Int Arch Allergy Appl Immunol 1991;96:119-127
©1991 S.Karger AG. Basel 0020-5915/91/0962-0119 $ 2.75/0
Experimental Lung Injury Induced by Trimellitic Anhydride Inhalation on Guinea Pigs Yoshiaki Tao3, Tsutomu Sugiura3, Hiroshi Nakamura3, Masamitsu Kidoh, Isamu Tanakac, Akio Kuroiwad ■'Department of Immunology. School of Medicine, b Division of Respiratory Disease. University Hospital. c Department of Environmental Health Engineering, Institute of Industrial Ecological Sciences,d Second Department of Internal Medicine. University of Occupational and Environmental Health, Kitakyushu. Japan
Introduction
Trimellitic anhydride (TMA) is a low molecular weight, water-insoluble, biologically reactive chemical compound widely used in the production of resins, ad hesives, polymers dyes and printing inks. Inhalation exposure to TMA has been recognized as causing four clinical syndromes in man; of these, asthma-rhinitis, late respiratory systemic syndrome and pulmonary' disease-anemia syndrome are immunologically medi ated. The mechanism of the fourth syndrome is based on an irritant reaction [1-4]. It has been suggested that the levels of antibodies specific for trimellitylated (TM) epitopes in the sera of patients are associated with either the asthma-rhinitis or the late respiratory systemic syndrome, and that different immunoglobu lin classes of antibodies are involved in a variety of re
spiratory syndromes [1. 3]. To further investigate the mechanisms of immunopathogenic response induced by TMA inhalation, an experimental animal model of lung injury caused by TMA inhalation was established by Leach et al. [5] and by Zeiss et al. [6] using specific pathogen-free (SPF) Sprague-Dawley (SD) rats. The results suggested that the inhalation of TMA was a significant stimulus of a systemic immune response that occurred parallel with lung injury. Furthermore, the results of inhibition studies on antibody specificity using various kinds of protein conjugates with TM epitopes suggested that the TMA inhalation resulted in antibody response with specificity for new antigenic determinants that arose from the coupling of TMA with autologous respiratory tract proteins [7-10], The activation mechanism of immune systems by TMA, however, is not yet fully understood.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Abstract. Trimellitic anhydride (TMA) causes lung injury by inhalation exposure in humans. In order to in vestigate more precisely the mechanism of lung injury by TMA, an experimental animal model of TMAinduced lung injury was established. Guinea pigs were intramuscularly injected with trimellitylated bovine se rum albumin (TM BSA) with complete Freund’s adjuvant (CFA). Appreciable amounts of antibodies against TM epitopes were detected in the sera of these animals. Guinea pigs that were passively sensitized with antiTM antisera as well as the actively immunized animals developed hemorrhagic pneumonitis after TMA in halation. It is well recognized that hyperimmune antisera of guinea pigs contain two subclasses of IgG antibod ies, IgGl and IgG2, which are known as homocytotropic and heterocytotropic antibodies, respectively. To de termine the role of these antibodies in the airway injury with alveolar hemorrhages, they were isolated by gel filtration and by ion exchange column chromatography, and the guinea pigs that were sensitized with each of these antibodies were exposed to TMA inhalation. The extent of lung injury in these animals was quantitatively determined by the measurements of lung extravasation of Evans blue dye injected intravenously after TMA in halation. Significant increases in the extravasation of dye were observed in both animal groups sensitized with IgGl and IgG2 antibodies. In addition, results obtained with heat-treated antisera and IgGl antibody did not significantly differ from those obtained with untreated samples. These results suggest that the lung injury re sulting from TMA inhalation exposure can be induced with humoral antibodies, not only IgGl but also IgG2, and that at least two types of allergic reactions are involved in the pathogenesis of lung injury.
Tao/Sugiura,Nakamura'Kido/Tanaka/Kuroiwa
120
Fig. 1. a Gel filtration chromatography of anti-TM antibody IgG fraction on a column of Sephacryl S-300 (3x51 cm) equilibrated with PBS (pH 7.0). The elution was performed with the same buffer solution and the optical absorbance at 280 nm wavelength and the antibody content of elute (3.1 ml) was measured. The antibody activity was detectable in the second peak corresponding to IgG with a molecular weight of about 160,000. These fractions were pooled, concentrated and offered to further purification. Blue dextran. OVA and human IgG were used as molecular weight markers, b DEAE-cellulose column chromatography of the IgG peak by gel filtration chromatography. The elution was performed stepwise by changing pH and ionic strength of the buffer solution as induced in the figure. Each of the peak components was pooled and denoted as P-1. P-2, P-3 or P-4.
Materials and Methods Experimental Design Guinea pigs were directly immunized with two or four intra muscular injections of TM conjugates with CFA. As antiserum do nors, guinea pigs were immunized by injections of TM conjugates with complete Freund’s adjuvant (CFA) into their footpads fol lowed by three intradermal injections with the same antigen at 3week intervals. In this study, four experimental approaches were undertaken: (1) to prepare TM protein antigens that would pro duce sufficient amounts of the specific antibodies against TM epi topes in guinea pigs; (2) to determine the immunization schedules of guinea pigs for developing lung injury after TMA inhalation; (3) to sensitize guinea pigs passively with anti-TM antisera, and (4) to sensitize guinea pigs with a passive transfer of isolated IgGl and/or IgG2 antibodies. In addition, the immunological character
ization of the isolated antibodies were carried out using passive hemagglutination, complement fixation and passive cutaneous anaphylaxis (PCA). Animals Random-bred female Hartley guinea pigs (200-500 g) were maintained in the animal research facilities at the university and were carefully examined for health defects prior to use in investi gations. Reagents TMA was purchased from Nacaiai Tesque. Kyoto, Japan, in the form of small Oakes. Bovine serum albumin (crystalline BSA: Armour Pharmaceutical Co.. Kankance. III. USA) and ovalbumin (crystalline OVA; Sigma Chemical Co.. St. Louis. Mo.. USA) were used for the preparation of TM protein conjugates. Sheep eryth rocytes (SRBC) were supplied by Nippon Bio-Supply Center, To kyo. Japan. Anti-SRBC antibody was prepared by using rabbit an tisera against SRBC. Anti-OVA antibody was prepared by using guinea pig antisera against OVA. Normal guinea pig IgG was pre pared by ammonium sulfate precipitation at 40% saturation. Goat antibodies specific for guinea pig IgGl and IgG2 were prepared in our laboratory [II], TM Protein Conjugates The conjugation of TM epitopes to protein was carried out us ing BSA or OVA according to the method deset ibed by Zeiss et al. [1] with slight modifications. Briefly, a BSA or OVA solution con taining a concentration of 5 mg protein/ml of 7% sodium bicarbo nate solution was prepared. To 20 ml of the protein solution were added various amounts of TMA dissolved in 1 ml of dioxane and the resulting mixture was stirred for 30 min at room temperature. The sample solution was dialyzed against 0.1 M NaHCO, over
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
In the present report, it is demonstrated that lung injury can be induced by TMA inhalation in guinea pigs that are not only actively immunized with TM bo vine serum albumin (BSA) conjugates but also pas sively sensitized with a homologous antibody, IgGl or IgG2, and that different types of allergic reactions may be involved in the process of lung injury. In addi tion, the involvement of another class of homocytotropic antibody, IgE, is probably excluded by the re sults of heat treatment of antibodies. This animal model of lung injury induced by TMA inhalation pro vides a useful system for investigating the immunopathological mechanisms of the syndrome.
night, then the outer solution changed to phosphate-buffered sa line (PBS, pH 7.0). The number of TM epitope conjugated with the protein molecule was assessed by the measurements of optical absorbances at 240 and 280 nm (Hitachi model U-1100 spectropho tometer; Tokyo, Japan) using molar extinction coefficients 11.400/ 1,700 (A240 nm/AMI nm ) for NaTM. and 74.800/43,500 and 62.100/ 33,300 for BSA and OVA. respectively. TM protein conjugates with various amounts of the epitopes were obtained and the num ber of conjugated TM residues were expressed in small figures such as TM„, BSA and TMI4 OVA. Hyperimmune Serum Guinea pigs had I mg TM36 BSA emulsified in CFA injected into their footpads on days 0 and 14 followed by intradcrmal in jections with the same antigen on days 35, 56 and 77. The animals were exsanguinated and the antisera were fractionated by ammo nium sulfate precipitations at 40% saturation. The precipitate, IgG fraction, was centrifuged and dissolved in PBS and extensively dialyzed against PBS with several changes in the outer solution. IgG fraction was further purified by gel filtration chromatography on a column of Scphacryl S-300 in PBS, and the elution profile was monitored by optical density measurements at 280 nm. as shown in figure la. The fractions containing IgG were pooled and concentrated by evaporation, and then thoroughly dialyzed against a 10 mM phosphate buffer (pH 8.0) for further purification by DEAE-cellulose (Whatman DE-52) column chromatography according to the method described by Liu et at. [12] with slight modifications. The elution was performed by changing pH and the ionic strength of the buffer solution with NaCI. One of the typical elution profiles is shown in figure lb. The immunochemical prop erty of each fraction was examined by immunoelectrophoresis and the double diffusion test using goat antisera specific for guinea pig IgG! and/or IgG2. The fractions containing single IgG subclass were then pooled and concentrated. Hemagglutination Test The sensitization of SRBC with TM BSA or TM OVA was per formed according to the method described by Avrameas et al. [13]. The anti-TM antibodies in antisera were measured by passive hemagglutination using sensitized SRBC. The sample solutions were incubated at 56 °C for 30 min to inactivate complement com ponents and absorbed with SRBC treated with 1% glutaraldehyde and glycine solution. They were then 2-fold serially diluted with gelatinized veronal buffer (pH 7.4. 4.9 mM veronal, 0.5 mM MgS04, 0.75 mM CaCL, 146 mM NaCI and 0.2% gelatin) in micro titer plates. To the sample solutions (25 pi) was added a 1% sus pension of TM|4 OVA SRBC (25 pi), mixed and incubated at room temperature for 90 min. The antibody titer was expressed as the reciprocal of the highest dilution of samples giving positive hemagglutination. A normal serum and an anti-OVA serum were used as control. Complement Fixation Test Complement fixation tests were carried out using the same ve ronal buffer solution as described above. A normal guinea pig se rum was used as a source of complement (130 CHS0). To the dilut ed sample solutions containing anti-TM antibodies in microtiter plates with round-bottom wells were added 25 ul of the serial 2fold dilutions of 4 mg/ml TM,4 OVA with diluted normal guinea pig serum containing 2.5 CHS0. After incubation at 37 °C for 2 h.
