0 Aeolus Press Ocular Immunology and Inflammation 0927-3948/93/US% 3.50 (Accepted 31 March 1993)

Transforming growth factor+ levels in aqueous humor during experimentally induced uveitis GUILLERMO ROCHA,' MALCOLM G. BAINES,'y2 JEAN DESCHENES,'

ALAIN J. DUCLOS2 and EMILIA ANTECKA*12

Departments of Ophthalmology and 2Microbiology Q Immunology, McGill University, MontrPal, QuPbec

ABSTRACT. The anterior chamber of the eye is known to be a site of immune privilege. Particu- larly, the aqueous humor (AqH) appears to possess unique immunoregulatory properties. The authors have previously shown that human AqH (HAqH) may increase or decrease the prolifera- tion of different cell types. Although no single factor has been established as solely responsible for these effects, much attention has been given to the 24-30 kD fraction of AqH, where trans- forming growth factor-beta (TGF-@) is found. The purpose of this study was to determine the changes occurring in the rabbit AqH (RAqH) in relation to intraocular inflammation. Heterolo- gous lens or human serum albumin (HSA) immunization-induced uveitis models were used in two groups of New Zealand albino rabbits to study the relationship between uveitis and TGF-@. AqH and serum samples were obtained serially before, during and after the induction of ocular inflam- mation. Systemic humoral immunity to HSA or lens antigens was monitored using enzyme-linked immunosorbent assay (ELISA). A mink lung epithelial cell (CCL-64) bioassay for TGF-@ was used to quantify the amount of this cytokine in RAqH. TGF-@ levels in RAqH increased fourfold after the first immunization. A sharp decrease in RAqH TGF-@ levels was found in association with the development of acute intraocular inflammation. The implications of this finding to the etiology of uveitis are discussed.

Key words: experimental uveitis; aqueous humor; TGF-0; immunosuppressive factors; ACAID

INTRODUCTION

Recent advances in ocular immunology have provided us with newer concepts regarding the pathophysiology of uveitis. The eye is an im- munologically privileged site1v2 due to proper- ties such as the presence of a highly selective

* Correspondence to: Jean Deschgnes, MD, FRCSC, Department of Ophthalmology, Royal Victoria Hospital, 687 Pine Avenue West, Room E4.60, Montreal, QC H3A IAl, Canada.

blood ocular barrier,3 the sequestration of reti- nal antigen^,^ and the presence of inhibitory factors in aqueous humor (AqH).5 The pheno- menon known as anterior chamber-associated immune deviation (ACAID)6 which induces a systemic, antigen-specific inhibition of delayed hypersensitivity,' imposes yet another form of immune protection on the eye.

Low concentrations of non-specific immuno- modulatory factors found in AqH have been shown to exert significant anti-proliferative ef- fects on various cell types. Two groups of fac-

Ocular Immunology and Inflammation - 1993, Vol. I , No. 4, pp. 343-354 0 Aeolus Press Buren (The Netherlands) 1993

343

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

tors have been reported, one consisting of pro- teins found at or below the 3.5 kDa range,8 another one in the 25-30 kDa range. It is within the latter range that transforming growth factor- @ (TGF-@), the most important of these factors, is f o ~ n d . ~ - l l TGF-@ is a multifunctional family of peptides that regulate diverse functions in many cells including proliferation and differen- tiation.I0 Different cells, including those within the normal iris and ciliary body, have been shown to produce TGF-@2, and most cells ex- press a high-affinity TGF-@ receptor.I2-l5 Five genetically distinct isoforms of TGF-@ have been described, however 01, @2 and 03 are the human isoforms, and it seems that most of the biologic activity is due to TGF-02. 16J7 TGF-8 has been recently implicated as a negative regu- lator of cell growth in the immune system.ll TGF-@ has also been shown to have ACAID- inducing effects on the ‘alternative antigen presenting cells’ (APCs)-F4/80+- of the anteri- or chamber (AC).15918.19 TGF-@2 may also have a role in promoting the healing of ocular tis- sues.2o It is logical to assume, then, that TGF-@ should play an important role in maintaining the privileged microenvironment in the AC, the primary purpose of which is to preserve the in- tegrity of the visual axis by suppressing or regulating acute intraocular inflammatory reactions.

