Chem. Res. Toxicol. 1988,1, 281-287
28 1
Intra- and Intermolecular Reactions of 4,4’-Diisocyanatodiphenylmethane with Human Serum Albumin Ruzhi Jint and Meryl H. Karol* Jilin Provincial Institute of Industrial Health and Occupational Disease, Changchun, Jilin Province, People’s Republic of China, and Department of Industrial Environmental Health Sciences, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 Received May 19, 1988
Diisocyanates are highly reactive industrial chemicals that have been shown to possess toxic activity, including potential for allergic sensitization. T o assist in diagnosis of sensitization, immunoassays for diisocyanate-specific antibodies are performed; such assays require preparation of diisocyanate-containing hapten-protein conjugates. Conditions were investigated for forr tion of conjugates yielding varying degrees of hapten binding. Relative concentrations of h tens and proteins were varied as were pH, temperature, and time of reaction. Quantitat n of 4,4’-diisocyanatodiphenylmethane(MDI) binding with human serum albumin (HSA) was assessed by absorbance of the isolated conjugates a t 250 nm after determination of the molar extinction coefficient for MDI. At pH 7.4 and 37 OC, the binding reaction was found to be biphasic with binding of 5-6 mol of MDI groups/mol of HSA within the first minute, followed by incorporation at a rate of 0.16 mol/min during the next 2 h. Evaluation of reaction products using SDS-PAGE revealed extensive inter- and intramolecular cross-linking of HSA by MDI. Intramolecular cross-linking was accompanied by an increased migration of conjugates from an initial molecular mass of 66 kDa, typical of HSA, to a molecular mass of 44 kDa. The change in migration was also produced by using disuccinimidyl tartarate (DST) as hapten and was eliminated when DST was cleaved with sodium periodate. It was attributed to altered protein shape. Conditions that favored binding of MDI with HSA were a high relative concentration of MDI:HSA, a pH of 9.4, and a temperature of 37 “C. Under such conditions it was calculated that 53 mol of MDI were bound per mole of HSA after 24 h. The ability of the various conjugates to detect antibodies to MDI in sera from individuals and animals exposed to MDI was assessed and is described in the following paper in this issue.
Introduction Isocyanates are in widespread industrial use in the manufacture of polyurethane foams, sprays, and paints as well as in production of herbicides and pesticides (I). Several isocyanates have been reported to cause allergic reactions in exposed workers (2,3). Symptoms of sensitization range from bronchial asthma to hypersensitivity pneumonitis. Although the sensitization properties of particular isocyanates have been recognized for many years, little is known regarding the mechanisms underlying sensitization. Several investigators have reported the presence of specific antibodies in workers exposed to isocyanates (4-6). Detection of such antibodies can be used both as an indicant of exposure (7) and, in detection of IgE antibody, as a risk factor for allergic response (5, 8). Moreover, if serial blood samples are drawn, the pattern of antibody titers may indicate the occasion of the actual exposure ( 4 ) . The chemically reactive nature of isocyanates necessitates use of hapten-protein conjugates for antibody detection. Since some industrial isocyanates are bifunctional, preparation of conjugates inevitably results in intra- and intermolecular cross-linking of the carrier proteins. The extent of these reactions is dependent upon the conditions *To whom corremondence should be addressed at the Universitv of Pittsburgh. t Jilin Provincial Institute of Industrial Health and OccuDational Disease.
selected for conjugate preparation, i.e., protein concentration, pH, temperature, and length of reaction time. In view of the importance associated with detection of antibodies and the difficulty encountered by some investigators in identifying antibodies in symptomatic workers, we undertook an extensive investigation into the reaction products formed from incubation of one diisocyanate, 4,4‘-diisocyanatodiphenylmethane (MDI), with human serum albumin, the latter being the most frequently employed protein carrier for hapten conjugate production to detect human antibodies. Subsequently, the various conjugates were evaluated for ability to detect antibodies to MDI. This study characterizes the hapten-protein conjugates. The following paper (21) compares the conjugates for ability to detect antibodies, produced in workers and in an animal model, following known exposure to MDI.
