide nor somatostatin had any statistically significant effect on growth of the X56 tumors in these experiments. The effects of gastrin l'i' alone at various concentrations on HT29 tumor DNA synthesis are shown in

Table 2. X56 Colon Cancer: Effects of Peptide Treatment in Experiment 2 Control Tumor volume (cm 3 ) Tumor weight (mg) Protein content (mg) DNA content (mg)

0.342 220 19.6 0.61

± ± ± ±

0.043 48 3.8 0.13

Pentagastrin (1000 ~g1kg) 0.643 435 36.4 1.32

± ± ± ±

0.183 133 9.9 a

0.34a

Proglumide (500 mg/kg) 0.337 208 19.0 0.95

± ± ± ±

0.066 54 4.4 0.28

Somatostatin (33 ~g1kg) 0.477 350 28.1 0.99

± ± ± ±

0.083 118 10.2 0.30

Somatostatin (100 ~g/kg) 0.325 207 16.8 0.83

± ± ± ±

0.121 66 4.8 0.41

Somatostatin (300 ~glkg) 0.262 164 16.2 0.49

± ± ± ±

0.072 46 3.9 0.13

DNA, deoxyribonucleic acid. Mice received 5 x 10 viable tumor cells subcutaneously on day 1. Groups of mice (n = 6 for each treatment) were injected from day 4 through day 27 with saline, pentagastrin, proglumide, or somatostatin in the indicated doses twice per day and then killed on day 28. a p :s 0.05 compared with control. 6

1546

GASTROENTEROLOGY Vol. 95, ,t.J'o. 6

SMITH AND SOLOMON

CPM X 100 140

r--------------------------------------,

130 t------------=;;;;iiii~--~~-__j 120 110

~----------~__~--------------------~

~----~~----------------------------~

100 ~~--------------------------------__1

4

.4

40

400

4000

GASTRIN CONCENTRATION, pM

Figure 5. Fifty thousand HT29 colon cancer cells were incubated with [3H)thymidine for 48 h. Results are shown as counts per minute (cpm). Gastrin 17 caused a biphasic effect on [3H)thymidine incorporation, with maximum increases of 39% at the 400 pM dose (p < 0.025).

Figure 5. Gastrin 17 caused a biphasic effect on [3H]thymidine incorporation, with maximum increases of 39% at the 400 pM dose (p < 0.025). Proglumide (40 nM) had no effect on [3H]thymidine incorporation alone; however, the increased growth seen in gastrin-treated cells (400 pM) was significantly (p < 0.005) inhibited by the addition of proglumide to the medium (Figure 6). Results of the cell count studies are shown in Figure 7. Gastrintreated cells had significantly greater cell counts, with increases up to 28% at the 10- 10 M dose as compared with controls (p < 0.005). Somatostatin alone (400 pM) significantly decreased cell counts, and somatostatin markedly inhibited the gastrin-

,. 16 ~

'"o.... x

I

E 14 Cl; (,)

,. "

r--

o

400pM Ga~trin

40nM Proglumide

400pM Ga~trin 40nM+ Proglumide

PEPTIDE CONCENTRATION Figure 6. Fifty thousand HT29 colon cancer cells were incubated with [3H)thymidine either alone (control) or in combination with peptide. Gastrin (400 pM) caused significantly greater [3H)thymidine incorporation as compared with controls (*p < 0.01). Proglumide had no effect on HT-29 cells alone, but when the same dose of proglumide (40 nM) was combined with gastrin (400 pM) the increased [3H)thymidine counts were significantly reduced (**p < 0.005) when compared with gastrin alone.

"'0

....x

a: W CD

::E :::> Z

-J -J

W

()

24

,

* **

16

,.,.

'*

8

Gastrin 400pM

Somatostatin 400pM

Gastrin 400pM+ Somatostatin 400pM

PEPTIDE CONCENTRATION Figure 7. Two hundred twenty-four thousand cells were plated into each well and incubated for 4 days in McCoy's medium plus 10% fetal calf serum and additives. G<\strin-treated cells had significantly greater cell counts with increases of up to 28% at the 400 pM dose as compared with controls (*p < 0.005). Somatostatintreated cells had significantly fewer cell counts, with a decrease of 14% compared with controls (* *p < 0.05). Somatostatin in combination with gastrin significantly inhibited the gastrin-induced proliferation (* **p < 0.001), with a 34% reduction in cell counts as compared with cells treated with gastrin alone.

stimulated cell growth when both peptides were added in combination (p < 0.001).