We hypothesized that TGF-@ levels should be negatively associated with acute intraocular in- flammation. The purpose of this study was to as- sess the levels of TGF-@ in the rabbit AqH (RAqH) during experimentally-induced uveitis.

MATERIAL AND METHODS

Animals

Female New Zealand rabbits weighing 1.5-2.0 kg were randomly distributed in four groups of

three animals each. Animals were housed at the animal care facilities of McGill University and procedures were done in accordance with the guidelines of the Canadian Council for Animal Care, and the Association for Research in Vision and Ophthalmology (ARVO) resolutions for the use of laboratory animals in research. When needed, general anesthesia was achieved with 35 mg/kg ketamine and 5 mg/kg xylazine by in- tramuscular in jection.

Induction of uveitis

Two uveitis models were used: heterologous lens antigen-induced and human serum albumin (HSA)-induced uveitis.

a. Lens-induced uveitis

Group I . Animals received one intramuscular injection of porcine lens protein (PLP) (10 mg in 0.4 ml) with Freund’s complete adjuvant (1 : 1). Ten days later, a right anterior capsulotomy was done under ketamine-xylazine anesthesia. Three weeks later, a left anterior capsulotomy was done. Using this model an intense uveitis was seen in the left eye approximately five to seven days following the second capsulotomy.

Group 11. Lens-induced uveitis with unilateral capsulotomy was essentially as described above, except that the second capsulotomy was not per- formed. A mild and transient uveitis was seen only in the capsulotomized eye. The non- capsulotomized eye served as an internal control.

b. HSA induced uveitis

Group 111. An intravitreal injection of 30 p1 of human serum albumin (HSA) (Cutter Biologi- cal, Berkeley, CA) (20% solution) was given to

344

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

Aqueous TGF-0 levels in experimental uveitis

the right eye under ketamine-xylazine anesthe- sia. The left eye was injected with 30 p1 of sterile saline. A mild primary inflammatory response was elicited in the right eye only. After three weeks, the animals were challenged intravenous- ly with HSA (1 mg/kg). A distinct anterior uvei- tis was observed within one week following sys- temic immunization.

c. Controls

Group ZV. Rabbits received systemic saline in- jections in place of PLP or HSA.

Evaluation of uveitis in rabbits

Inflammation was assessed clinically by slit- lamp examination, and was scored from ‘0’ to ‘4’ according to a modification of the system reported by Hogan,21 with ‘0’ being ‘no inflam- mation’. ‘ 1 ’ represented minimal conjunctival hyperemia, no corneal edema and no fibrin seen in the AC. ‘2’ and ‘3’ represented moderate degrees of inflammation, with evident conjunc- tival hyperemia, fibrin in AC and either no cor- neal edema (‘2’) or corneal edema (‘3’) present. ‘4’ was used to describe severe conjunctival in- jection, and poor visualization of AC structures due to significant corneal edema and fibrin in the AC (Table 1).

Aqueous humor sampling

Samples of AqH were obtained by paracentesis under anesthesia and blood samples were taken every week until the conclusion of the ex- periment.

RAqH was collected from the AC of each eye with a 1 cc syringe and a 27-gauge needle using standard paracentesis technique. Weekly speci- mens of approximately 125 p1 of AqH were ob-

tained from each eye. RAqH was placed in siliconized tubes and stored in a freezer at -80°C until assayed. Serum specimens were kept frozen at -20°C until used.

