Interleukin-2-Induced Lung Injury - Circulation Researchcircres.ahajournals.org/content/circresaha/74/2/329.full.pdfInterleukin-2-Induced LungInjury TheRole ofComplement ... IL-2.56

329

Interleukin-2-Induced Lung Injury

The Role of Complement

R. Rabinovici, M.D. Sofronski, P. Borboroglu, A.M. Spirig, L.M. Hillegas, J. Levine, J. Vernick,S.M. Scesney, N. Feuerstein, G. Feuerstein

Abstract Pulmonary edema and sepsislike syndrome aregrave complications of interleukin-2 (IL-2) therapy. Recentanimal studies have suggested IL-2-induced microvascular injuryas the underlying mechanism. Since complement factors havebeen shown to mediate increased vascular permeability in diverseconditions that lead to pulmonary injury and recombinant hu-man IL-2 is known to activate the complement system in patientsundergoing IL-2 therapy, we hypothesized that complementfactors play a pivotal role in the development of increasedvascular permeability after IL-2 treatment. To test this hypoth-esis, we evaluated the capacity of recombinant soluble humancomplement receptor type 1 (sCR1, BRL 55730), a new highlyspecific complement inhibitor, to attenuate IL-2-induced lunginjury in the rat. Recombinant human IL-2 (intravenously for 60

I nterleukin-2 (IL-2) is under investigation as an immu-notherapeutic agent used in the treatment of ad-vanced metastatic cancer.12 The clinical toxicity of

the IL-2 regimen includes sepsislike syndrome and pul-monary edema secondary to increased vascular perme-ability.3'4 Although the pathophysiological processes lead-ing to these side effects are still obscure, severalmechanisms have been proposed. For example, recent invitro data have suggested that tissue injury is mediated bylymphokine-activated killer cells activated in response toIL-2.56 However, several in vivo studies conducted insheep7-9 and rats10 cast doubt on this possibility, since lungmicrovascular injury has been demonstrated as early as 2to 6 hours after IL-2 infusion, a time interval insufficientfor lymphokine-activated killer cell generation." Neutro-phils, which are early-response inflammatory cells, havebeen recently implicated in the pulmonary response toacute IL-2 administration.10'12 In addition, increased vas-cular permeability could be consequent to the IL-2-induced production of several humoral inflammatory me-diators such as interleukin-1,"3 tumor necrosis factor-a(TNF-a),'4"15 thromboxane A2 (TXA2),8"0"6 and platelet-activating factor (PAF).17 This possibility draws credencefrom the observation that IL-2 itself did not inducepermeability defects in vitro7 and that athymic nude miceand leukocyte-depleted mice develop a milder form oflung microvascular injury after IL-2 infusion.'8

Received March 10, 1993; accepted November 4, 1993.From the Department of Surgery (R.R., M.D.S., P.B., A.M.S.,

J.V.), Jefferson Medical College, Philadelphia, Pa; SmithKlineBeecham (L.M.H., G.F.), King of Prussia, Pa; T Cell Sciences(J.L.), Cambridge, Mass; and the Department of Pathology (N.F.),University of Pennsylvania, Philadelphia.Correspondence to Reuven Rabinovici, MD, Department of

Surgery, Jefferson Medical College, 1025 Walnut St, Philadelphia,PA 19107-5083.

minutes) at 106 U per rat (n=4) elevated lung water content(37±6%, P<.05), myeloperoxidase activity (162±49%, P<.05),and serum thromboxane B2 (30+±1 pg/100 ,uL,P<.01) and had noeffect on serum tumor necrosis factor-a. sCR-1 at 30 mg/kg(n=5), but not at 10 mg/kg (n=6), attenuated the elevation oflung water content (18±2%, P<.05) and myeloperoxidase activ-ity (42±9%, P<.05) but failed to alter serum thromboxane B2response to IL-2. These data suggest the involvement of comple-ment in the pathogenesis of IL-2-induced pulmonary microvas-cular injury and point to the potential therapeutic capacity ofcomplement inhibitors in combating this toxic effect of IL-2therapy. (Circ Res. 1994;74:329-335.)Key Words * cytokines * anaphylatoxins * pulmonary

edema

Recently, IL-2 infusion to cancer patients has beenshown to induce a dose-dependent activation of thecomplement system that correlated with the develop-ment of microvascular injury (vascular leakage syn-drome).'9 Also, complement factors have been shown toplay a key role in other pathological conditions thatinvolve endothelial damage and microvascular injury,such as sepsis20'21 and adult respiratory distress syn-drome.2223 Therefore, we hypothesized that comple-ment might be involved in the mediation of IL-2-induced lung injury.

Recently, the human complement receptor-1 gene hasbeen cloned,24 and the solubilized truncated extracellularform has been expressed and purified.24 This 200-kDprotein has been shown to be a potent inhibitor of humanC3 and C5 convertases in vitro24 and to potentially inhibitC3a and C4a formation. Soluble complement receptortype 1 (sCR1) has already demonstrated efficacy in theinhibition of complement-mediated tissue injury in vivo,including pulmonary microvascular injury associated withIgG or cobra venom factor-induced complement activa-tion,25 burn injury,25 or endotoxin-induced adult respira-tory distress syndrome in PAF-primed rats.26 The presentstudy tested the hypothesis that sCR1 might also mitigatepulmonary vascular injury consequent to IL-2 infusion inrats and to investigate whether this action compromisesthe capacity of IL-2 to activate T cells in vitro.

