JOURNAL OF OCULAR PHARMACOLOGY Volume 8, Number 1, 1992 Mary Ann Liebert, Inc., Publishers

The Effects of Transferrin Receptor Antibody, Transferrin Receptor Antibody Bound to Pseudomonas Exotoxin and Transforming Growth Factor-a Bound to Pseudomonas Exotoxin on Human Tenon's Capsule Fibroblast Proliferation ROBERT J. SMYTH, SHINICHI KITADA, and DAVID A. LEE Department of Ophthalmology, Jules Stein Eye Institute, UCLA School of Medicine, Los

Angeles, California ABSTRACT

Pharmacological agents which modulate the wound healing process by the inhibition of proliferation of fibroblasts may improve the success of glaucoma filtration surgery. Since cell proliferation is essential to the wound healing process, we targeted the surface receptors that are associated with proliferating cells. We present the effects of three such agents—purified mouse anti-human transferrin receptor monoclonal antibody 42/6 (anti-TfR-42/6), anti-transferrin monoclonal antibody bound to a Pseudomonas exotoxin (anti-TfR-PE40) and transforming growth

factor-a Pseudomonas exotoxin (TGF-a-PE40)—on human fibroblasts from Tenon's capsule. The inhibition of human subconjunctival fibroblast proliferation by anti-TfR-42/6 (with a concentration up to 25 pg/ml) and by anti-TfR-PE40 and TGF-a-PE40 (both with a concentration range of 5000-0.00001 pg/ml) was determined by colorimetric (OD), and cell counting (CC) assays over a 9-day period. Neither anti-TfR-42/6 nor anti-TfR-PE40 had an antiproliferative effect on the fibroblasts. TGF-a-PE40 demonstrated an antiproliferative effect in a dose response manner. The mean 50% inhibitory dose (ID50) by OD was 32.91 pg/ml, while the ID50 by CC was 27.88 pg/ml. EGF was used as a negative control for TGF-a-PE40 toxin. The inhibitory effect of the toxin conjugate was completely blocked by the addition of 1000 pg/ml of EGF. These in vitro studies show that TGF-a-PE40 may be useful in modulating the proliferation of human ocular fibroblasts; they also give some indication of drug dosages for future in vivo testing.

INTRODUCTION

Filtering surgery is an important procedure in the control of intraocular pressure when the glaucomatous eye is unresponsive to medical and/or laser therapy. The sequence of wound-healing events—fibroblast activation, attachment, migration, proliferation and collagen synthesis—occurs in response to filtering surgery and may result in subconjunctival fibrosis and bleb failure (1-3). Pharmacological agents which modulate the wound healing process by inhibiting proliferation of fibroblasts may improve the success of glaucoma filtration surgery (3). An alternative approach, which has recently been proposed for therapy against malignant cells, is to target the surface receptors that are associated with proliferating cells (4). Monoclonal antibodies may be used to directly block cell surface receptors while both monoclonal antibodies and growth factors may be bound with toxins and be used to deliver toxins specifically to target cells (5). The first of the three

pharmacological agents we tested, purified mouse anti-human transferrin monoclonal receptor antibody 42/6 (anti-TfR-42/6), blocks transferrin binding and inhibits human tumor cell growth (6). The second pharmacological agent tested, anti-transferrin receptor monoclonal antibody bound to a Pseudomonas exotoxin (anti-TfR-PE40), acts against tumors by binding to the transferrin receptor (forming a receptor ligand complex) which is internalized through 83

anti-TfR-42/6

anti-TfR-PE40

CD o-lie transferrin receptor

P nQOQ TGF-a-PE40 (§^D EGF CELL

EGF receptor

Figure 1. A diagrammatic representation, not drawn to scale, of the proposed mechanism of action of the anti-transferrin receptor monoclonal antibody 42/6 (anti-TfR-42/6), the anti-transferrin receptor bound to the Pseudomonas exotoxin (anti-TfR-PE40), the transforming growth factor a bound to the Pseudomonas exotoxin (TGF-a-PE40), and epidermal growth factor (EGF). coated

pits. Once inside the cytosol, the Pseudomonas exotoxin kills the cell (7). The last pharmacological agent we tested was the transforming growth factor-alpha-Pseudomonas exotoxin protein (TGF-a-PE40). This protein has two properties-the receptor binding activity of TGF-a, and the ability to kill cells which possess epidermal growth factor (EGF) receptors (8, 9) (TGF-a binds to the EGF cell surface receptor) (lO)-(Figure 1). MATERIALS AND METHODS Cell Lines