121
25 pi of 1% suspension of SRBC sensitized with rabbit anti-SRBC antibody (hemolysin) was added to each well. One hour later, the titer was determined from the highest dilution of specimens with complete hemolysis. Tussive Cutaneous Anaphylaxis Two aliquots of serial 2-fold dilutions of the samples with PBS were prepared. One of them, without any pretreatment, was used for the induction of PCA reaction, and the other was incubated at 56 °C for 2 h prior to its use in the PCA reaction. A 0.1-ml portion of the sample was injected intradermally into normal guinea pigs, and 6 h later. 1 ml of PBS containing 1 mg of TM14 OVA and 10 mg of Evans blue dye (Sigma Chemical Co.) were injected intrave nously into the guinea pigs. At 30 min after the injection, animals were sacrificed and the diameter of the blue spot developed at the antibody injection site was measured on the inner surface of the flayed skin. The PCA titer of the specimen was determined as the highest dilution having a blue spot size of 10 x 10 mm [14]. Exposure System of TMA Inhalation The exposure of TMA inhalation was performed by the same system described by Tanaka et al. [15] using TMA instead of fly ash. The airflow in the chamber was downward and laminar with a volume of 0.1 m \ The mass concentration of TMA in the chamber was measured by an isokinetic dust sampler (A type. Shibata. Ja pan). The TMA flake was grounded to respirable-size particles by a grinder and the particle sizes were measured in the exposure chamber by an Anderson sampler. The average aerodynamic di ameter of the TMA particles was 5 um. Bronchoalveolar Lavage Fluid (BALF) Guinea pigs anesthetized with sodium pentobarbital were ex sanguinated by cardiac puncture. The lungs were removed from these animals and 5 ml of PBS was injected slowly into the left lung through the trachea. The PBS-filled lung was gently mas saged for 10 s and the PBS was then carefully withdrawn. The re covered lavage fluid was filtered through a sterile gauze, trans ferred to a test tube and centrifuged at 250 g for 10 min. The su pernatant was kept at -20°C for subsequent determination of the antibody content. Histologic Examination Guinea pigs were sacrificed by an intraperitoneal injection of pentobarbital. The lungs were removed and the right lungs were infused gently with 20% buffered formalin at a pressure of 30 cm H,0, fixed in the same solution and then embedded in paraffin. The fixed specimens were sliced into 4 pm thickness and stained with hematoxylin and eosin for microscopic observations. Measurement of Vascular Permeability Fifty milligrams of Evans blue dye per kilogram body weight of guinea pigs was injected into the ear veins of the animals. At 60 min after the injection, the animals were sacrificed by injecting pentobarbital. To wash out the dye from the intravascular spaces, the lungs and heart of each animal were removed and 40 ml of PBS was perfused through the pulmonary artery at a perfusion pressure of 100 cm H:0 . A peripheral section of each of the three lobes was resected, weighed and placed in 2 ml of 100% formamidc. and kept for 24 It in a 60"C water bath. The concentration of Evans blue dye in the formamide was determined by measuring
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
TMA-lnduced Lung Injury in Guinea Pigs
122
Tao/Sugiura/Nakamura/Kido/Tanaka/Kuroiwa
Results
Antisera against
Animals, n
Antibody tiler using SRBC conjugated with TM14 OVA
OVA
T M , BSA
4
1>
1>
T M w BSA
10
5 .8 ± 2 .1
1>
T M jg B S A
10
1 0 .4 ± 1.5
1>
The animals were intramuscularly injected 4 times with TM BSA conjugates emulsified with CFA. Anti-TM antibody titers of the animal sera were determined by passive hemagglutination us ing TM,., OVA-coated SRBC. Values represented are mean ±SD.
Table 2. Hemorrhagic pneumonitis of actively immunized guinea pigs with TM BSA after TMA inhalation and antibody ti ters in their sera and BALFs Animal
1 2 3 4 5 6* 7* 8 la 2a 3a
Hemorrhagic pneumonitis
+ + + + + + + + -
Antibody titer (log,) in scrum
BALF
9 7 6 10 8 10 8 7 1> 1> 1>
1> 1> I> l> 1> 1> 1> 1> 1> 1> 1>
The animals (n = 8) were intramuscularly injected twice with TMj, BSA emulsified with CFA. Two weeks later, the animals were exposed to TMA inhalation for 6 h. Pulmonary hemorrhages of the lungs were seen as positive (+) or negative (-) and anti-TM antibody titers were determined by hemagglutination using TM m OVA-coated SRBC. Asterisks (*) show animals in shock. Control animals without immunization are shown as fa, 2a and 3a.
the optical density at 620 nm [16. 17). The vascular permeability was assessed by the amount of Evans blue dye (ng) per milligram of tissue wet weight. Statistical Analysis Student’s t test was applied for statistical analysis to compare the difference in vascular permeabilities between antibody-in duced animals and control animals. The p value < 0.05 was consid ered significant.
Anti-TM Antibody Response in Immunized Guinea Pigs
Guinea pigs were immunized with 0.5 mg of TMconjugated BSA in 1 ml of PBS emulsified in CFA by four intramuscular injections at 2-week intervals. As shown in table 1, three TM BSA preparations with dif ferent epitope densities, TM2, TM10 and TM28, were used for the immunization. Two weeks after the last injection, the animals were exsanguinated and the se ra were isolated. The antibody contents in these sera were determined by the passive hemagglutination method using TM OVA coupled with SRBC because there was no cross-reactive antigenicity between BSA and OVA, and the results obtained are also included in table 1. These results showed that the highest anti body titer was found in the serum immunized with TM28 BSA, suggesting that the antibody response was significantly involved in the conjugated epitope densi ties of the synthetic antigens. Since the BSA that was conjugated with more than 36 TM residues became insoluble, all of the immunization experiments were carried out using the conjugates with about 30 TM re sidues per protein molecule. Exposure o f Guinea Pigs Immunized with TM BSA Conjugates to TMA Inhalation
The guinea pigs (n = 8) immunized with TM14 BSA were exposed to TMA inhalation in the chamber at a concentration of 130 mg/m3 for 6 h. After the expo sure, the sera and BALFs from these animals were re moved to determine the antibody contents, and the lungs were then isolated for the macro- and microhistological observations of pulmonary hemorrhages. As shown in table 2 and figure 2a and b, significant amounts of antibody specific for TM epitopes were detected in all of the antisera, but not in their BALFs, and the pulmonary hemorrhages could be observed in the lungs of ail of these animals. Two of the 8 animals seemed to be in shock after exposure, and died due to pulmonary hemorrhages accompanied by nasal bleed ing. In these cases, severe external hemorrhages cov ering each lung lobe were histologically found and a large number of erythrocytes and inflammatory cells and also significant amounts of edema fluids were ob served in the interstitium and alveoli. Even in the cases of animals with a slight clinical status, the anti bodies in their sera were detectable, and then pulmo nary hemorrhages were seen. As a control, normal
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Table 1. Production of anti-TM antibody in guinea pigs
TMA-Induccd Lung Injury in Guinea Pigs
123
guinea pigs without immunization neither have anti bodies in their sera and BALFs nor show pulmonary hemorrhages after TMA inhalation. Exposure o f Guinea Pigs Passively Sensitized with Anti-TM Antibody to TMA Inhalation
Normal guinea pigs (n = 6) were intravenously in jected with 1 ml of TM antibody containing the titer of 218. Fifteen hours after the injection, the animals were exposed to TMA inhalation for 3 h at a concentration of 130 mg/m3. The animals were exsanguinated, and the antibody concentrations in the sera and in the BALFs along with the lung injury were examined. As shown in table 3 and figure 3a and b, in all of the ani mals, the anti-TM antibodies were detectable in the sera and the BALFs, and pulmonary hemorrhages were observed in all but I case. In a parallel study, the guinea pigs passively sensitized with anti-TM antibody via the peritoneal cavity also developed pulmonary hemorrhages. As a control, normal guinea pigs (n = 5) were intravenously injected with normal guinea pig serum and exposed to TMA inhalation under the same conditions as above. No pulmonary hemorrhag es were found in these animals, but some neutrophil infiltrations were found in 3 cases.