Materials and MethodsDrugsIL-2Human recombinant IL-2, kindly provided by Hoffman

LaRoche Inc, Nutley, NJ, was reconstituted before use with 1mL sterile 0.9% NaCl per 106 U of IL-2.

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

330 Circulation Research Vol 74, No 2 February 1994

TABLE 1. Scheme of the Experimental Protocols

Protocol n Intraperitoneal Bolus Intravenous Infusion (+10 minutes)

1 4 sCR1v sCR1v + IL-2

2a 6 5 mg/kg sCR1 5 mg/kg sCR1 + IL-2

2b 5 15 mg/kg sCR1 15 mg/kg sCR1 + IL-2

3 6 sCR1v sCR1v + IL-2v

sCR1v indicates soluble complement receptor type 1 vehicle; sCR1, soluble complement receptor type 1; IL-2,interleukin-2; and IL-2v, interleukin-2 vehicle.

sCRJsCR1 (BRL 55730, SmithKline Beecham Laboratories,

King of Prussia, Pa) was stored at -5°C. Before use, sCR1 wasdiluted with sterile 0.9% NaCl to give a final concentration of5.56 mg/mL. Endotoxin concentration in the sCR1 preparationwas 0.15 endotoxin units (EU)/mg protein, which equals 15 pglipopolysaccharide (LPS) per mg sCR1 (since 10 EU=1 ngLPS). This endotoxin level is absolutely below doses that haveany effect in animals.27

IndomethacinIndomethacin (Sigma Chemical Co, St Louis, Mo) was

dissolved in 0.2 mol/L Na2CO3 (pH 8) to give a final concen-tration of 1 or 10 mg/mL. Indomethacin vehicle was preparedby dissolving 0.2 mol/L Na2CO3 in 0.9% NaCl at similarvolumes.

AnimalsMale Sprague-Dawley rats (220 to 350 g) were studied. The

rats were kept in groups of three in standard cages with foodand water ad libitum at 22°C and a 12-hour light/dark cycleuntil surgery.

In Vivo Experimental ProtocolA scheme of the experimental protocol is presented in Table

1. Under pentobarbital (30 mg/kg IP) anesthesia, catheters(PE-50) were inserted into a single femoral artery for bloodsampling and into both femoral veins for IL-2 and sCR1infusion. A basal blood sample (0.7 mL exchanged with anequivolume of 0.9% NaCl) was collected for serum thrombox-ane B2 (TXB2), the nonactive metabolite of TXA2, and TNF-adetermination. The animals were randomly assigned to severalgroups: (1) sCR1 vehicle (at a volume similar to that used inthe sCR1 dose of 30 mg/kg) was injected as an intraperitonealbolus. Ten minutes later, sCR1 vehicle at the same dose andIL-2 at 106 U were infused intravenously (using separateintravenous lines) for 1 hour. Blood samples (for the above-mentioned assays) were collected at 0.5, 2, and 4 hours. At theend of the observation period (4 hours), the left lung wasremoved and used to determine lung myeloperoxidase (MPO)activity and lung wet and dry weights. (2) A similar protocolwas repeated using sCR1 at 10 or 30 mg/kg (50% administeredas an initial intraperitoneal bolus and 50% as an intravenousinfusion for 1 hour). (3) The first protocol was repeated butwith sCR1 and IL-2 vehicles. Additional blood samples (0.15mL) for serum sCR1 determination were obtained at severaltime points in animals receiving sCR1 at 30 mg/kg. The IL-2dose was selected on the basis of previous studies that char-acterized the pulmonary responses to IL-2.17 The sCR1 dosewas selected on the basis of previous studies in which similardoses attenuated microvascular lung injury produced by endo-toxin in PAF-primed rats,26 immune complexes,25 or thermalinjury.25,28

Additional protocols were conducted to establish the timecourse of the pulmonary responses to IL-2 infusion. Protocol1 was repeated but with lungs removed at 1 (n=6) or 2 (n=6)hours.

Preliminary data showed increased TXB2 in the serum ofIL-2-treated animals. Thus, to further evaluate the role ofTXA2 in the development of IL-2-induced lung injury, theeffect of TXA2 inhibition by indomethacin was studied. Tothat end, animals were given indomethacin (1 or 10 mg/kg IV,n=5) or vehicle (n=5) 15 minutes before the administration ofIL-2. Control animals given indomethacin (10 mg/kg IV)followed by IL-2 vehicle were also evaluated. Blood wassampled for TXB2 assay as previously described, and lungswere harvested at 4 hours to determine water content.To confirm that IL-2 activates the complement system,

additional rats were given IL-2 (106 U IV) or vehicle (n=4),and blood samples for serum complement hemolytic activitywere collected before and at several time points after IL-2infusion.

In Vitro Experimental ProtocolThe activation of T cells by IL-2 was tested using a prolif-

eration assay in IL-2-dependent T-cell lines. The murineIL-2-dependent cytotoxic T-cell line CTLL-229 was main-tained in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mmol/L sodium pyruvate, 10

50

IC)

CD

(Dz:D

40

30

20

10

A.