Bioassay for TGF-8

The CCL-64 mink lung epithelial cell (Mv 1 Lu) bioassay for TGF-P was modified from previ- ously described methods. 18122323 Briefly, the cells were kept at 37°C in 5% CO,, maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Grand Island, NY) sup- plemented with 10% fetal calf-serum (FCS) and 1 Yo sodium pyruvate, and passaged at three-day intervals. To initiate the growth inhibition as- say, the cells were recovered after incubation with trypsin (1 .O mg/ml) and ethylene-diamine- tetra-acetic acid (EDTA, 1 mM; GIBCO), pellet- ed at 500 g for 10 min, followed by two washes in mink assay medium (DMEM supplemented with 5% FCS, 10 mM HEPES, penicillin and streptomycin). Test reagents consisting of 25 pl of heat-activated RAqH (1 :3 dilution), serum, TGF-P (2.5 ng/ml) or PBS were added in tripli- cate to 25 pl of mink assay medium in 96-well flat bottomed microtiter plates and serially diluted. To activate latent TGF-0, RAqH was diluted 1 :3 with bovine serum albumin (BSA 50 pg/ml), and placed in a hot water bath at 80°C for 8 min. Recombinant human TGF-P2 (Gen- zyme Corporation, Cambridge, MA) used as a standard was diluted to a concentration of 2.5 ng/ml with BSA (50 pg/ml). The TGF-0 neu- tralizing antibodies, monoclonal mouse anti- TGF-P1,2,3 and anti-TGF-p2,3 (Genzyme Cor- poration, Cambridge, MA), were diluted with mink assay medium to a concentration of 20 pg/ml and 30 pg/ml, respectively. Finally, 25,000 of the sub-confluent CCL-64 cells in 25 p1 of mink assay medium were added to each well.

345

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

TABLE 1 . Clinical scores of inflammation.*

Weeks post-immunization

0 1 2a 3 4b,c 5 6 7 8

Control, OD + 0 s 0 0 0 0 0 0 0 0 0 Group I, OD 0 0 0 1.3 0 0 0 0 0 Group I , 0 s 0 0 0 0 0 1.6 1.6 1 2 Group 11, OD 0 0 0 1.3 0 0 0 0 0 Group 11, 0 s 0 0 0 0 0 0 0 0 0 Group 111, OD 0 1.3 2 0 0 1.6 1.6 1 2 Group 111, 0 s 0 0 0 0 0 0 0 0 0

a: First capsulotomy performed in Groups I and 11, OD. b: Second capsulotomy performed in Group I , 0s. c: Intravenous HSA challenge in Group 111. * Refer to the methodology section for details of the scoring system.

Group I: Week 0, 10 mg PLP + Freund’s complete adjuvant (FCA) IM. Week 2, capsulotomy OD. Week 4, capsulotomy 0s.

Group 11: Week 0, 10 mg PLP + FCA IM. Week 2, capsulotomy OD.

Group 111: Week 0, 6 mg HSA intravitreally OD, saline 0s. Week 4, HSA Img/kg IV

An additional 25 p1 of mink assay medium with or without neutralizing anti-TGF-P antibodies were added in order to confirm the specific na- ture of the inhibitory factors. After an incuba- tion period of 3 hr at 37°C in 5% CO,, the plates were pulsed with 0.5 pCi of 3H-thymidine for 16 hr. The supernatants were discarded and replaced with trypsin and EDTA for 30 min to detach the mink lung cells; then the samples were harvested on a multichannel plate harvester (Skatron, AG), and the 3H-thymidine uptake was quantified using a Beckmann 8000 beta scin- tillation counter.

Computer analysis of results of TGF-8 bio- assays

Data analysis was performed in the following way. The triplicate results for each data point were used to compute the concentration of TGF-

in the AqH. The analysis was performed on a

486DX-33 computer running MS-DOS 5.0 and CSS/Statistica 3.0 E (Statsoft, Inc.). A non- linear curve-fitting analysis on the data using the following equation was performed: y = Ax + Bx2 + Cx3 + Dx4 + K (where ‘x’ represents the logarithm of the TGF-P concentration in ng/ml and ‘y’ the thymidine incorporation in CPM). Non linear regression analysis provided the best curve-fitting for our data. The computer was used to determine ‘x’ (the specific concentration of TGF-0) from a corresponding ‘y’ (CPM) for each AqH specimen using a simple BASIC program.

Standard methods for the enzyme-linked im- munosorbent assay (ELISA) were employed as previously described.% Serum obtained from groups I and I1 was tested for the presence of an- tibodies against porcine and rabbit lens protein

346

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

Aqueous TGF-P levels in experimental uveitis

using a goat anti-rabbit IgG conjugated with peroxidase (Sigma, St. Louis, MO). Group I11 sera were tested for antibodies against HSA us- ing the same protocol.