To examine whether immunoglobulin classes of anti-TM antibodies are necessary for the induction of lung injury, normal guinea pigs were intraveneously injected with the isolated antibody, IgGl or IgG2. As shown in table 4, the characterization of the immu nochemical properties of these antibodies was per formed by the measurements of hemagglutination, complement fixation, and PCA titers. The major parts of the TM antibodies were found in P-1 and P-2 frac tions corresponding to IgG2 and IgGl subclasses, re spectively. Judging from the results of the comple ment fixing and PCA liters, cross-contaminations of each subclass antibody in these fractions were less than 1%, respectively. In this investigation, the change of vascular permeability by lung injury was quantita tively determined by measuring the Evans blue dye that were exuded from the blood vessel into the extravascular tissues. The antibody preparations were in jected intravenously into normal guinea pigs. At 15 h after the injection, the animals were placed in the chamber and exposed to TMA inhalation at a concen tration of ISC) mg/m3 for 3 h. Two hours after the in-
Fig. 2. a Macrophotograph of a lung from an actively immu nized guinea pig with TM,., BSA after single TMA inhalation. The lung exhibits severe external hemorrhages covering each lobe, b Microphotograph of the lung. A large number of intraalveolar erythrocytes and other inflammatory cells, and edema fluids are seen. HE. x40. Table 3. Hemorrhagic pneumonitis of passively sensitized guin ea pigs with anti-TM antisera after TMA inhalation and antibody titers in their sera and BALFs Animal
1 2 3 4 5 ft
Hemorrhagic pneumonitis _ + + + + +
Antibody titer (log,) in serum
BALF
11 12 10 11 II 11
4 7 3 5 7 5
One milliliter of antiserum containing anti-TM antibody titer of 2"1was injected into their ear veins. Fifteen hours after the in jection. the animals were exposed to TMA inhalation for 3 h. The pulmonary hemorrhages and the antibody titers were expressed as described in table 2.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Exposure o f Guinea Pigs Passively Sensitized with Isolated IgGl or IgG2 Antibody to TMA Inhalation
Tao/Sugiura/Nakarnura/Kido/Tanaka/K.uroiwa
124
o
o
v
Q-
halation, the Evans blue dye solution was intrave nously injected, and the dye content in the peripheral lung tissues was determined by optical density mea surement, and expressed as nanograms of dye per mil ligram of the wet weight of lung, as shown in figure 4. In parallel experiments, the effects of heat treatment on the antibody preparations were examined and the results were also included in figure 4. Increased pul monary permeability after TMA inhalation was ob served in both of the actively and passively sensitized animal groups, but the differences between IgGl and IgG2 antibodies and between antiserum or IgGl with and without heat treatment were statistically insignif icant. These results indicated that the vascular perme ability was significantly increased by TMA inhalation to guinea pigs sensitized with either of the IgG sub classes, IgGl or IgG2 antibodies. Also no significant
alteration of the sensitizing activity of anti-TM anti bodies by heat treatment was demonstrated by the isolated antibody preparations as well as by the whole anti-TM antiserum.
Discussion
Trimcllitic, phthalic, tetrachlorophthalic and ma leic anhydrides are biologically active compounds that are known to cause immunologically mediated respi ratory syndromes in humans [18,19]. These chemicals have been widely used in industrial factories and have been reported to be the cause of asthma in those workers who have been exposed to them at their workplace [1]. In an experimental model, the inhala tion of TMA by SPF SD rats induced typical lung le-
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Fig. 3. a Macrophotograph of a lung from a passively sensitized guinea pig with anti-TM antisera after single TMA inhalation. The lung exhibits moderate external hemorrhages covering each lobe, b Microphotograph of the lung. Intraalveolar erythrocytes and other inflammatory cells are seen. HE. x 100.
Fig. 4. Comparison of vascular permeabilities by lung injury in duced in sensitized guinea pigs with TMA inhalation. The animals were passively sensitized with intact antibodies (□) and with heattreated antibodies (0) 15 h before exposure. The data obtained by actively immunized guinea pigs with TM BSA are also included (□). The Evans blue dye contents in the lungs of animals after TMA inhalation are expressed as ng/mg of lung (wet weight). A = PBS (n = 9); B = anti-TM antiserum (n = 9); C = heat-treated antiTM antiserum (n = 5); D = anti-TM antibody IgGl (n = 7); E = heat-treated anti-TM antibody IgGl (n = 4); F = anti-TM antibody IgG2 (n = 6); G = actively immunized guinea pigs (n = 4). Each value represents the mean +SD and p values are also calculated.
TMA-Induced Lung Injury in Guinea Pigs
125
Table 4. Immunochemical properties of isolated of anti-TM antibody Material
Hemagglutination
Complement fixation
PCA
Protein recovery (%)
IgG subclass
Whole serum P-1 P-2 P-3 P-4
216
2 '2
_
_
26
2»i>
2s 2U> 2 >
IgG2 IgGl
2M
2 >
2" T
22.4 26.4 20.5 17.5
2>o
2>
2 13
_ -
sions associated with TM-spccific antibody and the profiles resembled human syndromes [5]. TMA is a haptenic substance incapable of stimulating the im mune system by itself without carrier substances. However, when animals can be immunologically sen sitized through airways by TMA inhalation, the in haled TMA has to be converted into a certain immu nogenic substance capable of stimulating their im mune systems. It seems likely that protein molecules in lavage fluids and immunocytes are involved in this conversion process. At present, however, the mecha nism of sensitization by TMA inhalation is still un clear. The present study was undertaken to investigate the immunological mechanism of TMA-mediated lung injury in conventional guinea pigs, because they were highly susceptible to allergic responses, passive cutaneous and systemic anaphylaxis. It was demon strated that the injection with TM-carrier conjugates in guinea pigs resulted in the production of specific antibodies against the TM epitopes conjugated with BSA in the sera of these animals. A positive correla tion was seen between the number of TM epitopes conjugated to BSA and the antibody titer of the anti sera. This finding is compatible with the report of Ru bin [20] and suggested that the production of anti-TM antibody is associated with epitope density on the car rier protein. The guinea pigs actively immunized with TM BSA developed severe pulmonary hemorrhages after a single exposure to TMA inhalation. This sug gests that inhaled TMA acts directly or indirectly as an antigenic substance in lung injury. Furthermore, guinea pigs sensitized intraveneously or intraperitoneally with antiserum containing anti-TM antibody al so developed pulmonary hemorrhages after TMA in halation. This result is in agreement with that report
ed by Zeiss et al. [6], who showed that the lung lesions in normal rats could be induced by the passive sensiti zation with humoral antibodies followed by TMA in halation. In this investigation, the inhalation exposure was performed at a TMA concentration ranging from 130 to 180 mg/m-1 because no injury was found in animals’ lungs after repeated inhalation of low concentrations of TMA. This result conflicts with that of SPF SD rats reported by Leach et ai. [5]. In our preliminary exper iments, only 1 of 8 SPF SD rats exposed to TMA in halation developed slight pulmonary hemorrhages. The differences between our experiments and that of Leach et al. may be the size of the TMA particles and the inhalation systems. TMA particles of about 5 urn in size were used in these experiments, whereas Leach et al. used particles of about 1 pm. This suggests that the aerodynamic property of TMA particles plays an important role in the induction of lung injury. Anoth er explanation for the discrepancy of the results may be a difference in the animal species studied. Guinea pigs may be difficult to sensitize by TMA inhalation alone. It is well known that guinea pig antisera against protein antigens contain considerable amounts of an tibodies in both of the immunoglobulin subclasses, IgGl and IgG2, and that these antibodies are distin guishable by their contrasted immunochemical prop erties [21-23]. The antibodies associated with IgGl subclass have a cytotropic affinity to homologous ani mal tissues, but lack a complement fixing activity, whereas the antibodies associated with IgG2 subclass have a complement fixing activity, but lack a homocytotropic affinity. Henson and Cochrane [24] reported that homocytotropic antibodies in recipient rabbit se rum seemed to be necessary for the successful media
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
The peak fractions of anti-TM antibody isolated by DEAE-cellulose column chromatography were denoted as P-1, P-2, P-3 and P-4. The measurements of hemagglutination, complement fixation and PCA titers of these fractions were performed (see Materials and Methods). The protein recovery and the corresponding subclass are also included in this table.
Tao/Sugiura/Nakamura/Kido/Tanaka/Kuroiwa
126
reaction [31]. These findings lead us to suggest that the IgGl antibody binds to mediator cells such as mast cells, which release chemical mediators induced by TMA inhalation. In the case of the IgG2 antibody, it seems likely that a small portion of intravencously in jected antibody is secreted into the pulmonary alveoli via the lung capillary, where the antibody reacts with its antigen which is composed of the autologous BALF protein and inhaled TMA. The resulting im mune complex may deposit in different anatomic compartments of the lung, i.e. intravascular, intraalveolar and interstitial. At these points, the classical pathway of complement components is activated, re sulting in the generation of chemotactic peptides such as C5a that attract neutrophils from the circulation. The neutrophils then phagocytize immune complexes and release oxygen-derived metabolites, proteases and so forth [32], In conclusion, TMA inhalation induced pulmonary injury in guinea pigs actively immunized with TM BSA conjugates and passively sensitized with homolo gous cytotropic or noncytotropic antibody, and the results indicated that the complement-independent acute type hypersensitivity and the complement-de pendent immune complex type reactions were in volved in this pathogenesis. Guinea pigs appear to be a useful animal model for TMA-induced lung injury. Acknowledgements This work was supported in part by a Grant-in-Aid for Scien tific Research from the Ministry of Education. Japan (No. 02670352).