E50

040zLLI

zO 300

i2020

Z 10 .

0 -

B.#a,b #a,b

sCR1 (30 mg/kg) _ +sCR1 (10 mg/kg) - _ +sCR1 vehicle +IL-2 + + +

FIG 1. Bar graphs showing the effect of soluble complementreceptor type 1 (sCR1) on lung wet weight (A) and wet minus dryweight (B) response to interleukin-2 (IL-2, 106 U per rat). -indicates absence; +, presence. #P<.05 vs vehicle controlgroup (a) and all other groups (b). Percent change is assessedagainst the vehicle control groups.

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

Rabinovici et al Role of Complement in IL-2-Induced Lung Injury 331

250

200

150

100

50

OsCR1 (30 mg/kg) +sCR1 (10 mg/kg) - - +sCR1 vehicle +IL-2 + + +

FIG 2. Bar graph showing the effect of soluble complementreceptor type 1 (sCR1) on lung myeloperoxidase (MPO) re-sponse to interleukin-2 (IL-2, 106 U per rat). - indicates ab-sence; +, presence. *P<.05 vs vehicle control group (a) and allother groups (b). Percent change is assessed against thevehicle control group.

mmol/L HEPES, 50 ,umol/L 2-mercaptoethanol, and 50 U/mLrecombinant human IL-2 in a humidified 5% CO2 incubator at37°C. The proliferation of CTLL-2 cells in response to IL-2was determined by [3H]thymidine incorporation. Cells werewashed three times with culture media without IL-2 andplated in flat-bottom 96-well plates (104 cells per well). Trip-licate cultures were preincubated for 1 hour with variousamounts of sCR1, and then IL-2 (100 U/mL) was added. Afterincubation for 22 hours, the cultures were pulsed with 1 ,uCi[3H]thymidine for 2 hours and harvested onto glass-fiber filterstrips by a semiautomatic cell harvester, and the radioactivityincorporated into DNA was measured by liquid scintillationcounting.30The murine IL-2-dependent T helper clone L2 was main-

tained as described.3' Seven days after exposure to irradiatedallogenic spleen cells and 5 U/mL recombinant IL-2, the cellswere purified by Ficoll-Hypaque density gradient centrifuga-tion, incubated for 16 hours in culture media without IL-2, andthen used in IL-2-induced proliferation assay as describedabove.

Assays and TechniquesTNF-asSerum levels of TNF-a were measured by a "sandwich"

enzyme-linked immunosorbent assay (ELISA)32 using a ham-ster monoclonal anti-mouse TNF-a (Genzyme Corp, Cam-bridge, Mass) as the capture antibody and a polyclonal rabbitanti-murine TNF-a (Genzyme) as the detecting antibody.

35

30

25

20

1 5

1 0

0

#a,b

0 30

0-0 sCR1 vehicle + IL-2

[1 sCR1 vehicle + IL-2 vehicle

AL-A sCR1 (30 mg/kg) + IL-2V-V sCR1 (10 mg/kg) + IL-2n=5-6

120

TIME (min)

TNF-a levels in rat samples were calculated from a standardcurve generated with recombinant murine TNF-a (Genzyme).TNF-a levels determined by ELISA correlated with levelsdetected by the L-929 bioassay,33 with 1 U of activity in thebioassay corresponding to 5 pg of TNF-a in the ELISA. TheELISA detected levels of TNF-a down to 25 pg/mL.

TXB2TXB2 was determined by radioimmunoassay (sensitivity of

5.0 pg/100 ,uL) as previously described.34

MPOMPO activity in the lung was assayed as described previous-

ly35 on the basis of a modification of Bradley's method36adapted for rat lung MPO assay.37

Pulmonary Water ContentThe removed left lung was immediately frozen on dry ice.

When defrosted, the lung was weighed to determine wetweight. Dry weight was determined after the lung was dried at80°C for 36 hours, and the pulmonary water content wascalculated by subtracting the lung dry weight from the wet lungweight.

Serum sCRI Antihemolytic ActivitySerum sCR1 levels were assayed using an antihemolytic

assay, which measures the inhibition of complement-mediatedsheep erythrocyte hemolysis as a function of sCR1 concentra-tion. The assay was previously described in detail.26

Serum Complement Hemolytic ActivityComplement activity of rat serum was determined by mea-

suring the ability of the serum to lyse antibody-sensitizedsheep red blood cells. In a V-bottom plate, sensitized sheepred blood cells at 1x 107 cells per milliliter (Diamedix, Miami,Fla) were incubated with serial dilutions of rat serum usingHEPES buffer (0.1% mol/L HEPES, 0.15 mol/L NaCl, and0.1% bovine serum albumin, pH 7.4) for 60 minutes at 37°C.Cells were pelleted, the supernatants were transferred to aflat-bottom microtiter plate, and the absorbance at 405 nm wasmeasured to determine released hemoglobin. All serum sam-ples were run in duplicate.

Data AnalysisData in text and figures are mean-+±SEM for the indicated

number of animals. ANOVA followed by the Student-New-man-Keuls test for multiple comparisons was used for statis-tical analysis, with a value of P<.05 considered significant.