Cell lines were obtained and assays performed as previously described with slight modifications (11-15): Tenon's capsule tissue specimens were obtained from consenting patients

undergoing

various ocular surgical procedures. Within 4 hours of the surgical procedure, the tissue washed with sterile calcium-magnesium free phosphate-buffered saline (PBS; Flow Laboratories Inc., McLean, VA) and then triturated in 35 cm tissue culture dishes (Falcon, Becton Dickinson & Co., Lincoln Park, NJ) containing Eagle's minimal essential media with L-glutamine (MEM; Flow Laboratories Inc., McLean, VA), 10% fetal bovine serum (FBS; Flow Laboratories Inc., McLean, VA) and 100 lU/ml penicillin G, 100 pg/ml streptomycin, and 0.25 pg/ml amphotericin B (Sigma Chemical Co., St. Louis, MO). The tissue was then placed in a humidified incubator at 37°C in 5% C02. The cells were fed twice a week with 10% FBS-MEM. Fibroblasts usually appeared within three to seven days and achieved confluency within two weeks. At that time the medium from each dish was discarded and the cells were washed with PBS. Each dish received 15 ml of 0.025% trypsin solution and was incubated for ten minutes, after which 30 ml of 10% FBS-MEM was added to stop the trypsin reaction. The cell suspension was centrifuged at 1000 rpm for 10 minutes and the supernatant was removed. The pellets were resuspended in PBS, transferred to 150 cm* flasks containing 50 ml of 10% FBS-MEM, and permitted to grow to confluence. Cultures were passaged approximately every other week and screened once every month for Mycoplasma contamination with Hoechst fluorescent antibody 33258 (Mycotest, Flow Laboratories Inc., McLean VA). AU experiments used cells that were passaged three to five times. The cells were kept incubated at 37°C in 5% C02 and the medium was changed twice weekly. was

84

Cell Plating 2

The fibroblasts were trypsinized from the 150 cm' flasks as described. The cells were transferred to a tube and centrifuged at 1000 rpm for 10 minutes. After the supernatant was removed, the pellet of cells was resuspended with PBS so that any remaining media could be washed away. The fibroblasts were then centrifuged and the pellet was again resuspended in PBS. The cell number and viability were determined by staining with 0.008% trypan blue and counting with a hemocytometer. MEM containing 2% FBS was added to bring the final concentration to 1000 cells/100 pi. One hundred pi of the cell suspension were then added to each well of the 96-well tissue culture plates (Costar, Van Nuys, CA).

Drug Treatment Concentrations up to 25 pg (by ten-fold dilutions) per ml of purified anti-TfR-42/6 (Dr. Ian Trowbridge, Salk Institute, La Jolla, CA) dissolved in PBS were added to the culture plates 24 hours after seeding (16). Likewise, concentrations up to 5000 ng (5000-0.00001 pg/ml by ten-fold dilutions) per ml of anti-TfR-PE40 (Dr. David Fitzgerald, NIH, Bethesda, MD) were added to separate culture plates (7). Lastly, concentrations up to 5000 ng (5000-0.00001 pg/ml by ten-fold dilutions) per ml of TGF-a-PE40 (Dr. David C. Heimbrook, Merck Sharp and Dohme, West Point, PA) dissolved in PBS were added to the culture plates 24 hours after seeding (10). As a control, similar tissue culture plates were run simultaneously with the addition of 1000 pg/ml of EGF (Collaborative Research Incorporated, Bedford, MA) to the base media prior to the addition of TGF-a-PE40. Coulter Counter On day 9 the plates were inverted, blotted, and washed with PBS to remove any media and unattached fibroblasts from the wells. Fifty microliters of 0.05% trypsin was added to each well and incubated for 10 minutes. The reaction was stopped by the addition of 100 pi of 10% FBS-MEM. The cell suspension in each well was then added to 9.95 ml of filtered isotone (Fisher Scientific, Pittsburgh, PA). The cell number was measured four times per well by Coulter counter (CC; Model ZM, Coulter Electronics, Inc., Hialeah, FL).

Hexosaminidase Assay The

plates

were

inverted, blotted, and washed twice with PBS. Fifty microliters of 7.5 mM

p-nitrophenyl-N-acetyl-ß-D-glucosamide (Sigma, St. Louis, MO) with 0.25% Triton X-100 in 0.1 M citrate buffer (pH 5.0) was added to the wells and then incubated. After 24 hours the reaction was developed by adding 50 pi of 150 mM glycine/5 mM EDTA buffer (pH 10.5) (Sigma, St. Louis, MO)

to

reader

the wells. The

plates

were

read at 405

nm

by

an

enzyme-linked

immunoabsorbent assay

(Titertek Multiscan, Flow Laboratories, McLean, VA) to determine the optical density (OD).