References 1 Zeiss CR. Patterson R. Pruzansky JJ. Miller MM. Rosenberg M. Levitz D: Trimellitic anhydride-induced airway syndromes; clinical and immunologic studies. J Allergy Clin Immunol 1977;60:96-103. 2 Petterson R. Addington W. Banner AS, Byron GE. Franco M, Herbert FA. Nicotra MB. Pruzansky JJ. Rivera M. Roberts M. Yawn D, Zeiss CR: Antihapten antibodies in workers exposed to trimellitic anhydride fume; a potential immunopathogcnetic mechanism for trimellitic anhydride pulmonary disease-ane mia syndrome. Am Rev Rcspir Dis 1979;120:1259-1267. 3 Herbert FA. Orford R: Pulmonary hermorrhages and edema due to inhalation of resin containing trimellitic anhydride. Chest 1979;76:546-552. 4 Sale SR. Roach DE. Zeiss CR. Patterson R: Clinical and im munological correlations in trimellitic anhydride airway syn dromes. J Allergy Clin Immunol 1981;68:188-193.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
tion of glomerular and arterial injury. In addition, Roska et al. [25] suggested that pigeon serum-induced pulmonary inflammation in guinea pigs was a mani festation of a complement-dependent, humoral im mune mechanism of pathogenesis and thus was con sistent with an immune complex disease. Also. Richcrson [26] reported that scattered alveolitis induced by aerosolized antigens, depending on the type of de layed hypersensitivity without demonstrable antibody, developed in guinea pig lungs. Therefore, the follow ing experiments were attempted to determine which subclass antibody, IgGl, IgG2 or both were involved in the induction of pulmonary injury by TMA inhala tion. Each of the subclass antibodies was isolated from pooled hyperimmune antisera against TM BSA using gel filtration and ion exchange column chroma tographies, and the purity of the isolated preparation was examined by immunochemical procedures, dou ble diffusion and immunoclectrophoresis, comple ment fixation and PCA reactions. The results demon strated that the IgGl fraction contained less than 1% IgG2 and IgG2 fraction contained less than 1% IgGl. To evaluate the pulmonary injury more quantitatively, the enhanced permeability of Evans blue dye in the extravascular tissues after TMA inhalation were mea sured in the lungs of guinea pigs passively sensitized with isolated antibodies. Furthermore, the effect of heat treatment on the sensitizing activity of these an tibodies was also examined. With respect to vascular permeability, not only the IgGl, but also the IgG2 fraction was comparable with the antisera, suggesting that homocytotropic as well as nonhomocytotropic antibodies are involved in the induction of lung injury by TMA inhalation. Heat treatment of the antiserum and the IgGl fraction did not influence vascular per meability, which suggested that heat-labile IgE anti body was not involved in this process. It has been re ported that many chemical mediators cause a signif icant acute exudation of protein-rich plasma [27-29], A possible interaction between these mediators, mediator cells and inflammatory cells in acute and chronic asthma was demonstrated by Kay [30], who showed that the initial burst of lung mast cell-derived mediators leads to recruitment and activation of sec ondary inflammation cells, including activated neu trophils and monocytes. Recent evidence suggests that late reactions may be associated with the reacti vation of mast cells since a second elevation in serum neutrophil chemotactic activity and plasma histamine concentration was observed during late-phase asthma
127
TMA-Induccd L.ung Injury in Guinea Pigs
20 Rubin B: Studies on the induction of antibody synthesis
21
22
23
24
25
26 27
28
29
30 31
32
against sulfanilic acid in rabbits. I. Effect of the number of hapten molecules introduced in homologous protein on anti body synthesis against the hapten and the new antigenic deter minants. Eur J Immunol 1972:2:5-11. Ovary Z, Benacerraf B. Bloch KJ: Properties of guinea pig 7S antibodies. Identification of antibodies involved in passive cu taneous and systemic anaphylaxis. J Exp Med 1963:117:951964. Bloch KJ. Lourilsky FM. Ovary Z, Benacerraf B: Properties guinea pig 7S antibodies. Identification of antibodies involved in complement fixation and hemolysis. J Exp Med 1963:117: 965-981. Oliberia B. Oslcp AG. Siraganian RP, Sandberg AL: The bi ologic activities of guinea pig antibodies. I. Separation of y,and y2-immunoglobulins and their participation in allergic re actions of the immediate type. J Immunol 1970;104:320-328. Henson PM. Cochrane CG: Acute immune complex disease in rabbits. The role of complement and of a leukocyte-dependent release of vasoactive amines from platelets. J Exp Med 1971; 133:554-571. Roska AK. Moore VL. Abramoff P: Immune complex disease in guinea pig lung: Elicitation with pigeon serum. Am Rev Respir Dis 1979;120:129-136. Richerson HB: Acute experimental hypersensitivity pneumo nitis in the guinea pig. J Lab Clin Med 1972;79:745-757. Persson CGA. Erjefalt I: Inflammatory leakage of macromole cules from the vascular compartment into the tracheal lumen. Acta Physiol Scand 1986;126:615-616. Persson CGA. Erjefalt I. Karlsson JA: Effect of tachykinin an tagonists on the tracheobronchial microcirculation: in Hakanson R. Sundler F (eds): Tachykinin Antagonists. Amsterdam, Elsevier, 1985, pp 171-179. Persson CGA, Erjefalt I, Andersson P: Leakage of macromole cules from guinea pig tracheobronchial microcirculation. Ef fects of allergen, leukotricnes, tachykinins, and antiasthma drugs. Acta Physiol Scand 1986:127:95-105. Kay AB: Inflammatory cells in acute and chronic asthma. Am Rev Respir Dis 1987; 135:S36— S66. Durham SR, Lee TH. Cromwell O: Immunologic studies in al lergen-induced late-phase asthmatic reactions. J Allergy Clin Immunol 1984;74:49-60. Daniclc RP, Henson PM. Fantone JC, Ward PA, Preisin RB: Immune complex injury of the lung. Am Rev Respir Dis 1981; 124:738-755.
Received: April 22, 1991 Accepted after revision: May 23. 1991 Correspondence to: Dr Yoshiaki Tao Department of Immunology School of Medicine University of Occupational and Environmental Health Kitakyushu 807 (Japan) Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
5 Leach CL, Hatoum NS, Ratajezak HV, Zeiss CR. Roger JC. Garvin PJ: The pathologic and immunologic response to in haled trimellitic anhydride (TMA) in rats. Toxicol Appl Phar macol 1987;87:67-80. 6 Zeiss CR. Leach CL. Smith 1J, Levitz D, Hatoum NS, Garvin PL. Patterson R: A serial immunologic and histopathologic study of lung injury induced by trimellitic anhydride. Am Rev Respir Dis 1988;137:191-196. 7 Turner ES. Pruzansky JJ. Patterson R. Zeiss CR. Roberts M: Detection of antibodies in human serum using trimellitylerythrocytes: direct and indirect hacmagglutination and hae molysis. Clin Exp Immunol 1980;39:470-476. 8 Zeiss CR. Levity D. Chacon R. Wolkonsky P. Patterson R. Pru zansky JJ: Quantitation and new antigen determinant (NAD) specificity of antibodies induced by inhalation of trimellitic an hydride in man. Int Arch Allergy Appl Immunol 1980:61:380388. 9 Patterson R, Roberts M. Zeiss CR. Pruzansky JJ:Human anti bodies against trimcllityl proteins: Comparison of specificities of IgG, IgA and IgE classes. Int Arch Allergy Appl Immunol 1981:66:332-340. 10 Chandler MJ. Zeiss CR. Leach CL. Hatoum NS. Levitz D. Gravin PJ. Patterson R: Levels of specific antibody in bronchoalvcolar lavages (BAL) fluid in an animal model of trimel litic anhydride-induced by injury. .1 Allergy Clin Immunol 1987;80:223-229. 11 Nakamura H, Sugiura T: Inhibition of anaphylactic sensitiza tion activity of guinea pig antibody by binding with protein A from Staphylococcus aureus. J UOEH 1982;4:241-253. 12 Liu SH. Koo PH. Cebra JJ: Effect of carrier priming on the distribution of anti-hapten antibodies between IgG 1 and IgG2 isotypes in the hyperimmune guinea pig. J Immunol 1974; 113:677-687. 13 Avrameas S. Taudou B. Chuilon S: Glutaraldehyde, cyanuric chloride and tetraazotized o-daianisidine as coupling reagents in the passive hemagglutination test. Immunochemistry 1969;6: 67-76. 14 Ovary F: Local anaphylaxis; in Williams CA. Chase MW (eds): Methods in Immunology and Immunochemistry. New York. Academic Press, 1976, vol V, p 16. 15 Tanaka 1. Matsuno K. Kodama Y. Akiyama T: Pulmonary de position of a fly ash aerosol by inhalation. J UOEH 1983; 5:423-431. 16 Udaka K. Takeuchi Y. Movat HZ: Simple method for quantita tion of enhanced vascular permeability. Proc Soc Exp Biol Med 1970:133:1384-1387. 17 Thompson JA. Scypinski LA, Gordon T. Sheppord D: Tachykinins mediated the acute increase in airway responsiveness caused by toluene diisocyanate in guinea pigs. Am Rev Respir Dis 1987;136:43-49. 18 Howe E, Venables KM. Topping MD. Dally MB. Hawkins R. Law JS: Newman Taylor AJ: Tetrachlorophthalic anhydride asthma: evidence for specific IgE antibody. J Allergy Clin Im munol 1983:71:5-11. 19 Newman Taylor AJ. Venables KM. Durham SR. Grancek BJ. Topping MD: Acid anhydrides and asthma. Int Arch Allergy Appl Immunol 1987;82:435-439.