RESULTS

Induction of uveitis

In order to induce uveitis in the rabbits, it was necessary to focus an immune response upon an- tigens present within the eye. In the two models, the immune response was induced to en- dogenous autologous lens proteins or to an in- travitreally injected foreign protein. Whereas the former model could mimic an autoimmune uveitis, the latter was more representative of a uveitis induced by the presence of an exogenous pathogen in the eye. Two peaks of ocular in- flammation were observed clinically both in the lens- and HSA-induced uveitis models, cor- responding to the time after each immunologic challenge. Inflammation was assessed clinically by slit-lamp examination, and was graded as described. Similar scores were observed in both lens-induced groups when compared to the HSA group, and the three experimental groups showed significantly higher scores than controls. (Table 1). Eyes receiving intravitreal injection of saline were identical to uninjected eyes.

Serum antibody response to uveitogenic proteins

Serial measurement of serum antibodies showed a progressive increase in the titers to the antigen used in each group. Titers started from a base- line level of ‘zero’ for all groups, to maximum titers against the immunizing antigen of 12 x lo6 for the lens-induced group I with bilateral capsulotomy, 10 x lo6 for the group I1 with unilateral capsulotomy and 6 x lo6 for group 111. Anti-PLP titers peaked at 26 days after the

primary immunization and 20 days after the se- cond capsulotomy (Group I). The two peaks of anti-PLP antibodies observed correlated in time with the two peaks of ocular inflammation ob- served clinically (Fig. 1). Anti-HSA antibody titers peaked about one or two weeks after the secondary immunization, also corresponding in time to the ocular inflammation observed clini- cally. No antibody rise was observed after the in- itial intravitreal challenge. These results show that higher antibody titers to eye-associated an- tigens are associated with uveitis.

The release of autologous lens proteins by the second capsulotomy (Group I) stimulated anti- body production to lens protein. Conversely, significant titers of antibodies alone were not sufficient to induce uveitis, implying that ocular inflammation was regulated by other factors.

Quantification of TGF-/3 in aqueous humor

Since TGF-P can suppress lymphoprolifera- tion, cytokine production and inflammation, and since AqH contains significant quantities of TGF-P, the relationship between TGF-P activity and uveitis was investigated. The effect of AqH from normal eyes on cell proliferation was as- sessed. The results obtained with the CCL-64 cultures show a similarity in the shape of the curve for recombinant TGF-0 and AqH (Figs. 2 and 3). Higher concentrations of TGF-0 or AqH were associated with a significant reduction in DNA synthesis by the mink lung cells and lower 3H-thymidine uptake. When neutralizing anti- bodies were used, cell counts in both assays were shown to increase significantly, thus indicating that a great part of the growth inhibiting proper- ties of AqH from non-inflamed eyes was due to TGF-/3 (Table 2). Using the mink lung assay the concentration of TGF-P in normal rabbit AqH was found to be 2.72 k 0.61 ng/ml.

Uveitis was induced in rabbits and the levels of

347

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

t 2nd capsulotomy

0 0 1 2 3 4 5 6 7 a

1 W n k r pool-lrnmunkatkm

Fig. I . Serum antibody titers to porcine lens protein in lens-induced uveitis and to human serum albumin in HSA-induced uveitis. Antibody titers increased in all groups following immunological challenge. Each point on the curve represents the mean of three animals in each group.

TABLE 2. Inhibition of TGF-8 bioactivity by nautralizing antibodies

Neutralizing anti-TGF-b added to assay rHTGF;B2 RAqH

Activity of TGF-8 in cpm

(0.312 ng/ml) (1:24 dilution)

None 36,290 40,339

Anti-TGF-0 2,3 59,946 72,679 (To neutralization) (65%) (80%)

Anti-TGF-8 1,2,3 70,025 66,043 (Vo neutralization) (93%) (63 '7'0)

TGF-/3 in AqH were monitored. A greater varia- bility in the TGF-P concentration was present in inflamed versus non-inflamed eyes in the lens protein-induced model with unilateral capsulo-

tomy (data not shown). In the lens protein- induced group with bilateral capsulotomies, im- munization with PLP increased the basal level of TGF-/3 in the AqH of these rabbits to 8 ng/ml. This four-fold increase was sustained for three weeks until the second capsulotomy was per- formed. A sharp decline of TGF-/3 levels to con- trol levels was observed in this group following the second capsulotomy which was associated with the time when the elicited uveitis was clini- cally most apparent (Fig. 4). The AqH levels of TGF-/3 in rabbits immunized with HSA also showed the same sharp decline soon after the time of the secondary immunological challenge coincident with the occurrence of ocular inflam- mation (Fig. 5) .