Statistical

Analysis

Three different cell lines were tested. A set of control wells (zero concentration) was run simultaneously with each experimental plate. The cells were plated in quadruplicate and measurements were then averaged to obtain a mean cell count and a standard deviation at each (log)

concentration.

log inhibitory dose (ID50) was computed as described by Finney (17): Since the data did logit or probit model precisely, (i.e. did not always follow a straight line when plotted on a log-probability scale—see Figure 2) the jD50 was computed in two stages. First, the upper and lower log concentrations that most closely bracketed a 50% response were identified (where a 50% response was that response halfway from the mean of the control wells to the mean of the highest concentration). Since ten fold serial dilutions were performed, the ID50 was bracketed within one order of magnitude. Then, the log ID50 point estimate was determined by linear interpolation between these two closest bracketing doses. (Linear interpolation on a transformed (probit of logit) scale was also performed and gave essentially identical answers since the interpolation was within the range of one log unit.) The ID50 was determined by taking the antilog of the log ID50 for each experimental day and for each drug. The

not

fit

a

85

Hexosaminidase

120

100 c o

o

1/1

c o>

O «

O

*-

Q.

O

.0001

.001

.01

.1 TGF

1

u

-PE40

10

u

100

1000

10000

100

1000

10000

g/ml

Coulter Counter

120

100 o

o

c

O

O

O

O

.0001

TGF-CC-PE40 U g/ml

2. Inhibitory effect of of various concentrations of transforming growth factor a bound to the Pseudomonas exotoxin (TGF-a-PE40) on human Tenon's fibroblasts growth as determined by hexosaminidase and Coulter counter assays.

Figure

86

Hexosaminidase 400

-o-

TGF-aPE40

-•-

TGF-aPE40+EGF

300 o

o

200

100

0

-I—|—.—|—.—|—.—|—i—|—i—|—,—|—,—|—,—r0

0001 .001

01

.1

1

TGF-a -PE40

Coulter 400

10

u

100 10005000

g/ml

Counter

-i

a-

TGF«PE40 TGFaPE40+EGF

300 H o

ü

200 H c

3

O

O

O

100

O

i— —r

100 1000 5000

TGF-

«

-PE40 \i

g/ml

Figure 3. Inhibitory effect of of various concentrations of transforming growth factor a bound to the Pseudomonas exotoxin (TGF-a-PE40) on human Tenon's fibroblasts growth blocked by the addition of epidermal growth factor (EGF) 1000 pg/ml as determined by hexosaminidase and Coulter counter assays (*; cells exposed only to EGF).

87

Summary means, standard deviations and confidence intervals for each group of experiments and each drug were obtained on the log scale since the ID50 values more closely followed a log normal distribution. Therefore, all mean ID50 values were geometric means (antilogs of the mean on a

log scale).

RESULTS Concentrations up to 25 pg of anti-TfR-42/6 per ml in 2% FBS-MEM had no antiproliferative on the fibroblasts as determined by OD or CC. Likewise, anti-TfR-PE40 had no antiproliferative effects in concentrations up to 5000 ng per ml in 2% FBS-MEM utilizing the same assays (results not shown). TGF-a-PE40, as measured by both OD and CC, demonstrated a dose response inhibition of fibroblasts (Figure 2). The ID50 for OD was 32.91 pg/ml with a 95% confidence interval (CI) of 15.59-69.49 pg/ml, while the ID50 for CC was 27.847 pg/ml with a 95% CI of 0.136-568. EGF was used as a negative control for TGF-a-PE40. The inhibitory effect of the toxin conjugate was completely blocked by the addition of 1000 pg/ml of EGF (Figures 3, 4). effect

DISCUSSION In order to determine the feasibility of utilizing monoclonal antibody therapy to inhibit subconjunctival fibroblast proliferation, we performed preliminary work on the expression of transferrin receptor on the growth phase of fibroblasts and their responses to EGF (18). Briefly, we found that transferrin receptors (TfR) on human fibroblasts from Tenon's capsule were inducible with EGF. We found an average of 40-60% baseline TfR expression upon induction with EGF, TfR expression rose sharply to approximately 84% after 2 minutes. These data agreed with the turnover data reported previously using I-transferrin on teratocarcinoma stem cells (19).