FIG 3. Graph showing the effect of soluble complementreceptor type 1 (sCR1) on serum thromboxane B2 (TXB2)response to interleukin-2 (IL-2, 106 U per rat). #P<.01 vsbasal value (a) and sCR1 vehicle + IL-2 vehicle group (b).The dashed line represents the sensitivity of the assay (5pg/1 00 ,L).

240

0

0ziD-J

00

\-

m

:DCLLLIcr

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

332 Circulation Research Vol 74, No 2 February 1994

TABLE 2. Time Course of Lung Myeloperoxidase and Weight Response toInterleukin-2 Infusion

Control Al h,% A2h,% A4h,%MPO 10±1 U/g wet weight 23±26 13±1 42+9*

Wet weight 463+28 mg 5+12 -1±7 41±5*

Dry weight 92±5 mg 2±10 1±6 38±7*

Wet-dry weight 369±15 mg 5±7 3±6 36±6*

Control indicates values at 4 hours of the soluble complement receptor type 1 vehicle plusinterleukin-2 vehicle group; A, percent change vs control; and MPO, myeloperoxidase.*P<.05 vs control.

ResultsEffect of sCRi on IL-2-Induced LungWeight Response

Administration of IL-2 increased basal lung wet(463±28 mg), dry (92±5 mg), and wet minus dry(369±15 mg) weights by 41±5% (P<.05), 38+7%(P<.05), and 36±6% (P<.05), respectively (Fig 1).These results are consistent with previous reports.10,17sCR1 at 30 mg/kg, but not at 10 mg/kg, attenuated theseresponses (P<.05, Fig 1).

Effect of sCRi on IL-2-Induced LungMPO Response

IL-2 infusion elevated lung MPO activity by 42±9%(P<.05, Fig 2) as compared with vehicle-treated controllungs (10.3 ±0.6 U/g wet lung weight). These results arein agreement with previous reports.17 sCR1 at 30 mg/kgattenuated lung MPO activity by 74±8% (P<.05, Fig 2),whereas sCR1 at 10 mg/kg failed to alter lung MPOresponse.

Effect of sCR1 on IL-2-Induced TXB2 andTNF-a Response

Basal serum TXB2 and TNF-a levels were below thesensitivity of the assays used as previously reported.37-40IL-2 increased serum TXB2 (P<.01, Fig 3) but failed toelevate serum TNF-a (data not shown). sCR1 at alltested doses did not alter the IL-2-induced elevation ofserum TXB2 (Fig 3).

Time Course of Lung Weight and MPOResponse to IL-2At 4 hours after infusion, IL-2 elevated lung MPO

activity (P<.05, Fig 2), wet weight (P<.05, Fig 1A), andwater content (P<.05, Fig 1B); no significant changeswere recorded earlier at 1 or 2 hours (Table 2).

Effect of Indomethacin on IL-2-Induced LungWeight Response

Pretreatment with indomethacin at all tested dosesfailed to attenuate lung edema induced by IL-2 (Fig4A). In contrast, indomethacin at 10 mg/kg completelyprevented the IL-2-induced elevation of serum TXB2(Fig 4B). Indomethacin alone did not produce anydetectable change in lung weight (Fig 4A).

Serum sCR1Detectable levels of sCR1 were found in the serum at

30 minutes after sCR1 administration (and before lunginjury) (Fig 5). Serum sCR1 levels remained the samethroughout the experiment period.

Effect of IL-2 on Serum ComplementHemolytic Activity

IL-2 but not IL-2 vehicle significantly reduced serumcomplement activity at 2 hours (P<.01) and 4 hours(P<.01) after infusion (Fig 6).

Effect of sCRi on IL-2-Induced (3H]ThymidineIncorporation to L-2 or CTLL-2 Cells

IL-2 markedly stimulated [3H]thymidine incorpo-ration in both L-2 and CTLL-2 cell lines (Table 3).sCR1 failed to alter this response. Furthermore, sCR1had no effect of its own on [3H]thymidine incorporationinto L-2 or CTLL-2 cells.

50

I- 40z

zo 30

3:CDg 20

Z 10_i

A.#a #a

+

35

o 300

XI~ 25

3 20

m 15

> 10:DX 5Lw PCn,)

0 30

+ +

_ + _

#b0 0-0 Indo ve

- g\\ O O-E indo (1A-A indo (1

q ~~V-V indo (1n=5

It- ----

B.chicle + IL-2mg/kg) + IL-2

10 mg/kg) + IL-210 mg/kg) + IL-2 vehicle

120

TIME (min)240

FIG 4. Graphs showing the effect of indomethacin (indo) onlung weight (A) and thromboxane B2 (TXB2, B) response tointerleukin-2 (IL-2). #P<.05 vs basal value, control, andindo + IL-2 vehicle group (a) and basal value, 10 mg/kgindomethacin + IL-2 group, and 10 mg/kg indomethacin + IL-2vehicle group (b). The dashed line in panel B represents thesensitivity of the TXB2 assay (5 pg/100 ,L).