To confirm that the effects observed on cell

348

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

Aqueous TGF-0 levels in experimental uveitis

- T - TGF-Q

7oooo -0- TGF-Q + ontl TGF-0 12.3 - TGF-Q + anti TGFD 2.3

0.039 0.078 0.156 0.312 0.625 1.25 2.5 2 TGF-81 (ng/ml)

Fig. 2. Inhibition of mink lung cell proliferation by recombinant TGF-P2. Increased concentrations of TGF-8 led to a decrease in cell proliferation. Neutralization of TGF-0 produced an increase in cell 3H-thymidine uptake. Each point on the curve represents the mean of three animals in each group.

proliferation were due to the presence of TGF-0, diluted AqH (1:6) obtained at 32 days post- immunization was placed in the cell cultures, in the presence of neutralizing antibodies (anti- TGF-/32,3). The results showed that for all groups, there was an increase in cell prolifera- tion which was more than double of that ob- tained without the neutralizing antibodies (Fig. 6) . These findings support the contention that AqH from inflamed eyes exerts powerful anti- proliferative effects mainly through the actions of TGF-0.

DISCUSSION

In summary, we have shown that the sensitiza- tion of rabbits to lens antigens or intravitreal

HSA resulted in a four-fold increase in basal TGF-0 concentration in AqH followed by a sig- nificant decrease in TGF-0 levels in association with the development of overt ocular inflam- mation.

By eliciting inflammation using lens protein and HSA, ocular immunologic privilege is breached in several ways. Although these models induce a significant or distinct uveitis, the mechanisms that lead to it are diverse and difficult to analyze individually. By performing serial AC paracenteses or by the inoculation of HSA to the vitreous, the blood-aqueous barrier is physically broken. Furthermore, the antigens injected (PLP and HSA) may have contained factors such as bacterial lipopolysaccharides (endotoxin) which could stimulate TGF-0 pro-

349

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

la300

0

\ - RAqH + anti TGF-0 12.3 -- - RAqH + antl TGF-Q 2.3

1 I I I , I 1 I I I I 1 I

Fig. 3. Inhibition of mink, lung cell proliferation by rabbit aqueous humor. 3H-thymidine uptake was similar to that seen with TGF-P. Likewise, the addition of anti-TGF-P antibodies increased cell proliferation, confirming that a great part of the anti-proliferative effects of AqH are due to TGF-P2 or 3 . Each point on the curve represents the mean of three animals in each group.

duction. Finally, the PLP was injected with Freund’s complete adjuvant, an agent well known to stimulate a mild uveitis and arthritis. Any one of these factors could account for a general increase in TGF-@ production in the treated and untreated eyes in the immunized rab- bits. At this time there is no evidence of severe inflammation in the eyes of either group of rab- bits. This data indicates that the dramatic in- crease in TGF-@ in AqH in response to local or systemic immune stimulation may be important in preventing uveitis and loss of vision. Follow- ing secondary immune stimulation with HSA or by capsulotomy, significant uveitis was evident. Further, a more severe uveitis was seen to be coincident with a marked decline in TGF-@ ac- tivity in AqH. The sudden exposure of lens-

protein to the AC microenvironment may bring about a series of reactions which might hamper the production or effectiveness of the inhibitors normally present in AqH either by binding, inac- tivation or suppression of synthesis.

