Factor-fl Production - NCBI

25 downloads 296 Views 2MB Size Report
either agent alone.20 This prolongation was critically de- pendent on the administration ...... Furth ME, Davis LJ, Fleurdelys B, Scolnick EM. Monoclonal Anti-.
ANNALS OF SURGERY Vol. 218, No. 4, 492-503 © 1993 J. B. Lippincott Company

Combination Anti-CD2 and Anti-CD3 Monoclonal Antibodies Induce Tolerance While Altering lnterleukin-2, lnterleukin-4, Tumor Necrosis Factor, and Transforming Growth Factor-fl Production Kenneth D. Chavin, M.D., Lihui Qin, M.D., Jixun Lin, M.D., Jennifer E. Woodward, M.S., Prabhakar Baliga, M.D., and Jonathan S. Bromberg, M.D., Ph.D.

From the Departments of Surgery and Microbiology and Immunology, Medical University of South Carolina, Charleston, South Carolina

Objective These studies were designed to elucidate the mechanism by which signals delivered by anti-CD2 monoclonal antibody (MoAb) interfere with activational signals delivered by anti-CD3 MoAb and induce long-term graft survival and tolerance.

Summary Background Data Anti-CD2 or anti-CD3 MoAb can prolong allograft survival when administered alone. In combination, they synergistically prolong survival while reducing anti-CD3-associated cytokine toxicity. It was postulated that the mechanism of synergism and reduced cytokine toxicity was related to anti-CD2-induced alterations in anti-CD3-induced T-cell activation.

Methods C57BL/6 (H-2b) mouse hearts were transplanted to CBA (H-2k) mice. The recipients received anti-CD2 and/or anti-CD3 MoAb intravenously only at the time of initial allografting. Serum from treated animals and culture supernatants from lymphocytes stimulated in vitro with anti-CD3 were examined for interleukin (IL)-2, -4, -6, and -10, tumor necrosis factor (TNF), and transforming growth factor-: (TGF,B). RNA was isolated from lymphocytes from treated animals and examined for receptor and cytokine gene expression by northern hybridization or reverse transcribed and amplified by the polymerase chain reaction (PCR).

Results Anti-CD2 and anti-CD3 MoAbs alone prolonged graft survival (22.0 ± 0.5 days and 28.0 ± 0.5 days, respectively; p < 0.02 and p < 0.01 vs. control, by Wilcoxon signed-rank test). Combined anti-CD2/anti-CD3 MoAbs synergistically prolonged survival indefinitely (> 150 days, p < 0.01) while decreasing cytokine toxicity. Second donor-specific allografts also showed long-term survival. The peak serum TNF concentration (2100 units/mL) was reduced 78% by anti-CD2 treatment (455 units/mL). Anti-CD2 inhibited anti-CD3-stimulated proliferation and in vitro production of IL-2 and IL-4, with no alteration of IL-6, IL-1 0, or TNF. Conversely, there was an

492

Tolerance Induction and T-Cell Activation

Vol. 218 * No. 4

493

increase in the immunosuppressive cytokine TGF,B. PCR analysis showed that anti-CD2 reduced anti-CD3-stimulated IL-2 messenger RNA expression, and by northern analysis, anti-CD2 inhibited anti-CD3-stimulated increases in messenger RNA for the CD2 and CD3 receptors themselves.

Conclusions The combination of anti-CD2 and anti-CD3 MoAbs induced a state of tolerance while decreasing anti-CD3-associated cytokine toxicity. The mechanism was related to anti-CD2-generated alterations in T-cell activation and gene expression.

Anti-CD3 monoclonal antibodies (MoAbs) have been used in induction therapy' and the treatment of acute rejection episodes for several years.2'3 Many side effects have been associated with the use of anti-CD3 MoAbs. These side effects have been attributed to T-cell activation induced by crosslinking of the CD3 molecules by the antibody.4 Receptor crosslinking and T-cell activation are accompanied by increased expression of the interleukin (IL)-2 receptor and the release of various cytokines. The cytokines that have been demonstrated to be involved are tumor necrosis factor (TNF), y-interferon, IL- 1, IL-2, and IL-6.5 These products may result in the development of hypoglycemia, hypothermia, diarrhea, pulmonary edema, and death.6 In an attempt to alleviate these side effects, multiple approaches have been used to alter anti-CD3-induced T-cell activation. Changes in the structure of anti-CD3 MoAbs have been investigated, including the use of F(ab')2 fragments,7'8 alternative isotypes,9"10 and humanized mouse anti-human CD3 MoAbs." Additional avenues of investigation have focused on the use of antiTNF MoAb administered in conjunction with, or as a pretreatment to, anti-CD3 MoAb administration.'2 Prevention of anti-CD3-associated cytokine syndrome has also been attempted using pretreatment with anti-CD4 MoAb,'3 pentoxifylline,'4 or corticosteroids to abrogate cytokine release.'5",6 All these measures have had some success; however, they did not prolong graft survival in combination with anti-CD3 MoAb compared with antiCD3 MoAb alone. In addition, they have generally not acted on early events in anti-CD3-induced activation. These protocols have focused mainly on preventing or diminishing more distal side effects that occur after Tcell activation. Presented at the 11 3th Annual Scientific Session of the American Surgical Association, Baltimore, Maryland, April 1-3, 1993. Supported by a Biomedical Research Support grant from MUSC; Procept, Inc.; The American Surgical Association Foundation Fellowship Award; and National Institutes of Health grant A132655 (J.S.B.). Address reprint requests to Jonathan S. Bromberg, M.D., Ph.D., Department of Surgery, Medical University of South Carolina, Charleston, SC 29425. Accepted for publication April 9, 1993.