©1991 S.Karger AG. Basel 0020-5915/91/0962-0119 $ 2.75/0
Experimental Lung Injury Induced by Trimellitic Anhydride Inhalation on Guinea Pigs Yoshiaki Tao3, Tsutomu Sugiura3, Hiroshi Nakamura3, Masamitsu Kidoh, Isamu Tanakac, Akio Kuroiwad ■'Department of Immunology. School of Medicine, b Division of Respiratory Disease. University Hospital. c Department of Environmental Health Engineering, Institute of Industrial Ecological Sciences,d Second Department of Internal Medicine. University of Occupational and Environmental Health, Kitakyushu. Japan
Introduction
Trimellitic anhydride (TMA) is a low molecular weight, water-insoluble, biologically reactive chemical compound widely used in the production of resins, ad hesives, polymers dyes and printing inks. Inhalation exposure to TMA has been recognized as causing four clinical syndromes in man; of these, asthma-rhinitis, late respiratory systemic syndrome and pulmonary' disease-anemia syndrome are immunologically medi ated. The mechanism of the fourth syndrome is based on an irritant reaction [1-4]. It has been suggested that the levels of antibodies specific for trimellitylated (TM) epitopes in the sera of patients are associated with either the asthma-rhinitis or the late respiratory systemic syndrome, and that different immunoglobu lin classes of antibodies are involved in a variety of re
spiratory syndromes [1. 3]. To further investigate the mechanisms of immunopathogenic response induced by TMA inhalation, an experimental animal model of lung injury caused by TMA inhalation was established by Leach et al. [5] and by Zeiss et al. [6] using specific pathogen-free (SPF) Sprague-Dawley (SD) rats. The results suggested that the inhalation of TMA was a significant stimulus of a systemic immune response that occurred parallel with lung injury. Furthermore, the results of inhibition studies on antibody specificity using various kinds of protein conjugates with TM epitopes suggested that the TMA inhalation resulted in antibody response with specificity for new antigenic determinants that arose from the coupling of TMA with autologous respiratory tract proteins [7-10], The activation mechanism of immune systems by TMA, however, is not yet fully understood.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Abstract. Trimellitic anhydride (TMA) causes lung injury by inhalation exposure in humans. In order to in vestigate more precisely the mechanism of lung injury by TMA, an experimental animal model of TMAinduced lung injury was established. Guinea pigs were intramuscularly injected with trimellitylated bovine se rum albumin (TM BSA) with complete Freund’s adjuvant (CFA). Appreciable amounts of antibodies against TM epitopes were detected in the sera of these animals. Guinea pigs that were passively sensitized with antiTM antisera as well as the actively immunized animals developed hemorrhagic pneumonitis after TMA in halation. It is well recognized that hyperimmune antisera of guinea pigs contain two subclasses of IgG antibod ies, IgGl and IgG2, which are known as homocytotropic and heterocytotropic antibodies, respectively. To de termine the role of these antibodies in the airway injury with alveolar hemorrhages, they were isolated by gel filtration and by ion exchange column chromatography, and the guinea pigs that were sensitized with each of these antibodies were exposed to TMA inhalation. The extent of lung injury in these animals was quantitatively determined by the measurements of lung extravasation of Evans blue dye injected intravenously after TMA in halation. Significant increases in the extravasation of dye were observed in both animal groups sensitized with IgGl and IgG2 antibodies. In addition, results obtained with heat-treated antisera and IgGl antibody did not significantly differ from those obtained with untreated samples. These results suggest that the lung injury re sulting from TMA inhalation exposure can be induced with humoral antibodies, not only IgGl but also IgG2, and that at least two types of allergic reactions are involved in the pathogenesis of lung injury.
Tao/Sugiura,Nakamura'Kido/Tanaka/Kuroiwa
120
Fig. 1. a Gel filtration chromatography of anti-TM antibody IgG fraction on a column of Sephacryl S-300 (3x51 cm) equilibrated with PBS (pH 7.0). The elution was performed with the same buffer solution and the optical absorbance at 280 nm wavelength and the antibody content of elute (3.1 ml) was measured. The antibody activity was detectable in the second peak corresponding to IgG with a molecular weight of about 160,000. These fractions were pooled, concentrated and offered to further purification. Blue dextran. OVA and human IgG were used as molecular weight markers, b DEAE-cellulose column chromatography of the IgG peak by gel filtration chromatography. The elution was performed stepwise by changing pH and ionic strength of the buffer solution as induced in the figure. Each of the peak components was pooled and denoted as P-1. P-2, P-3 or P-4.
Materials and Methods Experimental Design Guinea pigs were directly immunized with two or four intra muscular injections of TM conjugates with CFA. As antiserum do nors, guinea pigs were immunized by injections of TM conjugates with complete Freund’s adjuvant (CFA) into their footpads fol lowed by three intradermal injections with the same antigen at 3week intervals. In this study, four experimental approaches were undertaken: (1) to prepare TM protein antigens that would pro duce sufficient amounts of the specific antibodies against TM epi topes in guinea pigs; (2) to determine the immunization schedules of guinea pigs for developing lung injury after TMA inhalation; (3) to sensitize guinea pigs passively with anti-TM antisera, and (4) to sensitize guinea pigs with a passive transfer of isolated IgGl and/or IgG2 antibodies. In addition, the immunological character
ization of the isolated antibodies were carried out using passive hemagglutination, complement fixation and passive cutaneous anaphylaxis (PCA). Animals Random-bred female Hartley guinea pigs (200-500 g) were maintained in the animal research facilities at the university and were carefully examined for health defects prior to use in investi gations. Reagents TMA was purchased from Nacaiai Tesque. Kyoto, Japan, in the form of small Oakes. Bovine serum albumin (crystalline BSA: Armour Pharmaceutical Co.. Kankance. III. USA) and ovalbumin (crystalline OVA; Sigma Chemical Co.. St. Louis. Mo.. USA) were used for the preparation of TM protein conjugates. Sheep eryth rocytes (SRBC) were supplied by Nippon Bio-Supply Center, To kyo. Japan. Anti-SRBC antibody was prepared by using rabbit an tisera against SRBC. Anti-OVA antibody was prepared by using guinea pig antisera against OVA. Normal guinea pig IgG was pre pared by ammonium sulfate precipitation at 40% saturation. Goat antibodies specific for guinea pig IgGl and IgG2 were prepared in our laboratory [II], TM Protein Conjugates The conjugation of TM epitopes to protein was carried out us ing BSA or OVA according to the method deset ibed by Zeiss et al. [1] with slight modifications. Briefly, a BSA or OVA solution con taining a concentration of 5 mg protein/ml of 7% sodium bicarbo nate solution was prepared. To 20 ml of the protein solution were added various amounts of TMA dissolved in 1 ml of dioxane and the resulting mixture was stirred for 30 min at room temperature. The sample solution was dialyzed against 0.1 M NaHCO, over
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
In the present report, it is demonstrated that lung injury can be induced by TMA inhalation in guinea pigs that are not only actively immunized with TM bo vine serum albumin (BSA) conjugates but also pas sively sensitized with a homologous antibody, IgGl or IgG2, and that different types of allergic reactions may be involved in the process of lung injury. In addi tion, the involvement of another class of homocytotropic antibody, IgE, is probably excluded by the re sults of heat treatment of antibodies. This animal model of lung injury induced by TMA inhalation pro vides a useful system for investigating the immunopathological mechanisms of the syndrome.
night, then the outer solution changed to phosphate-buffered sa line (PBS, pH 7.0). The number of TM epitope conjugated with the protein molecule was assessed by the measurements of optical absorbances at 240 and 280 nm (Hitachi model U-1100 spectropho tometer; Tokyo, Japan) using molar extinction coefficients 11.400/ 1,700 (A240 nm/AMI nm ) for NaTM. and 74.800/43,500 and 62.100/ 33,300 for BSA and OVA. respectively. TM protein conjugates with various amounts of the epitopes were obtained and the num ber of conjugated TM residues were expressed in small figures such as TM„, BSA and TMI4 OVA. Hyperimmune Serum Guinea pigs had I mg TM36 BSA emulsified in CFA injected into their footpads on days 0 and 14 followed by intradcrmal in jections with the same antigen on days 35, 56 and 77. The animals were exsanguinated and the antisera were fractionated by ammo nium sulfate precipitations at 40% saturation. The precipitate, IgG fraction, was centrifuged and dissolved in PBS and extensively dialyzed against PBS with several changes in the outer solution. IgG fraction was further purified by gel filtration chromatography on a column of Scphacryl S-300 in PBS, and the elution profile was monitored by optical density measurements at 280 nm. as shown in figure la. The fractions containing IgG were pooled and concentrated by evaporation, and then thoroughly dialyzed against a 10 mM phosphate buffer (pH 8.0) for further purification by DEAE-cellulose (Whatman DE-52) column chromatography according to the method described by Liu et at. [12] with slight modifications. The elution was performed by changing pH and the ionic strength of the buffer solution with NaCI. One of the typical elution profiles is shown in figure lb. The immunochemical prop erty of each fraction was examined by immunoelectrophoresis and the double diffusion test using goat antisera specific for guinea pig IgG! and/or IgG2. The fractions containing single IgG subclass were then pooled and concentrated. Hemagglutination Test The sensitization of SRBC with TM BSA or TM OVA was per formed according to the method described by Avrameas et al. [13]. The anti-TM antibodies in antisera were measured by passive hemagglutination using sensitized SRBC. The sample solutions were incubated at 56 °C for 30 min to inactivate complement com ponents and absorbed with SRBC treated with 1% glutaraldehyde and glycine solution. They were then 2-fold serially diluted with gelatinized veronal buffer (pH 7.4. 4.9 mM veronal, 0.5 mM MgS04, 0.75 mM CaCL, 146 mM NaCI and 0.2% gelatin) in micro titer plates. To the sample solutions (25 pi) was added a 1% sus pension of TM|4 OVA SRBC (25 pi), mixed and incubated at room temperature for 90 min. The antibody titer was expressed as the reciprocal of the highest dilution of samples giving positive hemagglutination. A normal serum and an anti-OVA serum were used as control. Complement Fixation Test Complement fixation tests were carried out using the same ve ronal buffer solution as described above. A normal guinea pig se rum was used as a source of complement (130 CHS0). To the dilut ed sample solutions containing anti-TM antibodies in microtiter plates with round-bottom wells were added 25 ul of the serial 2fold dilutions of 4 mg/ml TM,4 OVA with diluted normal guinea pig serum containing 2.5 CHS0. After incubation at 37 °C for 2 h.