While it is not possible at this time to deter- mine the mechanisms of TGF-@ loss, all the eyes which developed inflammation showed a signifi- cant decrease in TGF-@ levels. This was true in instances when lens proteins were exposed to the AC environment by capsulotomy, as well as when they were not, as in the case of the HSA- induced uveitis. Serial paracentesis was per- formed in all groups, yet controls only manifest- ed a slight increase in the basal level over time but not the sharp decrease in TGF-@ levels that was seen in the experimental eyes. This indicates

350

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

Aqueous TGF-P levels in experimental uveitis

12

10

8 ' 6 - - 9 g

4

2 --

-

--

--

--

-a- Group I, 00

-0- Group 1.05

---C Control

2nd copsulotomy (03

1st capsubtomy (OD) i

" I I I I 1

0 1 2 3 4 5 6 7 8 9 4 w..kspod-hmmmlralkn

Fig. 4. TGF-6 levels in lens-induced uveitis. The decreased TGF-0 levels correspond to the time when the highest clinical in- flammation occurred, and with the highest antibody titers. OD: right eye. 0s: left eye. Each point on the curve represents the mean of three animals in each group.

that intermittent trauma has little effect on TGF-P production. The type of inflammation induced did not seem to be a factor either, for es-

P levels in the first capsulotomized eye of the lens-induced uveitis group did not decline initial- ly, when mild inflammation was present, yet

sentially the same results were obtained with an autoimmune granulomatous (lens-induced) reaction and a non-specific (HSA-induced) im- mune response. There was no correlation be- tween the antibody levels and the severity of uve- itis as immunization without capsulotomy produced no uveitis (data not shown). However, there was a correlation between the severity of uveitis and the decrease in TGF-6 levels. The decrease was significant in all experimental groups, and was greater for the HSA model of uveitis even though the antibody titers against HSA were not as high as those against PLP. Another interesting finding is the fact that TGF-

subsequently decreased in association with the decrease in levels of TGF-(3 when the second cap- sulotomy was performed, and inflammation de- veloped, in the left eye. At this time, it is difficult to clarify whether systemic influences or other factors are involved in this phenomenon.

When TGF-(3 is produced in the eye,19*20 it is secreted in a latent form which is converted to its active form by mechanisms which are not yet clear but might include an acidic microenviron- ment and protease cleavage which may accom- pany i n f l a m m a t i ~ n . ~ ~ Little is known regarding the role of TGF-P during active ocular inflam- mation. To our knowledge, no other reports ad-

35 1

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

- 9 Intravenous HSA - Group 111. OS

--

--

I I I 1 I I I 1 I I I I I I I

a

7

6

F 5

9 , P P

1

0 0 1 2 3 4 5 6 7 8 9

5 Weekspod-lnwnunizallon

Fig. 5. TGF-8 levels in HSA-induced uveitis. The decreased TGF-8 levels correspond to the time when the highest clinical inflammation occurred, and with the increase in antibody titers. OD: Right eye. 0s: Left eye. Each point on the curve represents the mean of three animals in each group.

dress this association. Nevertheless, it could probably be assumed that due to its anti- proliferative effects, TGF-@ should have a pro- tective effect, either by decreasing intraocular cell proliferation, or by its role in ACAID- induction. In this regard an increase in the levels of TGF-@ should have been found. The fact that TGF-@ levels decreased during overt ocular in- flammation raises two interesting issues. The cause of the lower TGF-@ levels might have been due to either local or systemic inhibition of TGF- @ production. On the other hand, it could have been due to increased protease mediated cleavage of TGF-@, resulting in increased bind- ing to cellular receptors (available in greater number due to the intraocular ingress of cells during inflammation) and higher molecular

weight proteins. If TGF-0 is indeed an impor- tant protective factor in AqH, further produc- tion of it by cells of the iris and ciliary body may lead to a compensatory rise in its levels after the observed decrease. This return to pre-uveitis TGF-@ levels might also be related to the resolu- tion of the inflammation. In addition, it has been reported that cytokines such as IL-1, IL-2 and IL-4 interfere with the growth inhibiting properties of TGF-@ and abrogate ACAID.’ 1*26

For ACAID to be induced, ‘alternate (no class I1 molecule expression) antigen presenting cells’ (APCs) must be present in the AC. The subin- flammatory state produced by inoculating the eye gamrna-interfer~n,~~ and the granuloma- tous reaction that occurs to lens protein^^**^^ may allow pro-inflammatory cytokines to exert