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

Rabinovici et al Role of Complement in IL-2-Induced Lung Injury 333

# a,b# a,b T

v~~~~T7 c=ab #v

0-0 sCR1 vehicle + IL-2 (n=3)V-V sCR1 (30 mg/kg) + IL-2 (n=6)El-El sCR1 vehicle + IL-2 vehicle (n=2)

IL-2, i.v.CR1 (1 5mg/k )

" _E F: -- 7

FIG 5. Graph showing serum soluble complement re-ceptor type 1 (sCR1) levels. sCR1 was given as anintraperitoneal bolus at -10 minutes (15 mg/kg). Tenminutes later, sCR1 (15 mg/kg) and interleukin-2 (IL-2,106 U) were infused intravenously for 1 hour (throughseparate intravenous lines). #P<.01 vs basal value (a)and all other groups (b).

120

TIME (min)

DiscussionOur demonstration of acute lung microvascular injury

characterized by capillary permeability defect and leu-kosequestration as early as 4 hours after IL-2 infusioncorrelates well with recent reports from our laboratory17and by others7-10'16 and is in accord with observationsfrom clinical studies.344' In these reports, intravenousinfusion of IL-2 to rats and sheep induced microvascu-lar injury within 3 to 6 hours,7-10,16 and IL-2-treatedpatients developed pulmonary edema associated withacute respiratory failure within minutes of IL-2 admin-istration.3 4 4' In addition, our results indicate thatchanges in both IL-2-induced permeability defect andleukosequestration begin simultaneously (see Table 2).Taken together, these data suggest that IL-2 inducesacute lung microvascular injury and point to the poten-tial role of IL-2 in the mediation of other acute inflam-matory responses.The major finding of the present study is that

treatment with sCR1 attenuated the IL-2-inducedmicrovascular injury. Since sCR1 is a highly specific

inhibitor of the complement cascade,24 it is conceiv-able that complement factors play a key role in thepathophysiological processes by which IL-2 induceslung injury. This conclusion is supported by our ob-servation that IL-2 activates serum complement and iscomplementary to a previous report that demon-strated activation of the complement system duringimmunotherapy with IL-2 in humans.19 Also, theefficacy of sCR1 to block lung injury produced by IL-2without affecting the activation of T cells by LL-2suggests that sCR1 might be effective in combatingIL-2 toxicity without compromising the primary ther-apeutic goal of selected immune stimulation.The mechanisms by which IL-2 induces complement

activation are still obscure. Nevertheless, it is possiblethat IL-2 might indirectly activate the complementcascade through the production of other inflammatorymediators such as TNF-a or IL-1, which can recruit andactivate neutrophils to release proteases42 and toxicoxygen metabolites43 that cleave complement factors.Our finding of elevated lung MPO activity along with a

TABLE 3. Effect of Soluble Complement Receptor Type 1 on lnterleukin-2-lnduced [3H]ThymidineIncorporation to L-2 or CTLL-2 Cells

Experiment 1: Experiment 2: Experiment 3:Treatment CTLL-2, cpm CTLL-2, cpm L-2, cpm

IL-2v + sCR1v 1710 245 211

IL-2 65 580 29 113 4002

IL-2 + 0.01 ,ug/mL sCR1 ND 29 184 4163

IL-2 + 0.10 ,tg/mL sCR1 65 136 28 369 6210

IL-2 + 1.00 ,ug/mL sCR1 56 752 31 969 5710

IL-2 + 5.00 ,ug/mL sCR1 58 383 ND 5341

IL-2 + 10.00 ,ug/mL sCR1 66 532 ND ND

IL-2v + 0.01 ,ug/mL sCR1 ND 252 243

IL-2v + 0.10 ,tg/mL sCR1 ND 263 185

IL-2v + 1.00 ,ug/mL sCR1 ND 241 201

IL-2v + 5.00 ,ug/mL sCR1 ND ND 225

IL-2v indicates interleukin-2 vehicle; sCR1v, soluble complement receptor type 1 vehicle; IL-2, interleukin-2;sCR1, soluble complement receptor type 1; and ND, not done.

Cells were incubated with IL-2 (100 U/mL) and with various amounts of sCR1 for 22 hours and then pulsed with[3H]thymidine (1 liCi) for 2 hours. Results represent average of triplicates with <10% variability.

80

E

Uc

60 [

40 [

20

EM0E ,

L CO) 0

0 d.

30 60 240M~_k 1-4

1

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

334 Circulation Research Vol 74, No 2 February 1994

E

zl

4-j

0

a)

C

Uc0-Q0C/)

-n

0.300

0.200

0.100

0.000

=ZI IL-2

IL-2

n=4

basal 2 hr 4 hr

FIG 6. Bar graph showing the effect of interleukin-2 (IL-2) onserum complement hemolytic activity. The extent of hemolysiswas determined by the absorbance (405 nm) of the sampleusing a 1:320 dilution of serum subtracted from the backgroundabsorbance due to serum alone. A decrease in hemolysisindicates a decrease in serum complement activity because ofprior complement activation. *P<.01 vs all other groups.

previous report demonstrating increased pulmonaryneutrophil count in IL-2-treated rats16 strongly sup-ports this possibility. Complement activation might alsooccur in response to PAF, a potent phospholipid inflam-matory mediator (for review see Reference 44), whichhas been reported to mediate both IL-2-induced lunginjury17 and complement-dependent processes such asred cell phagocytosis.45