The approach adopted here was to combine anti-CD2 MoAb with anti-CD3 MoAb to prolong allograft survival while altering the anti-CD3-associated cytokine syndrome. The rationale for this investigation was twofold. First, the CD2 and CD3 cell surface receptors are physically and functionally associated in the cell membrane. Thus, perturbing one may alter the functional capacity of the other. Second, anti-CD2 MoAbs are themselves potent immunosuppressants. Anti-CD2 MoAbs can suppress cytotoxic T-lymphocytes'7"18 and prolong xenograft and allograft survival.'9 We previously demonstrated that anti-CD2 in combination with anti-CD3 MoAbs resulted in significant prolongation of allograft survival, which was greater than either agent alone.20 This prolongation was critically dependent on the administration regimen, that is, antiCD2 MoAb had to be administered before anti-CD3 MoAb to achieve a synergistic prolongation of survival. This particular regimen also led to amelioration of the cytokine syndrome. The current study extends those results, showing that the combination of these antibodies induces a state of tolerance and that the suppressive effect of anti-CD2 MoAb is related to inhibition of antiCD3-stimulated cytokine production and messenger RNA (mRNA) expression. Anti-CD2 also inhibits antiCD3-stimulated upregulation of CD2 and CD3 receptor mRNA expression. These interacting MoAbs and receptors define a regulatory circuit that can be manipulated as a new approach to immunosuppressive strategies.

MATERIALS AND METHODS Animals CBA/J (H-2k) female mice (8 to 10 weeks of age) were purchased from Jackson Laboratory, and timed pregnant C57BL/6 (H-2b) mice were purchased from Harlan Sprague-Dawley, Inc. (Indianapolis, IN).

Reagents The 12-15 rat IgG, antimouse CD2 hybridoma,2' a gift of Dr. P. Altevogt (Heidelberg, Germany), was grown in culture and purified over protein G (Pharma-

494

Ann. Surg. * October 1993

Chavin and Others

cia, Piscataway, NJ). The 11 B 11 rat antimouse IL-422 and Y 13-259 rat IgG, antihuman ras hybridomas23 were purchased from the American Type Culture Collection (ATCC) (Rockville, MD). The 145-2C11 hamster antimouse CD324 was a gift from Dr. J. A. Bluestone (University of Chicago, Chicago, IL), and the S4B6 rat antimouse IL-2 hybridoma25 was a gift from Dr. T. R. Mossman (DNAX, Palo Alto, CA). These were grown in culture and purified over protein G. Antibodies were injected intravenously in 0.5 mL volumes in phosphatebuffered saline (PBS) or were used in vitro at the specified concentrations.

Cardiac Transplantation Donor neonatal (postnatal day 1 to 2) C57BL/6 mice were killed, and their whole hearts removed and placed in a subcutaneous pocket in the ear pinnae of CBA/J recipients as previously described.'9'20'26'27 The survival of the allografts was followed by electrocardiographic monitoring (polygraph 78 series with a preamplifier and filters, Grass Instruments, Quincy, MA) every other day. Cessation of cardiac electrical activity for two consecutive readings was the determinant of rejection. The statistical comparison of the results used the Wilcoxon signed-rank test. The recipients received intravenous injections of MoAbs at specified times with respect to

transplantation.

Proliferation Assays The spleens were removed and mechanically dissociated into single-cell suspensions, and erythrocytes were lysed with Tris NH4Cl. For anti-CD3 stimulation, the cells were placed in culture and assayed as previously described.28 We coated 96-well round-bottomed plates with anti-CD3 MoAb at various concentrations in 50 uL of PBS for 1 hour at room temperature. The wells were then washed twice with complete medium (RPMI- 1640, 10% fetal calf serum, 100 IU/mL of penicillin, 100 mg/ mL of streptomycin, 1 mmol/L pyruvate, 1 x nonessential amino acids, 2 mmol/L L-glutamine, and 5 x 10' mol/L 2-mercaptoethanol). We placed 1 x 105 splenocytes in 200-,g.L triplicate cultures, in complete medium, for 72 hours, at 37 C, in a 5% C02/95% air atmosphere. Eighteen hours before the termination of the culture, the wells were pulsed with 0.5 uCi of 3H-thymidine and then harvested on glass fiber filters. Incorporation was quantified with a beta counter. The results were expressed as the mean of triplicate determinations; the SEM was < 10%. The statistical comparison used the two-tailed Student's t test. For calcium ionophore plus phorbol ester stimulation, ionomycin at a dose of 150 ng/mL plus

phorbol 12-myristate 13-acetate (1 ,ug/mL) were added at the initiation of culture. The cultures were pulsed and harvested as before.

Cytokine Syndrome Assessment Anti-CD3-induced cytokine production was assessed in vivo and in vitro using standard bioassays and enzymelinked immunosorbent assay (ELISA) analysis. Serum from animals receiving MoAbs was collected by cardiac puncture or retro-orbital exsanguination at 0, 1.5, and 6 hours after MoAb administration. Culture supernatants from anti-CD3-stimulated cells were assayed after 72 hours. All experiments were performed successfully a minimum of three times.

IL-2 and IL-4 IL-2 production was tested with the IL-2-dependent murine cytotoxic T-cell line (CTL-L),29 and IL-4 was tested with the IL-2- and IL-4-dependent T-cell line HT2,30 as previously described. We cultured 5 X I03 CTL-L or 1 x 104 HT-2 cells in triplicate in 200-,AL 96-well round-bottomed plates. We added 20 ,uL of serum or 100 ,uL of culture supernatants to each well, and the cells were cultured for 72 hours. Eighteen hours before the termination of culture, the wells were pulsed with 0.5 IACi of 3H-thymidine, and incorporation was quantified as described previously. The results are expressed as the mean ± SEM. Standard curves with purified recombinant mouse IL-2 and IL-4 (Genzyme, Cambridge, MA) were included in each assay, and the results (in units per milliliter) are derived from these curves. The specificity of each assay was confirmed by incorporating saturating amounts of anti-IL-2 and/or anti-IL-4 MoAbs in each culture.