352

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

Aqueous TGF-P levels in experimental uveitis

2oooo r I 7

18000

16000 v

il l4Ooo 4 12000 Q

n

r

.E 10000

*E 8000 ' 6000

4000 i5

2000

Alone Anti TGF-I3 293

I 6 TG F-0 Control Group I Group I1 Group 111

Fig. 6. Inhibition of TGF-0 bioactivity at five weeks post-immunization. TGF-0 concentration is 1.25 ng/ml. RAqH for all other groups is at a 1:6 dilution. Each bar on the curve represents the mean of three animals in each group.

their uveitogenic effects, and for classic APCs to enter the eye, thus further compromising ocular protective mechanisms. Furthermore, a2-Ma- croglobulin has been shown to be a biological regulator of TGF-P1 and 2 actions in v ~ v o . ~ O

Any one of these factors could have also been responsible for the decrease in TGF-P levels seen in this study.

Another possibility that must be explored is whether the bioassay employed was detecting only TGF-p.31 When comparing our range of values to those found in the literature,I6 similar

In conclusion, TGF-P is just one of the multi- tude of mechanisms which are present in the AC to maintain ocular immune privilege. Further research is warranted to better clarify the role of TGF-/3 during ocular inflammatory states. De- termining which factors increase or decrease its production may provide a better understanding of inflammatory states, thus allowing for a more precise management of uveitis.

ACKNOWLEDGEMENTS

levels of TGF-P in non-inflamed A ~ H were ob- served but no reports On TGF-P levels during ocular inflammation were found.

We are grateful to Maureen O'Connor, PhD (Biotechnology Research Institute-NRC, Montreal, Canada) for providing us with the CCL-64 cell line, and to Alexandra Hawiger (Genzyme Corporation) for her scientific assistance.

REFERENCES

1. Franklin RM, Prendergast RA. Primary injection of skin allografts in the anterior chamber of the rabbit eye. J Immunol

2. Wilbanks GA, Streilein JW. The differing patterns of antigen release and local retention following anterior chamber and intravenous inoculation of soluble antigen. Evidence that the eye acts as an antigen depot. Reg-Immunol 1989;

1970; 104:463-469.

2(6):390-398.

353

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.

G. Rocha et al.

3. Brubaker RF. Flow of aqueous humor in humans. Invest Ophthalmol Vis Sci 1991; 32(13):3145-3166. 4. Wacker WB, Lipton MM. Experimental allergic uveitis. I. Production in the guinea pig and rabbit by immunization with

retina in adjuvant. J Immunol 1968, 101:151. 5 . Streilein JW, Kaiser CJ. Immunosuppressive properties of aqueous humor as it relates to uveitis. In: Belfort Jr R, Petrilli

AMN, Nussenblatt R, editors. Proc First World Uveitis Symposium. S lo Paulo: Livraria Roca Ltda. 1989: 3-13. 6. Streilein JW. Anterior chamber associated immune deviation: The problem of immune privilege in the eye. Surv

Ophthalmol 1990; 35;67-73. 7. Streilein JW, Niederkorn JY. Induction of anterior chamber-associated immune deviation requires an intact, functional

spleen. J Exp Med 1981; 153:1058-1067. 8. Granstein RD, Staszewski R, Knisely TL, Zeira E, Nazareno R, Latina M, Albert DM. Aqueous humor contains trans-

forming growth factor-@ and a small (< 3500 daltons) inhibitor of thymocyte proliferation. J Immunol 1990; 144(8):302 1 -3027.

9. Massague J. The TGF-@ family of growth and differentiation factors. Cell 1987; 49:437-438. 10. Sporn MB, Roberts AB, Wakefield LM, Assoian RK. Transforming growth factor-@: Biological function and chemical