IL-2 caused leukosequestration, as evidenced by ele-vated pulmonary MPO activity. This is in accord withrecent reports demonstrating leukocyte accumulation inlung tissue16 and increased neutrophil-endothelial ad-herence in the skeletal muscle microcirculation of IL-2-treated rats.12,46 Currently, no information is avail-able as to the processes leading to IL-2-inducedleukocyte adherence and emigration into lung paren-chyma. Potential mechanisms are leukocyte and endo-thelial cell activation by (1) inflammatory mediatorssuch as IL-147 or PAF,44 which are produced in responseto IL-2,13,'7 and (2) complement anaphylatoxins, whichhave been shown to carry chemotactic properties and toinduce neutrophil aggregation (for review see Refer-ence 48). The latter possibility is supported by ourobservation that treatment with sCR1 attenuated theIL-2-induced elevation of lung MPO activity; however,our data do not allow us to differentiate between IL-2action on the leukocytes, endothelium, or both cellularelements.Although our data support a potential therapeutic

role for sCR1 and complement inhibitors in the pulmo-nary complications of IL-2, it should be noted thatinhibition of the complement system, a major compo-nent of the host defense mechanism,49 may lead toimmune suppression and enhanced susceptibility toinfections. Since sCR1 effectively blocks both the alter-nate and classic pathways of complement activation, itmight carry liabilities of complement suppression. How-ever, such a possibility is more likely to be associatedwith chronic inhibition. In contrast, the therapeuticpotential of sCR1 in IL-2-induced lung injury is likelyto overshadow any potential immune suppression, be-cause treatment is acute, and serum levels will precipi-

tously decline on cessation of infusion (time to halfactivation, -100 minutes).TXA2, a key prostaglandin mediator of inflammation

(for review see Reference 50), was elevated after IL-2infusion. A similar response, which suggests TXA2mediation of IL-2-induced lung injury, was previouslyreported by our group17 and by others.'0 In contrast,several observations suggest that TXB2 is not directlyinvolved in the pathogenesis of IL-2-induced lunginjury. First, inhibition of TXB2 production by indo-methacin failed to attenuate IL-2-induced lung injury.Second, sCR1 attenuated the lung injury produced byIL-2 without affecting the elevated serum TXB2. Thisphenomenon also suggests that complement is not in-volved in the processes leading to TXB2 productionafter IL-2 administration. The source of TXA2 produc-tion in response to IL-2 is still obscure. Nevertheless,one potential origin could be activated neutrophils.51However, the reduction of lung MPO activity by sCR1without attenuation of the elevated serum TXB2 doesnot support this possibility. Alternatively, TXA2 couldbe produced by aggregating platelets in response toIL-2.12 In that respect, platelet activation by IL-2 mightresult from stimulation by PAF, since rats pretreatedwith a PAF antagonist failed to elevate serum TXB2 inresponse to IL-2 infusion.17

In conclusion, the present study suggests that thepulmonary toxicity of IL-2 is mediated in part bycomplement factors and points to the potential thera-peutic capacity of complement inhibitors in combatingthis grave complication of IL-2 immunotherapy.

References1. Rosenberg SA, Lotze MT, Muul LM, et al. A progress report on

the treatment of 175 patients with advanced cancer usinglymphokine-activated killer cells and interleukin-2 or high doseinterleukin-2 alone. N EnglJ Med. 1987;316:889-897.

2. West WH, Tauer KW, Yanelli JR, Marshall GD, Orr DW,Thurman GB, Oldham RK. Constant infusion recombinantinterleukin-2 in adoptive immunotherapy of advanced cancer.N Engl J Med. 1987;316:898-905.

3. Lotze MT, Matory YL, Rayner AE, Ettinghausen S, Vetto J, SeippCA, Rosenberg SA. Clinical effects and toxicity of interleukin-2 inpatients with cancer. Cancer. 1986;58:2764-2772.

4. Siegal JP, Puri RK. Interleukin-2 toxicity. J Clin Oncol. 1988;15:10-26.

5. Damle NK, Doyle LV. IL-2-activated killer lymphocytes but nottheir secreted products mediate increase in albumin flux acrosscultured endothelial monolayers. J Immunol. 1989;142:2660-2669.

6. Damle NK, Doyle LV, Bender JR, Bradley EC. Interleukin-2-activated human lymphocytes exhibit enhanced adhesion to normalvascular endothelium cells and cause their lysis. J Immunol. 1987;138:1779-1785.

7. Klausner JM, Morel N, Paterson IS, Kobitz L, Valeri CR, SheproD, Hechtman HB. The rapid induction by interleukin-2 of pul-monary microvascular permeability. Ann Surg. 1989;209:119-128.

8. Klausner JM, Paterson IS, Morel NML, Goldman G, Gray AD,Valeri R, Eberlein TJ, Shepro D, Hechtman HB. Role of throm-boxane in interleukin-2-induced lung injury in sheep. Cancer Res.1989;49:3542-3549.

9. Harms BA, Pahl AC, Polham TH, Conhaim RL, Starling JR,Storm FK. Effects of interleukin-2 on pulmonary and systemictransvascular fluid filtration. Surgery. 1989;106:339-346.

10. Welbourne R, Goldman G, Kobitz L, Valeri CR, Shepro D,Hechtman HB. Involvement of thromboxane and neutrophils inmultiple-system organ edema with interleukin-2. Ann Surg. 1990;212:728-733.