Materials and Methods Purification of Human Serum Albumin (HSA). HSA (Sigma Co., crystallized, 15.0% nitrogen) was used as obtained or at times further purified on a column of Affi-Gel Blue (50-100 mesh, Bic-Rad Corp.). The gel was washed with 0.02 M phosphate buffer, pH 7.1, and poured into a 46 X 1.7 cm glass column. Five hundred milligrams of HSA in 2 mL of buffer was applied to the column. The HSA fraction was eluted with 1.4 M NaCl in the 0.02 M phosphate buffer. The column was regenerated by using 2 M guanidine hydrochloride incorporated into the buffer, and the purification was repeated. The eluted HSA was isolated by dialysis followed by lyophilization. The yield was typically 40%. Isocyanate-HSA Conjugates. MDI (Mondur M. Mobay Chemical Corp.) was recrystallized from n-hexane. Immediately 0 1988 American Chemical Society
282 Chem. Res. Toxicol., Vol. I , No. 5, 1988
J i n and Karol
Table I. Extinction Coefficient of MDI Determined from Reaction of Recrystallized MDI with Excess Glycine molar extinction solution" glycine (mg) MDI (pg) molar ratio, g1ycine:MDI ,A (nm) absorbance at A,, coefficient* a 0.06 21 101 243.3 0.350 20 833 b 0.2 21 33:l 245.1 0.415 24 714 C 2 21 333:l 246.8 0.480 28 577 d 20 21 333 x 1O:l 248.8 0.592 35 238 e 200 21 333 x 1001 248.8 0.615 36 613 f 400 21 333 x 2001 249.1 0.598 35 583 g 1000 21 333 X 500:l 249.0 0.606 36 059 nThe absorbance profile of each solution is shown in Figure 1. *The extinction coefficient was determined for each solution by using Beer's Law. The mean value for solutions d-g was 35 873. before use, it was dissolved in acetone and then added dropwise to a 0.5% solution of HSA dissolved in either 0.05 M phosphate buffer, pH 7.4, or 0.05 M borate-KC1 buffer, pH 9.4. The final concentration of acetone in the solution was always 3%. Reaction mixtures were stirred for periods up to 24 h at temperatures from 0 to 37 O C . Reactions were stopped by the addition of a 10-fold molar excess of monoethanolamine (MEA, Fisher Scientific Co., Pittsburgh). Preparations were then filtered, dialyzed, and lyophilized. A similar procedure was used to prepare conjugates of p-tolyl isocyanate (pT) with HSA. The isocyanate (Eastman Chemicals) was added dropwise to the rapidly stirred 0.5% HSA solution, and then the above procedure was followed. Disuccinimidyl Tartarate (DST). DST (lot no. 870112082, Pierce Chemicals) was reacted with HSA in a manner analogous to that used for isocyanate binding. The chemical was dissolved in acetone and then added dropwise to a 0.5% solution of HSA in 0.05 M phosphate buffer, pH 7.4, to achieve final molar ratios of DST to HSA of 1 0 1 to 180:l. At intervals of 4,24, and 48 h, 5-mL samples were removed, 5-fold excess MEA was added to quench the reaction, and the conjugates were isolated as described for MDI-HSA products. To cleave the hapten, 1 mg of DST-HSA conjugates was dissolved in 1 mL 0.02 M phosphate buffer, pH 7.5, containing 0.15 mM sodium periodate (meta, Fisher Scientific) and maintained at ambient temperatures for 4 h. Extinction Coefficient of MDI and p -Tolyl Isocyanate. The molar extinction coefficients were determined by reacting the isocyanates with excess glycine. MDI was recrystallized and then dissolved in acetone at 2.1 mg/mL. Ten microliters of this solution was added to 0.16 mM-2.8 M glycine in 0.02 M phosphate buffer, pH 7.5. For, pT conjugates, the isocyanate was diluted in acetone and then reacted with glycine in ratios of 201 to 2000:l (g1ycine:pT). The absorption profiles of the resulting conjugates were determined a t 200-350 nm. SDS-PAGE Gradient Electrophoresis. Mobilities of conjugates were determined by using gradient gels (4-22.5%) prepared according to the method of Lamelli (9). Just prior to use, 1 mg of each conjugate was dissolved in 1mL of glass distilled water, containing up to 20 MLof 0.1 N NaOH. They were diluted to 200 pg/mL in 0.16 M tris buffer, pH 8.8, containing 1.3% sodium dodecyl sulfate (SDS, Bio-Rad Corp.), 0.1% (w/v) sodium azide (Fisher Scientific), 1.7% (v/v) 2-mercaptoethanol (Bio-Rad),and 10% (w/v) sucrose (Fisher Scientific). Samples werre heated for 3 min at 100 "C and then chilled in an ice bath. Twenty microliters was then placed in each lane of the gel. Electrophoresis was performed at 30 W and 600 V (50 mA) for approximately 1 h in a cold room at 10 "C. Proteins were visualized by using 0.1% Coomassie Blue R250 in a solution containing 25% ethanol and 8% acetic acid. Absorption was determined by using a Quick Scan densitometer (Helena Labs). The mobility was determined by comparison with a calibration curve obtained with protein standards (Bio-Rad).
Results Extinction Coefficient of MDI. T o determine t h e MDI content of hapten-protein conjugates, i t was necessary t o first obtain t h e molar extinction value of t h e MDI-amine product. Accordingly, MDI was reacted with increasing amounts of glycine i n buffered solution at pH 7.5. The absorption profiles of t h e reaction products are
-225
250
300
Nanometers
Figure 1. Ultraviolet difference spectrum of MDI-glycine conjugates versus glycine. Ten microliters of a solution of MDI in acetone (0.21% w/v) was added to increasing concentrations of glycine dissolved in 0.02 M phosphate buffer, pH 7.5. The molar ratio of lycine to MDI was (a) l O : l , (b) 331, (c) 33 X 101:1, (d) 33 X 105:1, (e) 33 X 103:1, (f) 66 X 103:1, and (g) 16 X 104:l.