IL-6 IL-6 production was assayed with the B9 B-cell hybridoma.3' We cultured 3.5 x 103 cells in triplicate in 200-FL 96-well round-bottomed plates. We added 20 juL of serum or 100 ,AL of culture supernatants at the beginning of the culture, which was assayed after 4 days. At this time, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma, St. Louis, MO) was added at a dose of 0.6 mg/mL, and the cells were observed until the stain was visible. The reaction was stopped with 100 ,uL of 0.5% glacial acetic acid in isopropanol; 100 ,uL H20 was added, and absorbance was read at 570 nm with a 690-nm subtraction. Standard curves with purified rmIL-6 (Genzyme) were included in each assay.

Vol.218-No. 4

IL-1 0 IL- 10 was assayed using a two-antibody capture ELISA. We coated 96-well flat-bottomed plates with 50 AL of rat antimouse IL- 10 (catalog no. 18141 D, Pharmingen, San Diego, CA) at a dose of 2 ,ug/mL in 1% bovine serum albumin in PBS overnight at 4 C. The plates were washed twice with 0.05% Tween-20 in PBS and blocked with 200 IAL of 1% bovine serum albumin in PBS for 2 hours at room temperature. The plates then were washed, serum and culture supernatants were added, and they were incubated overnight at 4 C. The plates were again washed and incubated with biotinylated antimouse IL-10 MoAb (catalog no. 18152D, Pharmingen) at a dose of 2 ,ug/mL for 45 minutes at room temperature. They were again washed, 100 AL of a 1:1000 dilution of peroxidase-streptavidin (Kirkegaard & Perry, Gaithersburg, MD) was added to each well, and they were incubated at room temperature for 30 minutes. The plates were washed, and 100 ,L of freshly prepared 2,2'-azino-di(3-ethylbenzthiazoline-6-sulfonate) peroxidase solution (Kirkegaard & Perry) was added to each well. The reaction proceeded for 10 to 90 minutes and was stopped by adding 100 ,uL of 1% sodium dodecyl sulfate in PBS. Plate optical densities were measured at 405 nm. Murine IL-10 COS cell supernatants with an activity of 3000 units/mL, a gift from Dr. William Fanslow (Immunex, Seattle, WA), were used for the standard curve determinations.

TNF TNF levels were assayed with the L929 mouse fibroblast cell line.32 We plated 5 x I05 cells in 100-,uL complete medium in 96-well flat-bottomed plates. When the cells were 90% to 100% confluent, the medium was removed, and 80 ,tL of complete medium with dactinomycin (Sigma) at a dose of 1 mg/mL was added plus 20 ,uL of an experimental sample. The cultures were incubated for 18 hours and then assayed with MTT as described previously. Recombinant mouse TNFa, a gift from Dr. Grace Wong (Genentech, San Francisco, CA) was used for standard curve determination.

Transforming Growth Factor-,8 (TGF,8) TGFf was assayed by the mink lung cell assay.33 We plated 1 x 104 cells in 96-well flat-bottomed plates in complete Dulbecco's modified Eagle's medium with 10% fetal calf serum. When the cells had become 90% to 100% confluent, 20 ,uL of serum or 100 ,uL of culture supernatant samples, which had been heat activated for 3 minutes at 95 C, were added. The plates were incubated for 18 hours and pulsed with 2 ,uCi of 3H-thymi-

Tolerance Induction and T-Cell Activation

495

dine for 2 hours. The culture supernatants were then discarded, and 50 uL of trypsin-versene was added. The cells were incubated for 30 minutes at 37 C and harvested on glass fiber filters, and 3H-thymidine incorporation was quantified on a beta counter. Standard curves were generated with recombinant mouse TGF,B (Gibco BRL, Gaithersburg, MD).

Northern Hybridization of mRNA and Cytokine Gene Polymerase Chain Reaction (PCR) Splenic lymphocytes from in vivo MoAb treatment groups were obtained, and whole-cell RNA was isolated using RNAzol B (Cinna/Biotecx, Houston, TX). Equal quantities of whole-cell RNA underwent electrophoresis on 1% formaldehyde agarose gels and were transferred to nylon membranes. The membranes were then sequentially probed for CD2, CD3, and f-actin using 32p random primer (Stratagene, La Jolla, CA)-labeled probes. The CD2 probe was the 0.7-kilobase EcoRI-HpaI fragment from the murine CD2 complementary DNA clone pMCD2-2.34 The CD3 probe was the 1.4-kilobase EcoRI fragment of CD3E.35 The f-actin probe was a 540-base pair PCR product between base pairs 25 to 565 generated from C57BL/6 genomic DNA and ligated to the SmaI site of pBSSK+. The isolated RNA was reversed transcribed using random primers. The complementary DNA was amplified using PCR. Commercially purchased primers (Clontech, Palo Alto, CA) for murine IL- 1 a, IL-2, IL-2 receptora, IL-4, IL-6, TNFa, and f-actin were used with PCR conditions according to the manufacturer's recommendations. The PCR products were run on 2% agarose gels and stained with ethidium bromide.