11. Ellingsworth LR, Nakayama D, Segarini P, Dasch J, Carrillo P, Waegell W. Transforming growth factor& are equipo-

12. Runyan TE, McCombs WB, Holton OD, McCoy CE, Morgan MP. Human ciliary body epithelium in culture. Invest

13. Kondo K, Coca-Prados M, Sears M. Human ciliary epithelia in monolayer culture. Exp Eye Res 1984; 38:423-433. 14. Streilein JW, Bradley D. Analysis of immunosuppressive properties of iris and ciliary body cells and their secretory

products. Invest Ophthalmol Vis Sci 1991; 32:2700-2710. 15. Wilbanks GA, Streilein JW. Suppression of delayed hypersensitivity in vivo by an ACAID-inducing signal created in

vitro. Invest Ophthalmol Vis Sci 1992; 33(4):1283. 16. Cousins SW, McCabe MM, Danielpour D, Streilein JW. Identification of transforming growth factor-beta as an im-

munosuppressive factor in aqueous humor. Invest Ophthalmol Vis Sci 1991; 32:2201-2211. 17. Rocha G, Baines MG, DeschZnes J , Antecka E, Non-specific suppressor factors associated with uveitis. Can J

Ophthalmol 1992; 27(2):101. 18. Helbig H, Kittredge KL, Coca-Prados M, Davis J, Palestine AG, Nussenblatt RB. Mammalian ciliary-body epithelial

cells in culture produce transforming growth factor-beta. Graefe’s Arch Clin Exp Ophthalmol 1991; 229:84-87. 19. Knisely TL, Bleicher PA, Vibbard CA, Granstein RD. Production of latent transforming growth factor-beta and other

inhibitory factors by cultured murine iris and ciliary body cells. Curr Eye Res 1991; 10(8):761-771. 20. Jampel HD, Roche N, Stark WJ, Roberts AB. Transforming growth factor-@ in human aqueous humor. Curr Eye Res

21. Hogan MJ, Kimura SJ, Thygeson P. Signs and symptoms of uveitis. I. Anterior uveitis. Am J Ophthalmol 1959;

22. Danielpour D, Dart LL, Flanders KC, Roberts AB, Sporn MB. Immunodetection and quantitation of the two forms of transforming growth factor-@ (TGF-@I and TGF-@2) secreted by cells in culture. J Cell Physiol 1989; 138:79-86.

23. Cousins SW, McCabe MM, Danielpour D, Streilein JW. Identification of transforming growth factor-@ as an im- munosuppressive factor in aqueous humor. Invest Ophthalmol Vis Sci 1991; 32(8):2201-2211.

24. Cai F, Backman HA, Baines MG. Thimerosal: an ophthalmic preservative which acts as a hapten to elicit specific anti- bodies and cell mediated immunity. Curr Eye Res 1988; 7(4):341-351.

25. Rocha G, Baines MG, DeschZnes J. The immunology of the eye and its systemic interactions. Crit Rev Immunol 1992;

26. Benson JL, Niederkorn JY. Interleukin-1 abrogates anterior chamber-associated immune deviation. Invest Ophthalmol

27. Streilein JW, Cousins S, Bradley D. Effect of intraocular gamma-interferon on immunoregulatory properties of iris and

28. Easom HA, Zimmerman LE. Sympathetic ophthalmia and bilateral phacoanaphylaxis. Arch Ophthalmol1964; 72:9- 15. 29. Gery I, Nussenblatt R, BenEzra D. Dissociation between humoral and cellular responses to lens antigens. Invest

30. Danielpour D, Sporn M. Differential inhibition of transforming growth factor 01 and @2 activity by a2-macroglobulin.

31. Meager A: Assays for transforming growth factor 0. J Immunol Meth 1991; 141:l-14.

structure. Science 1986; 233532-535.

tent growth inhibitors of interleukin-1-induced thymocyte proliferation. Cell Immunol 1988; 114:41-54.

Ophthalmol Vis Sci 1983; 24:687-696.

1 990; 9( 10):963 -969.

47: 155- 170.

12(3,4):81-100.

Vis Sci 1990; 31:2123-2128.

ciliary body cells. Invest Ophthalmol Vis Sci 1992; 33:2304-2315.

Ophthalmol Vis Sci 1981; 20;32-39.

J Biol Chem 1990; 265(12):6973-6977.

354

Ocu

l Im

mun

ol I

nfla

mm

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

Yor

k U

nive

rsity

Lib

rari

es o

n 11

/14/

14Fo

r pe

rson

al u

se o

nly.