121
25 pi of 1% suspension of SRBC sensitized with rabbit anti-SRBC antibody (hemolysin) was added to each well. One hour later, the titer was determined from the highest dilution of specimens with complete hemolysis. Tussive Cutaneous Anaphylaxis Two aliquots of serial 2-fold dilutions of the samples with PBS were prepared. One of them, without any pretreatment, was used for the induction of PCA reaction, and the other was incubated at 56 °C for 2 h prior to its use in the PCA reaction. A 0.1-ml portion of the sample was injected intradermally into normal guinea pigs, and 6 h later. 1 ml of PBS containing 1 mg of TM14 OVA and 10 mg of Evans blue dye (Sigma Chemical Co.) were injected intrave nously into the guinea pigs. At 30 min after the injection, animals were sacrificed and the diameter of the blue spot developed at the antibody injection site was measured on the inner surface of the flayed skin. The PCA titer of the specimen was determined as the highest dilution having a blue spot size of 10 x 10 mm [14]. Exposure System of TMA Inhalation The exposure of TMA inhalation was performed by the same system described by Tanaka et al. [15] using TMA instead of fly ash. The airflow in the chamber was downward and laminar with a volume of 0.1 m \ The mass concentration of TMA in the chamber was measured by an isokinetic dust sampler (A type. Shibata. Ja pan). The TMA flake was grounded to respirable-size particles by a grinder and the particle sizes were measured in the exposure chamber by an Anderson sampler. The average aerodynamic di ameter of the TMA particles was 5 um. Bronchoalveolar Lavage Fluid (BALF) Guinea pigs anesthetized with sodium pentobarbital were ex sanguinated by cardiac puncture. The lungs were removed from these animals and 5 ml of PBS was injected slowly into the left lung through the trachea. The PBS-filled lung was gently mas saged for 10 s and the PBS was then carefully withdrawn. The re covered lavage fluid was filtered through a sterile gauze, trans ferred to a test tube and centrifuged at 250 g for 10 min. The su pernatant was kept at -20°C for subsequent determination of the antibody content. Histologic Examination Guinea pigs were sacrificed by an intraperitoneal injection of pentobarbital. The lungs were removed and the right lungs were infused gently with 20% buffered formalin at a pressure of 30 cm H,0, fixed in the same solution and then embedded in paraffin. The fixed specimens were sliced into 4 pm thickness and stained with hematoxylin and eosin for microscopic observations. Measurement of Vascular Permeability Fifty milligrams of Evans blue dye per kilogram body weight of guinea pigs was injected into the ear veins of the animals. At 60 min after the injection, the animals were sacrificed by injecting pentobarbital. To wash out the dye from the intravascular spaces, the lungs and heart of each animal were removed and 40 ml of PBS was perfused through the pulmonary artery at a perfusion pressure of 100 cm H:0 . A peripheral section of each of the three lobes was resected, weighed and placed in 2 ml of 100% formamidc. and kept for 24 It in a 60"C water bath. The concentration of Evans blue dye in the formamide was determined by measuring
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
TMA-lnduced Lung Injury in Guinea Pigs
122
Tao/Sugiura/Nakamura/Kido/Tanaka/Kuroiwa
Results
Antisera against
Animals, n
Antibody tiler using SRBC conjugated with TM14 OVA
OVA
T M , BSA
4
1>
1>
T M w BSA
10
5 .8 ± 2 .1
1>
T M jg B S A
10
1 0 .4 ± 1.5
1>
The animals were intramuscularly injected 4 times with TM BSA conjugates emulsified with CFA. Anti-TM antibody titers of the animal sera were determined by passive hemagglutination us ing TM,., OVA-coated SRBC. Values represented are mean ±SD.
Table 2. Hemorrhagic pneumonitis of actively immunized guinea pigs with TM BSA after TMA inhalation and antibody ti ters in their sera and BALFs Animal
1 2 3 4 5 6* 7* 8 la 2a 3a
Hemorrhagic pneumonitis
+ + + + + + + + -
Antibody titer (log,) in scrum
BALF
9 7 6 10 8 10 8 7 1> 1> 1>
1> 1> I> l> 1> 1> 1> 1> 1> 1> 1>
The animals (n = 8) were intramuscularly injected twice with TMj, BSA emulsified with CFA. Two weeks later, the animals were exposed to TMA inhalation for 6 h. Pulmonary hemorrhages of the lungs were seen as positive (+) or negative (-) and anti-TM antibody titers were determined by hemagglutination using TM m OVA-coated SRBC. Asterisks (*) show animals in shock. Control animals without immunization are shown as fa, 2a and 3a.
the optical density at 620 nm [16. 17). The vascular permeability was assessed by the amount of Evans blue dye (ng) per milligram of tissue wet weight. Statistical Analysis Student’s t test was applied for statistical analysis to compare the difference in vascular permeabilities between antibody-in duced animals and control animals. The p value < 0.05 was consid ered significant.
Anti-TM Antibody Response in Immunized Guinea Pigs
Guinea pigs were immunized with 0.5 mg of TMconjugated BSA in 1 ml of PBS emulsified in CFA by four intramuscular injections at 2-week intervals. As shown in table 1, three TM BSA preparations with dif ferent epitope densities, TM2, TM10 and TM28, were used for the immunization. Two weeks after the last injection, the animals were exsanguinated and the se ra were isolated. The antibody contents in these sera were determined by the passive hemagglutination method using TM OVA coupled with SRBC because there was no cross-reactive antigenicity between BSA and OVA, and the results obtained are also included in table 1. These results showed that the highest anti body titer was found in the serum immunized with TM28 BSA, suggesting that the antibody response was significantly involved in the conjugated epitope densi ties of the synthetic antigens. Since the BSA that was conjugated with more than 36 TM residues became insoluble, all of the immunization experiments were carried out using the conjugates with about 30 TM re sidues per protein molecule. Exposure o f Guinea Pigs Immunized with TM BSA Conjugates to TMA Inhalation
The guinea pigs (n = 8) immunized with TM14 BSA were exposed to TMA inhalation in the chamber at a concentration of 130 mg/m3 for 6 h. After the expo sure, the sera and BALFs from these animals were re moved to determine the antibody contents, and the lungs were then isolated for the macro- and microhistological observations of pulmonary hemorrhages. As shown in table 2 and figure 2a and b, significant amounts of antibody specific for TM epitopes were detected in all of the antisera, but not in their BALFs, and the pulmonary hemorrhages could be observed in the lungs of ail of these animals. Two of the 8 animals seemed to be in shock after exposure, and died due to pulmonary hemorrhages accompanied by nasal bleed ing. In these cases, severe external hemorrhages cov ering each lung lobe were histologically found and a large number of erythrocytes and inflammatory cells and also significant amounts of edema fluids were ob served in the interstitium and alveoli. Even in the cases of animals with a slight clinical status, the anti bodies in their sera were detectable, and then pulmo nary hemorrhages were seen. As a control, normal
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Table 1. Production of anti-TM antibody in guinea pigs
TMA-Induccd Lung Injury in Guinea Pigs
123
guinea pigs without immunization neither have anti bodies in their sera and BALFs nor show pulmonary hemorrhages after TMA inhalation. Exposure o f Guinea Pigs Passively Sensitized with Anti-TM Antibody to TMA Inhalation
Normal guinea pigs (n = 6) were intravenously in jected with 1 ml of TM antibody containing the titer of 218. Fifteen hours after the injection, the animals were exposed to TMA inhalation for 3 h at a concentration of 130 mg/m3. The animals were exsanguinated, and the antibody concentrations in the sera and in the BALFs along with the lung injury were examined. As shown in table 3 and figure 3a and b, in all of the ani mals, the anti-TM antibodies were detectable in the sera and the BALFs, and pulmonary hemorrhages were observed in all but I case. In a parallel study, the guinea pigs passively sensitized with anti-TM antibody via the peritoneal cavity also developed pulmonary hemorrhages. As a control, normal guinea pigs (n = 5) were intravenously injected with normal guinea pig serum and exposed to TMA inhalation under the same conditions as above. No pulmonary hemorrhag es were found in these animals, but some neutrophil infiltrations were found in 3 cases.