RESULTS Allograft Survival with Combined Anti-CD2 and Anti-CD3 MoAb Treatment It was previously demonstrated that the administration of high doses of anti-CD2 and anti-CD3 MoAbs alone and in combination significantly prolonged allograft survival.20 When the antibodies were combined, the regimen of MoAb administration that produced the most pronounced allograft survival was sequential administration of 100 ,g of anti-CD2 on days 0 and 1 after transplantation followed by 1 mg of anti-CD3 on days 2, 3, 4, 5, and 10. The initial observation was that this regimen of MoAbs prolonged allograft survival (> 60 days). These initial grafts and additional allografts have now been observed to survive longer than 150 days (Fig. 1 A). This is significantly greater (p < 0.001) than the survival

496

Ann. Surg. * October 1993

Chavin and Others 1st Graft Survival

A

-

-

-j-

-

| C

Control cCD2 100rg dO,1 a aCD3 1mg dO,1,2,5,10,15 &CD2 100pO dO1l/ aCD3 1mg 2,3,4,5,10 Control mAb I00vg dO,1/ aCD3 1mg d2,3,4,5,10

2 60* C,) to

by40'

801 --

*

1i >

Nontoerant

~~~~~~~~~~Tolernt

60-

40-

MoAb coated onto the plates. This is an accessory cellindependent proliferative assay of mature T-cells bearing intact T-cell receptors.36 Cells from the animals that had received anti-CD2 MoAb in vivo up to 1 week previously had a decreased proliferative response to anti-CD3 in vitro (Table 1). Cells pretreated with anti-CD3 MoAb in vivo had a variable proliferative response; both inhibition and augmentation could be observed. However, cells pretreated with both MoAbs had the most marked degree of inhibition of proliferation. Therefore, pretreatment with anti-CD2 MoAb inhibited the ability of antiCD3 MoAb to activate T-cells directly and converted a variable and sometimes stimulating anti-CD3 signal into an inhibitory signal. Identical results were seen in four additional experiments. Anti-CD2 MoAb in vivo does not, however, impair activation by a calcium ionophore plus phorbol ester (107,463 vs. 102,832 cpm), stimulating signals that activate T-cells by more distal secondmessenger pathways.37 It was previously shown'7'20 that both single and combined MoAb treatment does not deplete T-cells; therefore, the diminished proliferative responses are not caused by a lack of T-cells.

20-

0

10

20

30

40

50

60

70

80

90

Days Figure 1. Cardiac allograft survival with combined anti-CD2 and antiCD3 MoAb administration. Transplanted mice received MoAbs intravenously at the doses and times indicated. (A) Survival of primary grafts. (B) Survival of secondary, donor-specific grafts in unprimed (nontolerant) or tolerant recipients.

of isotype-matched control MoAb plus anti-CD3 MoAb (26.5 ± 0.5 days), which demonstrated no increase in graft survival over anti-CD3 alone (27.4 ± 3.3 days, Fig. 1A). Second allografts of original donor type transplanted to these primary recipients showed prolonged survival to more than 60 days without the need for any additional immunosuppression (Fig. 1 B). The prolonged first and second graft survival rates demonstrate tolerance induction as a result of combined MoAb administration.

Anti-CD3-Induced Proliferation With the observation that a specific combination of anti-CD2 plus anti-CD3 MoAbs resulted in tolerance, it was then asked whether this effect occurred at the level of stimulation of individual T-cell receptors by assaying Tcell activation. Splenic lymphocytes were isolated from CBA/J mice that had received MoAb or combinations of MoAbs in vivo before sacrifice. The regimen of MoAbs was similar to that used in the allograft experiments. The cells were then stimulated in culture with anti-CD3

Cytokine Syndrome During the course of the studies, it was noted that combined MoAb treatment resulted in fewer side effects in vivo compared with anti-CD3 MoAb administered alone. It was previously shown that anti-CD2 MoAb decreased the anti-CD3 MoAb-associated side effect of piloerection by delaying its onset and decreasing its duration.20 To explore the effects of pretreatment with antiCD2 MoAb administration further, CBA/J mice were treated in vivo with no MoAb or 100 ,ug of anti-CD2 MoAb for two consecutive days. The animals then received I mg of anti-CD3 MoAb on the following day, and serum samples were collected just before and at 1.5 and 6 hours after anti-CD3 administration, the times to peak serum cytokine levels.5-7 The sera were assayed for specific cytokines by bioassay and ELISA. Detectable increases in cytokine production were noted for IL-2, IL-4, TNF, and TGFj, but not for IL-6 or IL- 10, as a result of anti-CD3 MoAb administration. Anti-CD2 pretreatment resulted in a 78% decrease in peak TNF production at 1.5 hours (Table 2) from 2105 U/mL of TNF in antiCD3-treated to 455 U/mL in anti-CD2 plus anti-CD3treated recipients. There were no differences detectable for serum IL-2, IL-4, or TGF3 concentrations as a result of anti-CD2 pretreatment. These cytokine experiments were next performed with in vitro stimulation. The animals were treated in vivo with anti-CD2 and/or anti-CD3 MoAbs, and isolated splenic lymphocytes were stimulated in culture with anti-CD3 MoAb. Culture supernatants were harvested

Tolerance Induction and T-Cell Activation

Vol. 218 d No. 4

497

Table 1. ANTI-CD2 mAb INHIBITS ANTI-CD3 MAb DRIVEN PROLIFERATION Response (cpm) (% inhibition)

Stimulator

Pretreatment 1. 2.

3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13.

-

Control mAb Anti-CD2; d7, 6 Anti-CD2; d2, 1 Anti-CD3; d2 Anti-CD2, d4,3 Anti-CD3; d2 -