Discussion Our results demonstrate that large doses of pentagastrin and gastrin stimulate the growth of three human colon adenocarcinomas in vivo and in vitro. The gastrin-induced cell growth can be inhibited by proglumide and somatostatin. Proglumide alone had nO effect on any of the cell lines. Somatostatin alone significantly inhibited growth of CXl in vivo and HT29 in vitro but had no effect on X56. These findings support the hypothesis that some human colon cancers are sensitive to hormonal regulation of their growth. Considerable evidence has recently been presented demonstrating that growth of some human colon cancers may be stimulated by hormones. Based on growth of human colon cancer cell lines in vitro, stimulation of cell proliferation has been reported for epidermal growth factor (10,24), insulin (10,25), pentagastrin (9) and gastrin (10), glucagon (10), T3 (10), and hydrocortisone (10). The role of gastrin as a potential growth factor for colon cancer is further supported by studies using animal models of carcinogen-induced tumors. Three studies have reported increased growth of established colon cancers after administration of exogenous pentagastrin (7,8,26) to mice and rats bearing carcinogen-induced primary or transplanted tumors. In addition, administration of pentagastrin has been reported to in-

December 1988

HORMONAL EFFECTS ON HUMAN COLON CANCER GROWTH

crease cancer-related mortality in such a model (8). However, others have found no effect of chronic treatment with tetragastrin on growth of carcinogeninduced colon cancer in rats (27). In the present study, doses of pentagastrin required to stimulate growth of the two human colon cancers were large compared with reported effects of pentagastrin on a transplantable mouse colon cancer (8). Assuming that pentagastrin acts at least in part directly on cancer cells, specific colon cancers may show a preferential growth response to different molecular forms of gastrin or there may be potentiation among gastrin and other hormonal growth factors. It is also not known whether all colon cancers possess specific receptors for the growth-promoting effects of gastrin nor what is the nature of receptor-mediated events that result in proliferation. Answering these questions is of obvious importance for understanding the potential role of endogenous gastrin as a regulator of colon cancer growth. The lack of effect of proglumide on growth of eX1 and X56 human colon cancer xenografts in our experiments suggests that normal levels of endogenous gastrin in nude mice were not sufficient to stimulate tumor growth. Although proglumide has been reported to decrease the growth of a transplantable mouse colon cancer and enhance survival of tumor-bearing mice (13), this difference could be due to the greater apparent sensitivity of the mouse tumor to gastrin (8) described above. However, the importance of endogenous gastrin as a growth regulator of colon cancer is unclear at present. Some investigators have found that growth of experimental colon cancer in animal models is increased by conditions associated with hypergastrinemia, such as proximal resection of small intestine (28), antral exclusion (7,29,30), and feeding (31). Others have not observed such an association (32). Removal of the major source of circulating gastrin by surgical antrectomy has been reported to have no effect on incidence or growth of carcinogen-induced colon cancers in rats (7,32). Our observation that proglumide did decrease several measurements of tumor growth in mice receiving the low dose of pentagastrin and in gastrin-treated cells in tissue culture suggests that proglumide may only be effective in blocking the effects of elevated levels of gastrin. More detailed studies with different molecular forms of gastrin and more potent and specific gastrin receptor antagonists should be helpful in determining the importance of endogenous gastrin as a growth factor for human colon cancer. Much less information is available concerning possible hormonal inhibitors of colon cancer growth. Growth of a human colon adenocarcinoma cell line in vitro was reported to be slowed by addition of