To examine whether immunoglobulin classes of anti-TM antibodies are necessary for the induction of lung injury, normal guinea pigs were intraveneously injected with the isolated antibody, IgGl or IgG2. As shown in table 4, the characterization of the immu nochemical properties of these antibodies was per formed by the measurements of hemagglutination, complement fixation, and PCA titers. The major parts of the TM antibodies were found in P-1 and P-2 frac tions corresponding to IgG2 and IgGl subclasses, re spectively. Judging from the results of the comple ment fixing and PCA liters, cross-contaminations of each subclass antibody in these fractions were less than 1%, respectively. In this investigation, the change of vascular permeability by lung injury was quantita tively determined by measuring the Evans blue dye that were exuded from the blood vessel into the extravascular tissues. The antibody preparations were in jected intravenously into normal guinea pigs. At 15 h after the injection, the animals were placed in the chamber and exposed to TMA inhalation at a concen tration of ISC) mg/m3 for 3 h. Two hours after the in-
Fig. 2. a Macrophotograph of a lung from an actively immu nized guinea pig with TM,., BSA after single TMA inhalation. The lung exhibits severe external hemorrhages covering each lobe, b Microphotograph of the lung. A large number of intraalveolar erythrocytes and other inflammatory cells, and edema fluids are seen. HE. x40. Table 3. Hemorrhagic pneumonitis of passively sensitized guin ea pigs with anti-TM antisera after TMA inhalation and antibody titers in their sera and BALFs Animal
1 2 3 4 5 ft
Hemorrhagic pneumonitis _ + + + + +
Antibody titer (log,) in serum
BALF
11 12 10 11 II 11
4 7 3 5 7 5
One milliliter of antiserum containing anti-TM antibody titer of 2"1was injected into their ear veins. Fifteen hours after the in jection. the animals were exposed to TMA inhalation for 3 h. The pulmonary hemorrhages and the antibody titers were expressed as described in table 2.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Exposure o f Guinea Pigs Passively Sensitized with Isolated IgGl or IgG2 Antibody to TMA Inhalation
Tao/Sugiura/Nakarnura/Kido/Tanaka/K.uroiwa
124
o
o
v
Q-
halation, the Evans blue dye solution was intrave nously injected, and the dye content in the peripheral lung tissues was determined by optical density mea surement, and expressed as nanograms of dye per mil ligram of the wet weight of lung, as shown in figure 4. In parallel experiments, the effects of heat treatment on the antibody preparations were examined and the results were also included in figure 4. Increased pul monary permeability after TMA inhalation was ob served in both of the actively and passively sensitized animal groups, but the differences between IgGl and IgG2 antibodies and between antiserum or IgGl with and without heat treatment were statistically insignif icant. These results indicated that the vascular perme ability was significantly increased by TMA inhalation to guinea pigs sensitized with either of the IgG sub classes, IgGl or IgG2 antibodies. Also no significant
alteration of the sensitizing activity of anti-TM anti bodies by heat treatment was demonstrated by the isolated antibody preparations as well as by the whole anti-TM antiserum.
Discussion
Trimcllitic, phthalic, tetrachlorophthalic and ma leic anhydrides are biologically active compounds that are known to cause immunologically mediated respi ratory syndromes in humans [18,19]. These chemicals have been widely used in industrial factories and have been reported to be the cause of asthma in those workers who have been exposed to them at their workplace [1]. In an experimental model, the inhala tion of TMA by SPF SD rats induced typical lung le-
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
Fig. 3. a Macrophotograph of a lung from a passively sensitized guinea pig with anti-TM antisera after single TMA inhalation. The lung exhibits moderate external hemorrhages covering each lobe, b Microphotograph of the lung. Intraalveolar erythrocytes and other inflammatory cells are seen. HE. x 100.
Fig. 4. Comparison of vascular permeabilities by lung injury in duced in sensitized guinea pigs with TMA inhalation. The animals were passively sensitized with intact antibodies (□) and with heattreated antibodies (0) 15 h before exposure. The data obtained by actively immunized guinea pigs with TM BSA are also included (□). The Evans blue dye contents in the lungs of animals after TMA inhalation are expressed as ng/mg of lung (wet weight). A = PBS (n = 9); B = anti-TM antiserum (n = 9); C = heat-treated antiTM antiserum (n = 5); D = anti-TM antibody IgGl (n = 7); E = heat-treated anti-TM antibody IgGl (n = 4); F = anti-TM antibody IgG2 (n = 6); G = actively immunized guinea pigs (n = 4). Each value represents the mean +SD and p values are also calculated.
TMA-Induced Lung Injury in Guinea Pigs
125
Table 4. Immunochemical properties of isolated of anti-TM antibody Material
Hemagglutination
Complement fixation
PCA
Protein recovery (%)
IgG subclass
Whole serum P-1 P-2 P-3 P-4
216
2 '2
_
_
26
2»i>
2s 2U> 2 >
IgG2 IgGl
2M
2 >
2" T
22.4 26.4 20.5 17.5
2>o
2>
2 13
_ -
sions associated with TM-spccific antibody and the profiles resembled human syndromes [5]. TMA is a haptenic substance incapable of stimulating the im mune system by itself without carrier substances. However, when animals can be immunologically sen sitized through airways by TMA inhalation, the in haled TMA has to be converted into a certain immu nogenic substance capable of stimulating their im mune systems. It seems likely that protein molecules in lavage fluids and immunocytes are involved in this conversion process. At present, however, the mecha nism of sensitization by TMA inhalation is still un clear. The present study was undertaken to investigate the immunological mechanism of TMA-mediated lung injury in conventional guinea pigs, because they were highly susceptible to allergic responses, passive cutaneous and systemic anaphylaxis. It was demon strated that the injection with TM-carrier conjugates in guinea pigs resulted in the production of specific antibodies against the TM epitopes conjugated with BSA in the sera of these animals. A positive correla tion was seen between the number of TM epitopes conjugated to BSA and the antibody titer of the anti sera. This finding is compatible with the report of Ru bin [20] and suggested that the production of anti-TM antibody is associated with epitope density on the car rier protein. The guinea pigs actively immunized with TM BSA developed severe pulmonary hemorrhages after a single exposure to TMA inhalation. This sug gests that inhaled TMA acts directly or indirectly as an antigenic substance in lung injury. Furthermore, guinea pigs sensitized intraveneously or intraperitoneally with antiserum containing anti-TM antibody al so developed pulmonary hemorrhages after TMA in halation. This result is in agreement with that report
ed by Zeiss et al. [6], who showed that the lung lesions in normal rats could be induced by the passive sensiti zation with humoral antibodies followed by TMA in halation. In this investigation, the inhalation exposure was performed at a TMA concentration ranging from 130 to 180 mg/m-1 because no injury was found in animals’ lungs after repeated inhalation of low concentrations of TMA. This result conflicts with that of SPF SD rats reported by Leach et ai. [5]. In our preliminary exper iments, only 1 of 8 SPF SD rats exposed to TMA in halation developed slight pulmonary hemorrhages. The differences between our experiments and that of Leach et al. may be the size of the TMA particles and the inhalation systems. TMA particles of about 5 urn in size were used in these experiments, whereas Leach et al. used particles of about 1 pm. This suggests that the aerodynamic property of TMA particles plays an important role in the induction of lung injury. Anoth er explanation for the discrepancy of the results may be a difference in the animal species studied. Guinea pigs may be difficult to sensitize by TMA inhalation alone. It is well known that guinea pig antisera against protein antigens contain considerable amounts of an tibodies in both of the immunoglobulin subclasses, IgGl and IgG2, and that these antibodies are distin guishable by their contrasted immunochemical prop erties [21-23]. The antibodies associated with IgGl subclass have a cytotropic affinity to homologous ani mal tissues, but lack a complement fixing activity, whereas the antibodies associated with IgG2 subclass have a complement fixing activity, but lack a homocytotropic affinity. Henson and Cochrane [24] reported that homocytotropic antibodies in recipient rabbit se rum seemed to be necessary for the successful media
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
The peak fractions of anti-TM antibody isolated by DEAE-cellulose column chromatography were denoted as P-1, P-2, P-3 and P-4. The measurements of hemagglutination, complement fixation and PCA titers of these fractions were performed (see Materials and Methods). The protein recovery and the corresponding subclass are also included in this table.
Tao/Sugiura/Nakamura/Kido/Tanaka/Kuroiwa
126
reaction [31]. These findings lead us to suggest that the IgGl antibody binds to mediator cells such as mast cells, which release chemical mediators induced by TMA inhalation. In the case of the IgG2 antibody, it seems likely that a small portion of intravencously in jected antibody is secreted into the pulmonary alveoli via the lung capillary, where the antibody reacts with its antigen which is composed of the autologous BALF protein and inhaled TMA. The resulting im mune complex may deposit in different anatomic compartments of the lung, i.e. intravascular, intraalveolar and interstitial. At these points, the classical pathway of complement components is activated, re sulting in the generation of chemotactic peptides such as C5a that attract neutrophils from the circulation. The neutrophils then phagocytize immune complexes and release oxygen-derived metabolites, proteases and so forth [32], In conclusion, TMA inhalation induced pulmonary injury in guinea pigs actively immunized with TM BSA conjugates and passively sensitized with homolo gous cytotropic or noncytotropic antibody, and the results indicated that the complement-independent acute type hypersensitivity and the complement-de pendent immune complex type reactions were in volved in this pathogenesis. Guinea pigs appear to be a useful animal model for TMA-induced lung injury. Acknowledgements This work was supported in part by a Grant-in-Aid for Scien tific Research from the Ministry of Education. Japan (No. 02670352).