11. Lotze MT, Matory YL, Ettinghausen SE, Rayner AA, SharrowSO, Seipp CA, Custer MC, Rosenberg SA. In vivo administrationof purified interleukin-2, II: half-life, immunologic effects andexpansion of peripheral lymphoid cells in vivo with recombinantIL-2. J Immunol. 1985;135:2865-2875.

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

Rabinovici et al Role of Complement in IL-2-Induced Lung Injury 335

12. Edwards MJ, Miller FN, Sims DE, Abney DL, Schuschke DA,Corey TS. Interleukin-2 acutely induces platelet and neutrophil-endothelial adherence and macromolecular leakage. Cancer Res.1992;52:3425-3431.

13. Numeroff RP, Aronson FR, Mier JW. IL-2 stimulates the pro-duction of IL-1 alpha and IL-1 beta by human peripheral bloodmononuclear cells. J Immunol. 1988;141:4250-4257.

14. Heslop HE, Gottlieb DJ, Bianchi ACM, Meagher A, Prentice HG,Mehta AB, Hoffbrand AV, Brenner MK. In vivo induction ofgamma interferon and tumor necrosis factor by interleukin-2infusion following intensive chemotherapy or autologous marrowtransplantation. Blood. 1989;74:1374-1380.

15. Nedwin GE, Svedersky ELP, Bringman TS, Palladino MA, GoedelDV. Effect of interleukin-2, interferon a-, and mitogens on theproduction of tumor necrosis factor a and f. J Immunol. 1985;135:2492-2496.

16. Welbourne R, Goldman G, Kobitz L, Paterson I, Shepro D,Hechtman HB. Interleukin-2 induces early multisystem organedema mediated by neutrophils. Ann Surg. 1991;214:181-186.

17. Rabinovici R, Sofronski MD, Renz JF, Hillegas LM, Esser KM,Vernick J, Feuerstein G. Platelet activating factor mediatesinterleukin-2 induced lung injury in the rat. J Clin Invest. 1992;89:1669-1673.

18. Rosenstein M, Ettinghausen SE, Rosenberg SA. Extravasation ofintravascular fluid mediated by the systemic administration ofrecombinant interleukin-2. J Immunol. 1986;137:1735-1742.

19. Thijs LG, Hack CE, Strack Van Schijndel RJM, Nuijens JH,Wolbink GJ, Belmer AJM, Van Der Vall H, Wagstaff J. Activationof the complement system during immunotherapy with recom-binant IL-2. J Immunol. 1990;144:2419-2424.

20. Antrum RM, Solomkin JS. Monocyte dysfunction in severetrauma: evidence for the role of C5a in deactivation. Surgery.1986;100:29-37.

21. Shatney CH, Benner C. Sequential serum complement (C3) andimmunoglobulin levels in shock/trauma patients developing acutefulminant systemic sepsis. Circ Shock. 1985;16:9-17.

22. Stevens JH, O'Hanley P, Shapiro JM, Mihm FG, Satoh PS, CollinsJA, Raffin TA. Effects of anti-C5a antibodies on the adult respi-ratory distress syndrome in septic primates. J Clin Invest. 1986;77:1812-1816.

23. Parsons PE, Giclas PC. The terminal complement complex(sC5b-9) is not specifically associated with the development of theadult respiratory distress syndrome. Am Rev Respir Dis. 1990;141:98-103.

24. Weisman HF, Bartow T, Leppo MK, Marsh HC, Carson GR,Concino MF, Boyle MP, Roux KH, Weisfeldt ML, Fearon DT.Soluble human complement receptor type 1: in vivo inhibitor ofcomplement suppressing post-ischemic myocardial inflammationand necrosis. Science. 1990;249:146-151.

25. Mulligan MS, Yeh CG, Rudolph AR, Ward PA. Protective effectsof soluble CR1 in complement- and neutrophil-mediated tissueinjury. J Immunol. 1992;148:1479-1485.

26. Rabinovici R, Yue G, Hillegas LM, Griswold DE, DiMartino MJ,Vernick J, Fong K-LL, Feuerstein G. The role of complement inendotoxin/PAF-induced lung injury. J Immunol. 1992;149:1744-1750.

27. Feuerstein G, Hallenbeck JM, Vanatta B, Rabinovici R, PereraPY, Vogel SN. Effect of gram-negative endotoxin on levels ofserum corticosterone, TNFa, circulating blood cells, and thesurvival of rats. Circ Shock. 1990;30:265-278.

28. DiMartino MJ, Wolff CE, Slivjak MJ, Minthorn EA, Feuerstein G.Effects of BRL 55730 (sCR1) on thermal skin injury inducedhemoconcentration and lung inflammation in rats. PharmacolCommun. In press.

29. Gillis S, Smith KA. Long term culture of tumour-specific cytotoxicT cells. Nature. 1987;268:154-156.

30. Feuerstein N, Mond JJ. "Numatrin," a nuclear matrix proteinassociated with induction of proliferation in B lymphocytes. J BiolChem. 1987;262:11389-11397.

31. Shipman-Appasamy P, Cohen KS, Prystowsky MB. Interleukin2-induced expression of proliferating cell nuclear antigen is reg-

ulated by transcriptional and post-transcriptional mechanisms.J Biol Chem. 1990;265:19180-19184.