1547

prostaglandins and gonadal steroids to the culture medium (10), but somatostatin had no effect in this system (10). Another study found that f3-agonists decreased the tumor volume of two human rectal carcinomas grown as xenografts in nude mice (33). Our data indicate that somatostatin decreased the growth rate of HT29 and eX1 human colon cancer but not X56. The inhibitory effect of somatostatin on eX1 growth appears to depend on the dose administered and the initial tumor volume. The doses of somatostatin that we used were chosen because they have previously been shown to inhibit growth of normal exocrine pancreas (34) and experimental pancreatic cancer (20). However, these doses are likely to have produced blood levels of somatostatin much higher than are present normally. Nevertheless, these results indicate that somatostatin 14 or more potent analogues may have therapeutic potential for some human colon cancers. The mechanism of inhibition produced by somatostatin in our experiments is not known. Somatostatin has been found to inhibit the growth of several different types of experimental cancers, many of which are endocrinedependent (35). The well-known inhibitory effects of somatostatin on release of many hormones suggest that inhibition of colon cancer growth, as for other cancers, might be due to somatostatin-induced decreases in one or several growth-stimulating hormones. It is also possible that somatostatin might have acted directly on eX1 and HT29 cells to block the effects of gastrin and other growth factors. Either mechanism would explain our results. The lack of effect of somatostatin on growth of X56 could have been due to lower sensitivity of tumor cells to hormonal growth factors, and thus less effect from decreasing the levels of such stimulants, or to lack of specific receptors for somatostatin on X56 cells. It would be of considerable interest to determine whether specific receptors for somatostatin exist on colon cancer cells. In conclusion, our studies suggest that growth of human colon adenocarcinoma may be regulated by gastrointestinal peptides. The gastrin-stimulated cell growth may be mediated via a specific gastrin receptor as its effect is abolished by the addition of a gastrin-receptor antagonist. With the increasing availability and safety of long-acting somatostatin analogues, the inhibition of tumor growth by somatostatin has obvious implications for use as a therapeutic agent in the treatment of colon cancer.

References 1. DeRabertis F, Yamashita K, Dekken A, Larsen PR, Field JB.

Effects of thyroid-stimulating hormone on adenylcyclase activity and intermediary metabolism of cold thyroid nodules

1548

2.

3.

4.

5. 6.

7.

8. 9.

10.

11. 12. 13.

14.

15.

16.

17.

18.

19.

20.

GASTROENTEROLOGY Vol. 95, No.6

SMITH AND SOLOMON

and normal human thyroid tissue. J Clin Invest 1972;51:110917. Allegra JC, Barlock A, Huff KK, et al. Changes in multiple or sequential estrogen receptor determinations in breast cancer. Cancer 1980;45:792--4. Huggins C, Hodges CV. Studies on prostate cancer: the effect of castration, of estrogen and of androgen injection on serum phosphatases in metastatic carcinoma of the prostate. Cancer Res 1941;1:293-7. Johnson LR. Regulation of gastrointestinal growth. In: Johnson LR, ed. Physiology of the gastrointestinal tract. New York: Raven, 1981:169-96. Dembinski AB, Johnson LR. Growth of pancreas and gastrointestinal mucosa in antrectomized and gastrin-treated rats. Endocrinology 1979;105:769-73. Singh P, Rae-Venter B, Townsend CM, Khalil T, Thompson JC. Gastrin receptors in normal and malignant gastrointestinal mucosa: age-associated changes. Am J Physiol 1985;249: G761-9. McGregor DB, Jones RD, Karlin DA, Romsdahl MM. Trophic effects of gastrin on colorectal neoplasms in the rat. Ann Surg 1982;195:219-23. Winsett OE, Townsend CM, Glass EJ, Thompson JC. Gastrin stimulates growth of colon cancer. Surgery 1986;99:302-7. Sirinek KR, Levine BA, Moyer MP. Pentagastrin stimulates in vitro growth of normal and malignant human colon epithelial cells. Am J Surg 1985;149:35-8. Murakami H, Masui H. Hormonal control of human colon carcinoma cell growth in serum-free medium. Proc Natl Acad Sci USA 1980;77:3464-8. Kusyk q, McNiel NO, Johnson LR. Stimulation of growth of a colon cancer cell line by gastrin. Am J Physiol 1986;251: G597-601. Johnson LR, Guthrie PD. Proglumide inhibition of trophic action of pentagastrin. Am J Physiol 1984;246:G62-6. Beachamp RD, Townsend CM, Singh P, Glass EJ, Thompson JC. Proglumide, a gastrin receptor antagonist, inhibits growth of colon cancer and enhances survival in mice. Ann Surg 1985;202:303-9. Schusdziarra V. Somatostatin-physiological and pathophysiological aspects. Scand J Gastroenterol [Suppl] 1983;82:6984. Lehy T, Dubrasquet M, Bonfils S. Effect of somatostatin on normal and gastrin-stimulated cell proliferation in the gastric and intestinal mucosae of the rat. Digestion 1979;19:99-109. Lehy T, Dubrasquet M, Brazeau p, Bonfils S. Inhibitory effect of prolonged administration of long-acting somatostatin on gastrin-stimulated fundic epithelial cell growth in the rat. Digestion 1982;24:246-55. Stange EF, Schneider A, Schusdziarra V, Ditschuneit H. Inhibitory effects of somatostatin on growth and differentiation in cultured intestinal mucosa. Horm Metab Res 1984;16: 74-8. Lehy T. Trophic effect of some regulatory peptides on gastric exocrine and endocrine cells of the rat. Scand J Gastroenterol 1984;19(Suppll0l):27-30. Senegas-Balas F, Balas D, Pradayrol L, Laval J, Bertrand C, Ribet A. Long-term effect of somatostatin 14 on mouse stomach, antrum, intestine and exocrine pancreas. Acta Anat 1985;121:124-32. Redding TW, Schally AV. Inhibition of growth of pancreatic