References 1 Zeiss CR. Patterson R. Pruzansky JJ. Miller MM. Rosenberg M. Levitz D: Trimellitic anhydride-induced airway syndromes; clinical and immunologic studies. J Allergy Clin Immunol 1977;60:96-103. 2 Petterson R. Addington W. Banner AS, Byron GE. Franco M, Herbert FA. Nicotra MB. Pruzansky JJ. Rivera M. Roberts M. Yawn D, Zeiss CR: Antihapten antibodies in workers exposed to trimellitic anhydride fume; a potential immunopathogcnetic mechanism for trimellitic anhydride pulmonary disease-ane mia syndrome. Am Rev Rcspir Dis 1979;120:1259-1267. 3 Herbert FA. Orford R: Pulmonary hermorrhages and edema due to inhalation of resin containing trimellitic anhydride. Chest 1979;76:546-552. 4 Sale SR. Roach DE. Zeiss CR. Patterson R: Clinical and im munological correlations in trimellitic anhydride airway syn dromes. J Allergy Clin Immunol 1981;68:188-193.
Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
tion of glomerular and arterial injury. In addition, Roska et al. [25] suggested that pigeon serum-induced pulmonary inflammation in guinea pigs was a mani festation of a complement-dependent, humoral im mune mechanism of pathogenesis and thus was con sistent with an immune complex disease. Also. Richcrson [26] reported that scattered alveolitis induced by aerosolized antigens, depending on the type of de layed hypersensitivity without demonstrable antibody, developed in guinea pig lungs. Therefore, the follow ing experiments were attempted to determine which subclass antibody, IgGl, IgG2 or both were involved in the induction of pulmonary injury by TMA inhala tion. Each of the subclass antibodies was isolated from pooled hyperimmune antisera against TM BSA using gel filtration and ion exchange column chroma tographies, and the purity of the isolated preparation was examined by immunochemical procedures, dou ble diffusion and immunoclectrophoresis, comple ment fixation and PCA reactions. The results demon strated that the IgGl fraction contained less than 1% IgG2 and IgG2 fraction contained less than 1% IgGl. To evaluate the pulmonary injury more quantitatively, the enhanced permeability of Evans blue dye in the extravascular tissues after TMA inhalation were mea sured in the lungs of guinea pigs passively sensitized with isolated antibodies. Furthermore, the effect of heat treatment on the sensitizing activity of these an tibodies was also examined. With respect to vascular permeability, not only the IgGl, but also the IgG2 fraction was comparable with the antisera, suggesting that homocytotropic as well as nonhomocytotropic antibodies are involved in the induction of lung injury by TMA inhalation. Heat treatment of the antiserum and the IgGl fraction did not influence vascular per meability, which suggested that heat-labile IgE anti body was not involved in this process. It has been re ported that many chemical mediators cause a signif icant acute exudation of protein-rich plasma [27-29], A possible interaction between these mediators, mediator cells and inflammatory cells in acute and chronic asthma was demonstrated by Kay [30], who showed that the initial burst of lung mast cell-derived mediators leads to recruitment and activation of sec ondary inflammation cells, including activated neu trophils and monocytes. Recent evidence suggests that late reactions may be associated with the reacti vation of mast cells since a second elevation in serum neutrophil chemotactic activity and plasma histamine concentration was observed during late-phase asthma
127
TMA-Induccd L.ung Injury in Guinea Pigs
20 Rubin B: Studies on the induction of antibody synthesis
21
22
23
24
25
26 27
28
29
30 31
32
against sulfanilic acid in rabbits. I. Effect of the number of hapten molecules introduced in homologous protein on anti body synthesis against the hapten and the new antigenic deter minants. Eur J Immunol 1972:2:5-11. Ovary Z, Benacerraf B. Bloch KJ: Properties of guinea pig 7S antibodies. Identification of antibodies involved in passive cu taneous and systemic anaphylaxis. J Exp Med 1963:117:951964. Bloch KJ. Lourilsky FM. Ovary Z, Benacerraf B: Properties guinea pig 7S antibodies. Identification of antibodies involved in complement fixation and hemolysis. J Exp Med 1963:117: 965-981. Oliberia B. Oslcp AG. Siraganian RP, Sandberg AL: The bi ologic activities of guinea pig antibodies. I. Separation of y,and y2-immunoglobulins and their participation in allergic re actions of the immediate type. J Immunol 1970;104:320-328. Henson PM. Cochrane CG: Acute immune complex disease in rabbits. The role of complement and of a leukocyte-dependent release of vasoactive amines from platelets. J Exp Med 1971; 133:554-571. Roska AK. Moore VL. Abramoff P: Immune complex disease in guinea pig lung: Elicitation with pigeon serum. Am Rev Respir Dis 1979;120:129-136. Richerson HB: Acute experimental hypersensitivity pneumo nitis in the guinea pig. J Lab Clin Med 1972;79:745-757. Persson CGA. Erjefalt I: Inflammatory leakage of macromole cules from the vascular compartment into the tracheal lumen. Acta Physiol Scand 1986;126:615-616. Persson CGA. Erjefalt I. Karlsson JA: Effect of tachykinin an tagonists on the tracheobronchial microcirculation: in Hakanson R. Sundler F (eds): Tachykinin Antagonists. Amsterdam, Elsevier, 1985, pp 171-179. Persson CGA, Erjefalt I, Andersson P: Leakage of macromole cules from guinea pig tracheobronchial microcirculation. Ef fects of allergen, leukotricnes, tachykinins, and antiasthma drugs. Acta Physiol Scand 1986:127:95-105. Kay AB: Inflammatory cells in acute and chronic asthma. Am Rev Respir Dis 1987; 135:S36— S66. Durham SR, Lee TH. Cromwell O: Immunologic studies in al lergen-induced late-phase asthmatic reactions. J Allergy Clin Immunol 1984;74:49-60. Daniclc RP, Henson PM. Fantone JC, Ward PA, Preisin RB: Immune complex injury of the lung. Am Rev Respir Dis 1981; 124:738-755.
Received: April 22, 1991 Accepted after revision: May 23. 1991 Correspondence to: Dr Yoshiaki Tao Department of Immunology School of Medicine University of Occupational and Environmental Health Kitakyushu 807 (Japan) Downloaded by: King's College London 137.73.144.138 - 1/14/2018 9:26:48 AM
5 Leach CL, Hatoum NS, Ratajezak HV, Zeiss CR. Roger JC. Garvin PJ: The pathologic and immunologic response to in haled trimellitic anhydride (TMA) in rats. Toxicol Appl Phar macol 1987;87:67-80. 6 Zeiss CR. Leach CL. Smith 1J, Levitz D, Hatoum NS, Garvin PL. Patterson R: A serial immunologic and histopathologic study of lung injury induced by trimellitic anhydride. Am Rev Respir Dis 1988;137:191-196. 7 Turner ES. Pruzansky JJ. Patterson R. Zeiss CR. Roberts M: Detection of antibodies in human serum using trimellitylerythrocytes: direct and indirect hacmagglutination and hae molysis. Clin Exp Immunol 1980;39:470-476. 8 Zeiss CR. Levity D. Chacon R. Wolkonsky P. Patterson R. Pru zansky JJ: Quantitation and new antigen determinant (NAD) specificity of antibodies induced by inhalation of trimellitic an hydride in man. Int Arch Allergy Appl Immunol 1980:61:380388. 9 Patterson R, Roberts M. Zeiss CR. Pruzansky JJ:Human anti bodies against trimcllityl proteins: Comparison of specificities of IgG, IgA and IgE classes. Int Arch Allergy Appl Immunol 1981:66:332-340. 10 Chandler MJ. Zeiss CR. Leach CL. Hatoum NS. Levitz D. Gravin PJ. Patterson R: Levels of specific antibody in bronchoalvcolar lavages (BAL) fluid in an animal model of trimel litic anhydride-induced by injury. .1 Allergy Clin Immunol 1987;80:223-229. 11 Nakamura H, Sugiura T: Inhibition of anaphylactic sensitiza tion activity of guinea pig antibody by binding with protein A from Staphylococcus aureus. J UOEH 1982;4:241-253. 12 Liu SH. Koo PH. Cebra JJ: Effect of carrier priming on the distribution of anti-hapten antibodies between IgG 1 and IgG2 isotypes in the hyperimmune guinea pig. J Immunol 1974; 113:677-687. 13 Avrameas S. Taudou B. Chuilon S: Glutaraldehyde, cyanuric chloride and tetraazotized o-daianisidine as coupling reagents in the passive hemagglutination test. Immunochemistry 1969;6: 67-76. 14 Ovary F: Local anaphylaxis; in Williams CA. Chase MW (eds): Methods in Immunology and Immunochemistry. New York. Academic Press, 1976, vol V, p 16. 15 Tanaka 1. Matsuno K. Kodama Y. Akiyama T: Pulmonary de position of a fly ash aerosol by inhalation. J UOEH 1983; 5:423-431. 16 Udaka K. Takeuchi Y. Movat HZ: Simple method for quantita tion of enhanced vascular permeability. Proc Soc Exp Biol Med 1970:133:1384-1387. 17 Thompson JA. Scypinski LA, Gordon T. Sheppord D: Tachykinins mediated the acute increase in airway responsiveness caused by toluene diisocyanate in guinea pigs. Am Rev Respir Dis 1987;136:43-49. 18 Howe E, Venables KM. Topping MD. Dally MB. Hawkins R. Law JS: Newman Taylor AJ: Tetrachlorophthalic anhydride asthma: evidence for specific IgE antibody. J Allergy Clin Im munol 1983:71:5-11. 19 Newman Taylor AJ. Venables KM. Durham SR. Grancek BJ. Topping MD: Acid anhydrides and asthma. Int Arch Allergy Appl Immunol 1987;82:435-439.