32. Hornbeck P, Winston SE, Fuller SA. Enzyme linked immuno-sorbent assays (ELISA) for detection of antigens. In: Ausubel FM,Brente R, Kingston EE, Moore DD, Seidman JG, Smith JA,Strubel K, eds. Current Protocols in Molecular Biology. New York,NY: John Wiley & Sons, Inc; 1987:11.2.1.

33. Ruff MR, Gifford GE. Purification and characterization of rabbittumor necrosis factor. J Immunol. 1980;125:1671-1677.

34. Shohami E, Rosenthal J, Lavy S. The effect of incomplete cerebralischemia on prostaglandin levels of rat brain. Stroke. 1982;13:494-499.

35. Griswold DE, Hillegass L, Hill DE, Sherief HT, Kopia GA. Eval-uation of the effects of Evan's blue and triphenyltetrazoliumchloride dyes on myeloperoxidase activity in canine cardiac tissue.J Pharmacol Methods. 1989;21:13-19.

36. Bradley PP, Priebat DA, Christenson RD, Rothstein G. Mea-surement of cutaneus inflammation: estimation of neutrophilcontent with an enzyme marker. J Invest Dermatol. 1982;78:206-209.

37. Rabinovici R, Esser KM, Lysko PG, Yue TL, Griswold DE,Hillegas LM, Bugelski PJ, Hallenbeck JM, Feuerstein G. Primingby platelet activating factor of endotoxin induced lung injury andcardiovascular shock. Circ Res. 1991;69:12-25.

38. Rabinovici R, Yue T-L, Farhat M, Smith EF, Esser KM, Slivjak M,Feuerstein G. Platelet activating factor (PAF) and tumor necrosisfactor-a (TNFa) interactions in endotoxemic shock: studies withBN 50739, a novel PAF antagonist. J Pharmacol Exp Ther. 1990;255:256-263.

39. Rabinovici R, Rudolph AS, Feuerstein G. Characterization ofhemodynamic, hematologic and biochemical responses to admin-istration of liposome encapsulated hemoglobin in the consciousfreely moving rat. Circ Shock. 1989;29:115-132.

40. Anderegg K, Anzeveno P, Cook JA, Halushka PV, McCarthy J,Wagner E, Wise WC. Effects of a pyridine derivative thromboxanesynthetase inhibitor and its inactive isomer in endotoxic shock. BrJ Pharmacol. 1983;78:725-732.

41. Parkinson DR. Interleukin-2 in cancer therapy. Semin Oncol. 1991;9:694-704.

42. Le J, Vilcek J. Biology of disease: tumor necrosis factor andinterleukin-1: cytokines with multiple overlapping biologicalactivities. Lab Invest. 1987;56:234-248.

43. McCabe WR, Treadwell T, DeMaria A. Pathophysiology of bac-teremia. Am J Med. 1983;75(suppl 1B):7-18.

44. Braquet P, Touqui L, Shen T, Vargaftig BB. Perspectives inplatelet activating factor. Pharmacol Rev. 1987;39:97-145.

45. Lutz HU, Bossolino RF, Fasler S, Stammler P, Kazatchk MD,Arese P. Naturally occurring anti-band-3 antibodies and com-plement together mediate phagocytosis of oxidatively stressedhuman erythrocytes. Proc NatlAcad Sci U SA. 1987;84:7368-7372.

46. Ohkubo C, Bigos D, Jain RK. Interleukin-2 induced leukocyteadhesion to the normal and tumor microvascular endothelium invivo and its inhibition by dextran sulphate: implications forvascular leak syndrome. Cancer Res. 1991;51:1561-1563.

47. Bussolino F, Brevario F, Aligietta M, Sanavio F, Bosia A, DejanaE. Studies on the mechanism of interleukin-1 stimulation ofplatelet activating factor synthesis in human endothelial cells inculture. Biochem Biophys Acta. 1987;927:43-54.

48. Gallinaro R, Cheadle WG, Applegate K, Polk HC. The role of thecomplement system in trauma and infection. Surg Gynecol Obstet.1992;174:435 -440.

49. Klein J. Complement. In: Klein J, ed. Immunology. Boston, Mass:Blackwell Scientific; 1991:217-220.

50. Lefer AM. Role of prostaglandins and thromboxanes in shockstates. In: Altura BM, ed. Hand Book of Shock and Trauma. NewYork, NY: Raven Press, Publishers; 1983:355-376.

51. Simmet T, Luck W. Effect of Hi-antihistamines and calmodulinantagonists on the ionophore A23187-induced eicosanoid releasefrom human leukocytes. In: Schror K, Sinzinger H, eds. Prosta-glandins in Clinical Research. New York, NY: Alan R. Liss, Inc;1989:529-533.

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from

M Scesney, N Feuerstein and G FeuersteinR Rabinovici, M D Sofronski, P Borboroglu, A M Spirig, L M Hillegas, J Levine, J Vernick, S

Interleukin-2-induced lung injury. The role of complement.

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1994 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.74.2.3291994;74:329-335Circ Res. 

http://circres.ahajournals.org/content/74/2/329World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on May 16, 2018

http://circres.ahajournals.org/D

ownloaded from