Control mAb Anti-CD2; d7, 6 Anti-CD2; d2, 1 Anti-CD3; d2 Anti-CD2; d4, 3 Anti-CD3; d2

Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3,

10 Ag/mL 10 1g/mL 10 ig/mL 10 gg/mL 10 ug/mL

Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3, Anti-CD3,

1 1 1 1 1 1

10,ug/mL

,Ag/mL zg/mL ,ug/mL

qg/mL ,g/mL ,g/mL

Animals were pretreated in vivo with control mAb (Y13-259, 100 ug), anti-CD2 mAb (100 ug), lymphocytes were stimulated in culture as described in the methods. * For experiment 2, anti-CD2 given d4, 3 instead of d2, 1.

after 3 days and assayed for cytokine content. The assays demonstrated a 68% decrease in IL-2 content as a result of anti-CD2 treatment (Fig. 2A). A decrease of 97% in IL-2 content was observed in the anti-CD3 treatment group. A comparable decrement was also observed in the anti-CD2 plus anti-CD3 treated group. Anti-CD2 reduced IL-4 production by 30%, and anti-CD3 reduced it by 70%. The combination reduced IL-4 by 96% (Fig. 2B). Conversely, TGF3 was augmented in cells pretreated with anti-CD2 by more than 30% or anti-CD2 plus anti-CD3 by more than 50% (Fig. 2C). Anti-CD3 alone did not alter TGF3. This is significant because TGF,B is an immunosuppressive cytokine.38 No detectable levels of IL-6, IL-10, or TNF were noted in these

Table 2. SERUM CYTOKINE LEVELS 6 hrs

1.5 hrs

aCD2/ aCD3

aCD2/ aCD3

Cytokine

No mAb

aCD3

IL-2 IL-4 IL-6 IL-10 TNF TGF,B

0 0

15 310

15 290

15 50

15 50

0

0

0

0

0

11.8 455 4.3

13.0 5.3 4.5

10.5

10.5 32.8 0

10.2 2105 4.3

aCD3

Expt. 2

Expt. 1

2.3 5.0

Animals were treated with designated mAbs and sera harvested 1.5 or 6 hours after anti-CD3 mAb administration. Sera were analyzed for cytokine content. Results given as units/mL except for TGF,B, which is in pM/mL.

or

971 116,984 ND 77,821 (34) 96,870 (17) 113,176 (3) 26,473 (78)

1251 78,030 76,851 (1) 64,474 (20)

31,203 ND 12,568 (62) 18,737 (41) 7045 (80) 5269 (86)

27,621 22,656 (19) 15,614 (46) 13,440 (54)* 18,303 (35) 10,089 (66)

anti-CD3 mAb (1 mg)

64,769 (19)* 95,024 (-22) 59,425 (24)

on the

indicated days before sacrifice. Splenic

cultures (not shown). Comparable results for cytokine production were obtained in four additional experiments.

Cytokine mRNA Expression result of the alterations in anti-CD3-stimulated and culture cytokine content, we explored whether these alterations where the result of disruptions at the transcriptional level. Splenic lymphocyte RNA from animals treated with MoAbs in vivo was isolated, reverse transcribed, and amplified with specific cytokine amplimers by PCR. The results show there were no changes in IL-l1a, IL-6, or TNFa gene expression (Fig. 3). f-actin served as an internal control and was also unchanged; therefore, these results are quantitatively reliable. Anti-CD3 MoAb induced increased expression of IL-2, which was inhibited by anti-CD2 pretreatment. IL2 receptora and IL-4 gene expression were also increased by anti-CD3 administration; however, anti-CD2 had no effect on these. Comparable results were observed in additional experiments. As a serum

Receptor mRNA Expression Combined in vivo administration of anti-CD2 and anti-CD3 MoAbs altered cell surface expression of the CD2 and CD3 receptors.20 Anti-CD2 administered in vivo caused downmodulation ofCD2 expression without alterations of CD3, CD4, and CD8 expression. AntiCD3 MoAb caused downmodulation of CD3, CD4, and

498

A

Chavin and Others

Ann. Surg. * October 1993

sion of both CD2 and CD3 mRNA. Importantly, pretreatment with anti-CD2 inhibited the anti-CD3-induced increase in CD2 and CD3 mRNA expression. The effect was specific because there was no effect on fl-actin expression. Identical results were obtained in two additional experiments.

IL-2

4

2

p40.04

I1-

pCO.02_ 0.

B

60-

-

40-

P0C03

C

-

cCD2

cxCO3

ceCD2

*

cLCCD3

CO) t

IL-4

0-

=

30

20

OCE) CgDo la 0

N.DMaOOa D1

10

*m P

0S

40.00C

exC>2 +

O

aCD3

C soo

TGFI 500-

IL-1a

MoAM

400-

-9 I 300'

200-

IL-2

100. a.L_ -

Control mAb

czCD2

oCD3

acCD2

caCD3

Treatment Group Figure 2. Production of cytokines in vitro. The animals were pretreated with anti-CD2 on days 4 and 3 and/or anti-CD3 on day 2 before sacrifice. The cells were restimulated in culture with anti-CD3 MoAb at a dose of 1 iug/mL. The culture supernatants harvested after 3 days and assayed for (A) IL-2 by CTL-L assay; (B) IL-4 by HT-2 assay; and (C) TGF3 by mink lung cell assay. p versus control. Anti-IL-2 was incorporated into the HT-2 assay to restrict the specificity to IL-4.

IL-2Ra

IL-4 CD8 while causing a marked increase in CD2 expression. The combination of anti-CD2 and anti-CD3 MoAbs resulted in persistent downmodulation of CD3, CD4, and CD8 while preventing anti-CD2-induced downmodulation ofCD2 expression.'7'20 Because MoAb interaction with its cognate ligand is dependent on receptor expression and these interactions are important for the immune suppression demonstrated here, it was important to determine whether anti-CD2 and/or antiCD3 regulated receptor transcriptional expression. Whole-cell RNA was isolated from splenic lymphocytes from in vivo-treated animals. Equivalent amounts of RNA were run on a formaldehyde gel, transferred to a nylon membrane, and probed sequentially for CD2, CD3, and ,B-actin. The results demonstrated that antiCD2 MoAb in vivo had no effect on CD2 and CD3 mRNA levels compared with the isotype-matched control MoAb (Fig. 4). Anti-CD3 caused increased expres-