21.

22. 23.

24. 25.

26.

27.

28.

29. 30.

31. 32. 33. 34. 35.

carcinomas in two animal models by analogs of hypothalamic hormones. Proc Nat! Acad Sci USA 1984;81:248-52. Osieka R, Houchens DP, Goldin A, Johnson RK. Chemotherapy of human colon cancer xenografts in athymic nude mice. Cancer 1977;40:2640-50. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Falin phenol reagent. J Bioi Chem 1951;193:265-75. Burton K. A study of the conditions and mechanisms of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem J 1956;62:315-23. St. Hilaire RJ, Gospodarowicz D, Kim YS. Epidermal growth factor: effect on the growth of human colon adenocarcinoma cell line (abstr). Gastroenterology 1980;78:1271. Cezard JP, Forgue-Lafitte ME, Chamblier MC, Rosselin GE. Growth-promoting effect, biological activity, and binding of insulin in human intestinal cancer cells in culture. Cancer Res 1981;41:1148-53. Svet-Moldavsky GP. Dependence of gastro-intestinal tumors on gastro-intestinal hormones: pentagastrin stimulates growth of transplanted colon adenocarcinoma in mice. Biomedicine 1980;33:259-61. Tatsuta M, Yamamura H, lishi H, Noguchi S, Ichii M, Tanigachi H. Gastrin has no promoting effect on chemically induced colonic tumors in Wistar rats. Eur J Cancer Clin OncoI1985;21:741--4. Williamson RCN, Bauer FL, Oscarson JEA, Ross JS, Malt RA. Promotion of azoxymethane-induced colonic neoplasia by resection of the proximal small bowel. Cancer Res 1978;38: 3212-7. Elwyn KE, Jones RD, Romsdahl MM. Inhibitory effects of secretin on gastrin-stimulated rat colon neoplasms. Cancer 1985;55:1186-9. Karlin DA, McBath M, Jones RD, Elwyn KE, Romsdahl MM. Hypergastrinemia and colorectal carcinogenesis in the rat. Cancer Lett 1985;29:73-8. De Greve J, van der Elst J, Willems G. Refeeding of fasting rats stimulates DNA synthesis in implanted colon carcinoma. J Cancer Res Clin Oneal 1986;111:191-5. Oscarson JEA, Veen HF, Ross JS, Malt RA. Dimethylhydrazine-induced colonic neoplasia: dissociation from endogenous gastrin levels. Surgery 1982;91:525-30. Tutton PJM, Steel GG. Influence of biogenic amines on the growth of xenografted human colorectal carcinomas. Br J Cancer 1979;40:743-9. Morisset J, Genik P, Lord A, Solomon TE. Effects of chronic administration of somatostatin on rat exocrine pancreas. Regul Pept 1982;4:49-58. Schally AV, Comaru-Schally AM, Redding TW. Antitumor effects of analogs of hypothalamic hormones in endocrinedependent cancers. Proc Soc Exp Bioi Med 1984;175:259-81.

Received December 9, 1986. Accepted June 27, 1988. Address requests for reprints to: Jill Palmer Smith, M.D., Department of Medicine, Division of Gastroenterology, Milton S. Hershey Medical Center, Pennsylvania State University, P.O. Box 850, Hershey, Pennsylvania 17033. This work was supported by the Research Service of the Veterans Administration. The authors thank Catherine Mohesky for technical assistance and Sue Huntzinger for secretarial assistance.