We attempted to modulate cell proliferation with the use of anti-TfR-42/6. Concentrations up 25 pg per ml of purified antibody in 2% FBS-MEM had no antiproliferative effect on the fibroblasts. Lower concentrations of anti-TfR-42/6 (as little as 2.5 pg per ml) have been observed to have significant inhibitory effect on the growth of CCRF-CEM cells (16). Subsequently, using anti-TfR-PE40, we were not able to modulate cell proliferation with concentrations up to 5000 ng per ml in 2% FBS-MEM. Variable results have been reported with different cell lines tested: An ID50 of 4.0 ng/ml of anti-TfR-PE40 was found on A431 cells, while no cytotoxic effect was found on murine Swiss 3T3 cells with concentrations up to 2 pg/ml of anti-TfR-PE40 (7). Since TfR are expressed in human ocular fibroblasts and EGF induces an increase in TfR expression, we hypothesized that fibroblast proliferation might be inhibited by delivering toxins which bind to the EGF receptor. TGF-a binds to the EGF cell surface receptor (10), forming a receptor ligand complex which is then taken up via coated pits and internalized by endocytosis (20, 21). Both TGF-a and EGF bind to and activate the EGF receptors in an identical manner (10). TGF-a-PE40 binds 100-fold less effectively to the EGF receptor than does TGF-a (7). The cytotoxic effect of TGF-a-PE40 has been found to be very close to that of Pseudomonas exotoxin. The 50% effective cytotoxic effect of Pseudomonas exotoxin and TGF-a-PE40 on A431 and HeLa cells were both found to be about 1 ng/ml. No cytotoxic effects were exhibited with concentrations up to 2xl05 pM of Pseudomonas exotoxin without the recognition domain (7). We performed a dose response curve of TGF-a-PE40 on human fibroblasts using 5000 pg/ml as the highest concentration. We found a ID50 of approximately 30 pg/ml. (The ID50 of A431 cells as determined by H-leucine incorporation was found to be about 1 ng/ml). To confirm that the cytotoxicity was due to the internalization of the receptor, we performed the same dose response curves in the presence of excess EGF, and found that the cytotoxic effect of TGF-a-PE40 was blocked by the excess EGF. The greatest advantage of utilizing growth factor toxin conjugates as pharmacological agents is that they can be concentrated on proliferating cells and spare the non-proliferating cells. The cell surface receptors are readily accessible to pharmacological agents such as TGF-a-PE40. One can take advantage of a normal physiologic function of the cell—the formation of a receptor ligand complex which is then taken up via coated pits and internalized by endocytosis-to deliver toxins. Since the EGF receptors are not known to be species specific, TGF-a-PE40 may also be tested in vivo on different species prior to human testing (22). TGF-a-PE40 has the potential of being an inhibitor of fibroblast proliferation and may be used in the modulation of wound healing. to

88

ACKNOWLEDGMENTS This research was supported in part by NIH grants EY 07701 and EY 00331, and by the Lucille Ellis Simon Glaucoma Research Fund. The authors gratefully acknowledge the assistance of Dr. Ian S. Trowbridge for the anti-TfR-42/6, to Dr. David Fitzgerald for the anti-TfR-PE40, to Dr. David C. Heimbrook for the TGF-a-PE40, and to Dr. Jeff Gornbein, for his help in the statistical

analysis.

BIBLIOGRAPHY

1.

Addicks, E.M., Quigley, H.A., Green, W.R, and Robin, A.L. Histological characteristics of filtering blebs in glaucomatous eyes. Arch Ophthalmol 101:795-8, 1983.

2.

Lee, D.A., Leong, K.W., Panek, W.C., Eng, CT. and Glasgow, BJ. The use of bioerodible polymers and 5-fluorouracil in glaucoma filtration surgery. Invest Ophthalmol Vis Sei

29:1962-1967, 1988.

3.

Tahery,

4.

Ross, R. The fibroblast and wound repair. BjoJ Rev 43:51-96, 1968.

5.

Boss, B.D., Langman, R., Trowbridge, I.S. and Dulbecco, R. In Monoclonal Antibodies and

M. and Lee, D.A. Review: Pharmacologie control of wound filtration surgery. .T Ocular Pharm 5:155-179. 1989.

Cancer: Therapeutic Potential of Monoclonal Antibodies that Block Academic Press, Inc., p. 53-61, 1983.