IL-6

TNFa

,B-actin Figure 3. Cytokine PCR products. RNA from splenic lymphocytes from mice treated in vivo with the designated MoAbs was reverse transcribed and amplified with the designated cytokine amplimers.

Tolerance Induction and T-Cell Activation

Vol. 218 No. 4 -

E 0 a o

M

0 U C

CD2

CD3

f1-actin

Figure 4. Northern hybridization. RNA from splenic lymphocytes from mice treated in vivo with the designated MoAbs was probed for CD2, CD3, and fB-actin.

DISCUSSION Our previous work described significant prolongation of allograft survival with the combined administration of anti-CD2 and anti-CD3 MoAbs in a sequential fashion.20 In addition to prolonged graft survival, anti-CD2

499

MoAb pretreatment altered the anti-CD3 MoAb-induced systemic side effect of piloerection. Anti-CD2 MoAb also partially abrogated the anti-CD3 MoAb-induced increase in CD2 receptor expression. Because the timing of MoAb administration was critical in determining the type of immunosuppression observed, these results suggested that the mechanism of action ofthis combination of MoAbs was the result of alterations in T-cell activation. It was postulated that anti-CD2 inhibited activational signals delivered by anti-CD3 or by antigen to the T-cell receptor. The interaction of anti-CD2 and anti-CD3 with CD2 and CD3 receptors, respectively, was also postulated to define a regulatory feedback circuit in which anti-CD2-driven suppression inhibited CD3-related activation and anti-CD3 enhanced CD2-related suppression. The current results, with the combination ofanti-CD2 plus anti-CD3 MoAbs, show, not only a synergistic effect in prolonging allograft survival, but also a state of tolerance development (Fig. 1). The possible mechanisms of tolerance include clonal deletion or active suppression. Alternatively, the observed decrease in the systemic antiCD3-associated cytokine syndrome as a result of antiCD2 pretreatment suggests that this regimen alters T-cell activation and gene expression. A number of assays of T-cell activation were therefore investigated. Anti-CD3-stimulated proliferative assays demonstrated that anti-CD2 MoAb treatment inhibited this proliferation (Table 1). Furthermore, anti-CD2 could not inhibit proliferation driven by calcium ionophore plus phorbol ester. This suggests that anti-CD2 perturbed a membrane proximal CD3-associated signaling event. Because anti-CD2 plus anti-CD3 was even more inhibiting, this suggests further disruption of membrane-associated signaling complexes. Anti-CD2 inhibited antiCD3-stimulated IL-2 and IL-4 production and augmented TGF,B production in vitro (Fig. 2). Again, the combination of MoAbs was even more effective in this regard. Anti-CD2 inhibited anti-CD3-driven transcription from CD2, CD3, and IL-2 genes in vivo (Figs. 3, 4). These results were consistent with changes in gene expression as a result of anti-CD2 disruption of activational signals. Others have also shown that anti-CD2 can inhibit CD3-driven proliferation and IL-2 production by human T-cell clones in vitro.39 The serum cytokine response to anti-CD3 MoAb is also consistent with the hypothesis that anti-CD2 MoAb treatment delivers a negative regulatory signal to gene expression (Table 2). The most significant inhibition is the 78% decrease in TNF. This may be a function of anti-CD2 MoAb directly inhibiting T-cell activation. Anti-CD2 may also be acting on macrophage-monocytes and other antigen-presenting cells and directly inhibiting their function because murine T-cells, B-cells,

500

Ann. Surg. * October 1993

Chavin and Others

natural killer cells, and macrophages express CD2.40 An alternative explanation is that, because the particular anti-CD2 MoAb used here can crossreact with Fc receptors, it may block anti-CD3 crosslinking. This alternative does not appear to be valid because other anti-CD2 MoAbs that recognize different epitopes demonstrate similar inhibition of the anti-CD3-associated cytokine syndrome (unpublished results). Cytokine gene expression can be differentially regulated, depending on the antigenic stimulus.4' During allograft rejection, different cytokine mRNAs are expressed during different types of rejection and at different time points in the rejection process.42 OKT3 (Ortho Pharmaceuticals, Raritan, NJ) has been shown to induce cytokine mRNA expression as a consequence of its administration in vivo, and cytokines are induced in a sequential manner.43 The differences between the current in vivo and in vitro findings with respect to cytokine expression are probably a result of the kinetic and stimulating aspects of the assays. Thus, serum cytokines assayed a few hours after anti-CD3 MoAb administration revealed that anti-CD2 induced inhibition of TNF but IL2, IL-4, IL-6, and TGF3 were unaffected (Table 2). The PCR-amplified mRNA from lymphocytes treated 2 days earlier in vivo showed inhibition of IL-2 expression but not of IL- 1 a, IL-2 receptora, IL-4, IL-6, or TNFa (Fig. 3). Culture supernatant cytokines obtained after an additional 3 days of in vitro stimulation showed inhibition of IL-2 and IL-4 and an increase in TGFf (Fig. 2). These differences are likely related to the time point at which the assay was performed and the dissociation between the protein and mRNA content of a cell.44 Nonetheless, anti-CD2 inhibited the expression of important cytokines and receptors at a transcriptional and/or translational level. These results also show that anti-CD2driven suppression may be the result only of downregulation of activational cytokines but also may be caused by upregulation of suppressive cytokines such as TGFf. Therefore, anti-CD2 MoAb does not merely inhibit the expression of all genes but alters the pattern of expression to one that promotes immune suppression. The model previously advanced20 was that anti-CD2 blocked CD2 adhesion function and delivered negative regulatory signals, which impeded anti-CD3-driven activation. Anti-CD3 blocked CD3 antigen-recognition function and activated certain second-messenger pathways. The combination of MoAbs delivered in a specific temporal sequence resulted in a regulatory circuit favoring immunosuppression by anti-CD2 over activation by anti-CD3 MoAb. The results with cytokine production in vivo and in vitro, cytokine mRNA expression in vivo, and northern analysis of CD2 and CD3 receptor mRNA expression in vivo are all concordant with this model. The previous fluorescent flow cytometric results of CD2

and CD3 protein expression'7,20 and the current results of receptor mRNA expression also support the model. The regulatory circuit results preferentially in downmodulation of CD3 expression and function and upmodulation of CD2 expression and function. This circuit should promote suppression and tolerance. Others have also described combined MoAb-receptor approaches to tolerance induction, although they have not described specific effects of receptor engagement on T-cell gene expression as shown here.45'46 Two major areas need further exploration at this time. The mechanism of maintenance oftolerance is not necessarily the result solely of continued alterations of T-cell activation pathways. T-cell subset distribution and responses need to be evaluated for clonal deletion, clonal abortion, anergy, or suppression. The precise nature of anti-CD2-induced CD3 signaling defects requires analysis of events such as CD3 subunit associations,47 protein kinase activation, 48 calcium currents,49 and inositol phosphate metabolism.50

References 1. Vigeral P, Chkoff N, Chatenoud L, et al. Prophylactic use of OKT3 monoclonal antibody in cadaver kidney recipients. Transplantation 1986; 41:730-733. 2. Cosimi AB, Burton RC, Colvin RB, et al. Treatment of acute renal allograft rejection with OKT3 monoclonal antibody. Transplantation 1981; 32:535-539. 3. Ortho Multicenter Transplant Study Group. A randomized clinical trial of OKT3 monoclonal antibody for acute rejection of cadaveric renal transplants. N Engl J Med 1985; 313:337-342. 4. Hirsch R, Gress RE, Plyznik DH, et al. Effects of in vivo administration of anti-CD3 monoclonal antibody on T cell function in mice. J Immunol 1989; 142:737-743. 5. Alegre M, Vandenabeele P, Flamand V, et al. Hypothermia and hypoglycemia induced by anti-CD3 monoclonal antibody in mice: role of tumor necrosis factor. Eur J Immunol 1990; 20:707-7 10. 6. Ferran C, Sheehan K, Dy M, et al. Cytokine-related syndrome following injection ofanti-CD3 monoclonal antibody: further evidence for transient in vivo T cell activation. Eur J Immunol 1990;

20:509-515. 7. Hirsch R, Bluestone JA, DeNenno L, Gress RE. Anti-CD3 F(ab')2 fragments are immunosuppressive in vivo without evoking either the strong humoral response or morbidity associated with whole MoAb. Transplantation 1990; 49:1117-1123. 8. Woodle ES, Thistlethwaite JR, Ghobrial IA, et al. OKT3 F(ab')2 fragments-retention of the immunosuppressive properties of whole antibody with marked reduction in T cell activation and lymphokine release. Transplantation 1991; 52:354-360. 9. Woodle ES, Thistlethwaite JR, Jolliffe LK, et al. Anti-CD3 monoclonal antibody therapy. Transplantation 1991; 52:361-366. 10. Frenken LAM, Koene RAP, Tax WJM. The role of antibody isotype in IFN-,y and IL-2 production during anti-CD3-induced T cell proliferation. Transplantation 1991; 51:881-887. 11. Woodle SE, Thistlethwaite JR, Jolliffe LK, et al. Humanized OKT 3 antibodies: successful transfer of immune properties and idiotype expression. J Immunol 1992; 148:2756-2763. 12. Ferran C, Sheehan K, Schreiber R, et al. Anti-TNF abrogates the

Tolerance Induction and T-Cell Activation

Vol. 218 * No. 4

13. 14.

15. 16.

17.

18.

19. 20. 21.

22.

23.

24.

25.

26. 27.

28. 29.

30. 31. 32. 33.

34.

cytokine-related anti-CD3 induced syndrome. Transplant Proc 1991; 3:849-850. Hirsch R, Chatenoud L, Gress RE, et al. Suppression of the humoral response to anti-CD3 monoclonal antibody. Transplantation 1989; 47:853-857. Alegre ML, Gastaldella K, Abramowicz D, et al. Evidence that pentoxifylline reduces anti-CD3 monoclonal antibody-induced cytokine release syndrome. Transplantation 1991; 52:674-679. Ferran C, Dy M, Merite S, et al. Reduction of morbidity and cytokine release in anti-CD3 MoAb-treated mice by corticosteroids. Transplantation 1990; 50:642-648. Alegre ML, Vandenabeele P, Depierreux M, et al. Cytokine release syndrome induced by the 145-2Cl 1 anti-CD3 monoclonal antibody in mice: prevention by high doses of methylprednisolone. J Immunol 1991: 146:1184-1191. Bromberg JS, Chavin KD, Altevogt P, et al. Anti-CD2 monoclonal antibodies alter cell-mediated immunity in vivo. Transplantation 1991; 51:219-25. Guckel B, Berek C, Lutz M, et al. Anti-CD2 antibodies induce T cell unresponsiveness in vivo. J Exp Med 1991; 174:957-967. Chavin KD, Lau, HT, Bromberg JS. Prolongation of allograft and xenograft survival in mice by anti-CD2 monoclonal antibodies. Transplantation 1992; 54:286-291. Chavin KD, Qin L, Lin J, et