6.

Trowbridge, I.S. and Lopez, F. Monoclonal antibody binding and inhibits human tumor cell growth in vitro. 1989.

healing

in

Biological

glaucoma

Function.

to transferrin receptor blocks transferrin Proc Nati Acad Sei USA 86:8545-8549,

7.

Batra, J.K., Jinno, Y., Chaudhary, V.K., Kondo, T., Willingham, M.C., Fitzgerald, DJ. and Pastan, I. Antitumor activity in mice of an immunotoxin made with anti-transferrin receptor and a recombinant form of Pseudomonas exotoxin. Proc Nati Acad Sei USA 86:8545-8549, 1989.

8.

Edwards, G.M., DeFeo-Jones, D., Tai, J.Y., Vuocolo, G.A., Patrie, D.R., Heimbrook, D.C. and Oliff, A. Epidermal growth factor receptor binding is affected by structural determinants in the toxin domain of transforming Cell Biol 9:2860-2867. 1989.

9.

growth factor-alpha-Pseudomonas

exotoxin fusion

proteins. Mol

Pastan, I. and Fitzgerald, D. Pseudomonas exotoxin: Chimeric toxins. Minireview. J Biol Chem 264:15157-15160. 1989.

10.

Todaro, G.J., Fryling, C. and DeLarco, J.E. Transforming growth factors produced by certain tumor cells: Polypeptides that interact with epidermal growth factor receptors. Proc Nati Acad Sei USA 77:5258-5262. 1980.

11.

Lee, D.A., Tehrani, S.S., Stephenson, T. and Kitada, S. The effect of fluorinated pyrimidines FUR, FUdR, FUMP and FdUMP on human Tenon's fibroblasts. Invest Ophthalmol Vis Sa 32:2599-2609, 1991.

12.

Wong, V.K., Shapourifar, T.S., Kitada, S., Choo, P.H. and Lee, D.A. Inhibition of rabbit ocular fibroblast proliferation by 5-fluorouracil and cytosine arabinoside. I Ocular Pharm 1:27-39, 1991.

13.

Givens, K.T., Kitada, S., Chen, A.K., Rothshiller, J. and Lee, D.A. Proliferation of human ocular fibroblasts. Invest

14.

Ophthalmol Vis Sei 31:1856-1862.

1990.

Givens, K.T., Lee, DA., Rothschiller, J., Kitada, S., Larian, B. and Cortes, A. Antiproliferative

drugs and human ocular fibroblasts: Colorimetric 9:599-606, 1990.

89

vs.

cell

counting

assays. Curr Eye Res

15.

Lee, D.A., Shapourifar, T.S. and Kitada, S. The effect of 5-fluorouracil and cytarabine human fibroblasts from Tenon's capsule. Invest Ophthalmol Vis Sei 31:1848-1855. 1990.

16.

Trowbridge, I.S. and Lopez, F. Monoclonal antibody to transferrin receptor blocks transferrin binding and inhibits human tumor cell growth in vitro. Proc Nati Acad Sei USA 79:1175-1179,

on

1982.

17.

Finney, DJ. Quantal responses and the tolerance distribution. In Statistical Methods Biological Assay (3rd ed.) Finney DJ, ed. Charles Griffin and Co., London, p. 349-369, 1978.

18.

Smyth, R.J., Kitada, S. and Lee, D.A. Expression of transferrin receptors from the human eye. Invest Ophthalmol Vis Sei 31(Suppl):63. 1990.

19.

Karin, M. and Mintz, B. Receptor-mediated endocytosis of transferrin in developmentally totipotent mouse teratocarcinoma stem cells. J Biological Chem 256:3245-3252, 1981.

20.

Mendelsohn, J. Growth factor Prog Allergy 45:147-160. 1988.

21.

Trowbridge, I.S. Monoclonal antibody therapy: target. Prog Allergy 45:121-146. 1988.

22.

Blakey, D.C., Wawarzynczak, E.J., Wallace, P.M. A perspective. Prog Allergy 45:50-90. 1988.

Received:

Accepted

as

capsule of

targets for antitumor therapy with monoclonal antibodies.

September 13, 1991 for Publication:

tenons

in

November

Transferrin receptor and

Thorpe,

P.E.

as a

potential therapeutic

Antibody toxin conjugates:

21, 1991 Reprint Requests:

David A. Lee, M.D. Jules Stein Eye Institute UCLA School of Medicine 100 Stein Plaza Los Angeles, CA 90024-7004

90