al. Anti-CD2 and anti-CD3 monoclonal antibodies synergize to prolong survival with decreased side effects. Transplantation 1993; 55:901-908. Altevogt P, Kohl U, Von Hoegen P, et al. Antibody 12-15 crossreacts with Fc receptors and CD2: study of thymus expression, genetic polymorphism and biosynthesis of the CD2 protein. Eur J Immunol 1989; 19:341-346. Ohara J, Paul WE. Production of a monoclonal antibody to and molecular characterization of B-cell stimulatory factor-1. Nature 1985; 315:333-336. Furth ME, Davis LJ, Fleurdelys B, Scolnick EM. Monoclonal Antibodies to p21 products of the transforming gene of the Harvey murine sarcoma virus and of the cellular ras gene family. J Virol 1982; 43:294-304. Leo 0, Foo M, Sachs DH, et al. Identification of a monoclonal antibody specific for a murine T3 polypeptide. Proc Natl Acad Sci U S A 1987; 84:1374-1378. Mossman TR, Cherwinski H, Bond MW, et al. Two types of murine helper T cell clone I: definition according to profiles of lymphokine activities and secreted proteins. J Immunol 1986; 136:2348-2357. Fulmer RI, Cramer AT, Liebelt AG. Transplantation of cardiac tissue into the mouse ear. Am J Anat 1963; 113:273-285. Judd KP, Trentin JJ. Cardiac transplantation in mice. Transplantation 1971; 11:298-302. Guy R, Foo-Philips M, Sharrow SO, Hodes RJ. Subpopulations of fetal thymocytes defined by expression of T cell receptor/CD3 and IL-2 receptor. J Immunol 1991; 146:418-424. Baker PE, Gillis S, Smith KA. Monoclonal cytotoxic T-cell lines. J Exp Med 1979; 149:273-278. Watson J. Continuous proliferation of murine antigen-specific helper T lymphocytes in culture. J Exp Med 1979; 150:1510-1519. Lansdorp PM, Aarden LA, Calafat J, Zeiljemaker WP. A growth factor depended B-cell hybridoma. Curr Top Microbiol Immunol 1986; 132:105-113. Ruff MR, Gifford GE. Rabbit tumor necrosis factor: mechanism of action. Infect Immun 1981; 31:380-385. Danielpour D, Dart LL, Flanders KC, et al. Immunodetection and quantitation of the two forms of transforming growth factor-beta (TGF-,B1 and TGF-32) secreted by cells in culture. J Cell Physiol 1989; 138:79-86. Yagita H, Okimura K, Nakauchi H. Molecular cloning of the mu-

35.

36. 37. 38.

39.

40. 41. 42.

43.

44.

45. 46. 47.

48. 49. 50.

501

rine homologue of CD2: homology of the molecule to its human counterpart TI 1. J Immunol 1988; 140:1321-1326. Gold DP, Clevers H, Alarcon B, et al. Evolutionary relationship between the T3 chains of the T cell receptor complex and the immunoglobulin supergene family. Proc Natl Acad Sci U S A 1987; 84:7649-7653. Quintans J, Yokoyama A, Evavold B, et al. Direct activation of murine resting T cells by ConA or anti-CD3 Ig. J Mol Cell Immunol 1989; 4:225-237. Coligan JE, Kruisbeek AM, Margulies DH, et al. Current Protocols in Immunology. New York: Greene Publishing Associates and Wiley-Interscience, 1991. Wallick SC, Figari IS, Morris RE, et al. Immunoregulatory role of transforming growth factor (TGF-f3) in development of killer cells: comparison of active and latent TGF-,B1. J Exp Med 1990; 172:1777-1784. Yssel H, Aubry JP, DeWaal Malefijt R, et al. Regulation by antiCD2 monoclonal antibody of the activation of a human T cell clone induced by anti-CD3 or anti-T cell receptor antibodies. J Immunol 1987; 139:2850-2855. Yagita H, Nakamura T, Karasuyama H, Okumura K. Monoclonal antibodies specific for murine CD2 reveal its presence on B as well as T cells. Proc Natl Acad Sci U S A 1989; 86:645-649. Gauchat JF, Gauchat D, De Weck AL, Stadler BM. Cytokine mRNA levels in antigen-stimulated peripheral blood mononuclear cells. Eur J Immunol 1989; 19:1079-1085. Krams SM, Falco DA, Villanueva JC, et al. Cytokine and T cell receptor gene expression at the site of allograft rejection. Transplantation 1992; 53:15 1-156. Pisa EK, Pisa P, Hansson M, Wigzell H. OKT 3-induced cytokine mRNA expression in human peripheral blood mononuclear cells measured by polymerase chain reaction. Scand J Immunol 1992; 36:745-749. Schall TJ, O'Hehir RE, Goeddel DV, Lamb JR. Uncoupling of cytokine mRNA expression and protein secretion during the induction phase of T cell anergy. J Immunol 1992; 148:381-387. Isobe M, Yagita H, Okumura K, Ihara A. Specific acceptance of cardiac allograft after treatment with antibodies to ICAM- 1 and LFA-1. Science 1992; 255:1125-1127. Qin S, Cobbold SP, Pope H, et al. "Infectious" transplantation tolerance. Science 1993; 259:974-977. Mizoguchi H, O'Shea JJ, Longo DL, et al. Alterations in signal transduction molecules in T lymphocytes from tumor bearing mice. Science 1992; 258:1795-1798. Friedrich B, Cantrell DA, Gullberg M. Evidence for protein kinase C: activation in T lymphocytes by stimulation ofeither the CD2 or CD3 antigens. Eur J Immunol 1989; 19:17-23. Gerli R, Bertotto A, Crupi S, et al. Activation of cord T lymphocytes: I. Evidence for a defective T cell mitogenesis induced through the CD2 molecule. J Immunol 1989; 142:2583-2589. Ward SG, Ley SC, MacPhee C, Cantrell DA. Regulation of D-3 phosphoinositides during T cell activation in via the T cell antigen receptor/CD3 complex and CD2 antigens. Eur J Immunol 1992;

22:45-49.

Discussion DR. CLYDE F. BARKER (Philadelphia, Pennsylvania): I'm grateful to Dr. Bromberg for letting me see the manuscript, mainly because it gives me a chance to publicly congratulate him on the excellent work that he's done as an American Surgi-

cal Association Fellow for the last 2 years and to encourage the membership to support this fellowship through the donations