-"
0.6
J
GLYCINE : MDI [MOLAR RATIO)
Figure 2. Absorbance at 250 nm of MDI-glycine conjugates formed by varying the ratio of reactants. Reaction conditions are described in Figure 1. shown i n Figure 1 with quantities of reactants and t h e calculated extinction coefficients listed in Table I. Excess glycine was added to assure electrophilic reaction of isocyanates as opposed to hydrolysis or self-polymerization reactions. Both t h e absorbance a n d t h e wavelength of maximum absorbance increased with increasing ratios of glycine to MDI. T h e increase in absorbance was linear as t h e amount of glycine t o M D I was varied from 1O:l to 33 X 102:1 (Figure 2). Higher concentrations of glycine were without effect. A t maximum absorption (250 nm), the molar extinction coefficient was calculated a n d yielded a mean value of 35 873. By use of this value, the amount of M D I
MDI-Serum Albumin Conjugates Table 11. Haptenic Substitution of MDI-HSA Conjugates Formed at 37 "C, pH 7.4 mol of mol of amino molar ratio, MD1:mol of groups MD1:HSA at reaction HSA in reacted/mol start time (h) conjugatea of HSAb 101 4 4 13 14 24 5 48 7 18 501 4 12 43 24 41 17 48 37 16 2m1 4 48 35 24 43 37 48 40 34
Determined from ultraviolet difference absorbance a t 250 nm of the reaction product versus HSA. Proteins were weighed analytically, dissolved in 0.05 M phosphate buffer, pH 7.4,and diluted to equal concentration. bThe number of amino groups left unreacted on the MDI-HSA conjugates was determined by using the trinitrobenzenesulfonic acid reagent (IO).
reacted with HSA under various reaction conditions was calculated. Reaction of MDI with Human Serum Albumin (HSA). MDI was reacted with HSA in a 0.05 M phosphate buffer, pH 7.4. The amount of MDI added to the solution was varied to achieve an initial molar ratio of MDI to HSA of 1O:l to 200:l. Following dialysis and lyophilization of the reaction products, the absorption spectra were examined. The maximum absorption was noted at 250-257 nm with a slight shift to higher wavelengths occurring as more MDI was reacted with HSA (see Table I). This pattern was similar to that seen with MDI-glycine reactions as noted in Figure 1. By use of the molar extinction coefficient of 35 873 (Table I), the moles of MDI reacted with HSA in the three preparations were calculated and are listed in Table 11. Increasing the amount of MDI relative HSA in the solution resulted in greater binding of MDI to the protein. The effect of increasing reaction time on the degree of MDI bound to HSA is summarized in Table 11. Reactions were allowed for periods up to 48 h. With the 1O:l molar ratio of MDI:HSA, reactions continued to 48 h. When an initial ratio of 50:l or 200:l of MD1:HSA was employed, the reactions at 48 h did not differ from those at 24 h. The number of amino groups reacted on HSA was assessed by use of the trinitrobenzenesulfonic acid reagent (IO). Results, listed in Table 11, always indicated more amino groups on the protein had reacted when compared with the number of MDI moieties bound to the HSA molecules. By contrast, calculation of the number of moles of the monofunctional pT bound to serum albumin yielded similar results when based on the absorbance of pT conjugates or on the TNBS reaction (Table 111). These results suggested bifunctional reactions of MDI. Intermolecular Reactions. The recognized bifunctional cross-linking ability of MDI prompted investigation of possible polymerization of HSA especially under conditions of high MD1:HSA ratios and prolonged reaction times. MDI-HSA conjugates were examined by using SDS-PAGE employing denaturing conditions and 4-22.5 % gradient gels. Intermolecular cross-linking was apparent and increased as a function of time (Figure 3). The major product at the 4-h time period, from the 1O:l (MD1:HSA) reaction, migrated with an M , equivalent to a 66-kDa mass (lane a). A minor band was detected having a mass equivalent to 130 kDa. With increased time, both bands became sharper (lanes b and c). When reactions were conducted with
Chem. Res. Toxicol., Vol. 1, No. 5, 1988 283 Table 111. Haptenic Substitution of Conjugates Formed by Reaction of p-Tolyl Isocyanate with Serum Albumin mol of amino mol of pTmol of groups protein in reacted/mol of conjugatea conjugateb proteinC pT-GSA 28 31 26 pT-HSA (I) 29 21 pT-HSA (11) 16 pT-HSA (111) 27 30
aConjugates with guinea pig (GSA) or human serum albumin (HSA) were prepared by reaction of p-tolyl isocyanate (pT) with protein, at ratios of 501 to 100:1,a t 0 "C,pH 9.4,for 1 h. bValues were obtained by using a molar extinction coefficient of 15640 a t 240 nm determined from reaction of pT with excess glycine (see text). Values determined from reaction with trinitrobenzenesulfonic acid reagent. a
b
c
d
e
f
g
h
i
j
kl
20c
116
66
45
Figure 3. SDS-PAGE of MDI-HSA conjugates. MDI was reacted with HSA in molar ratios of lO:l, 50:1, and 2001 in 0.05 M phosphate buffer, pH 7.4. Aliquots were taken from each reaction vessel a t 4,24,a t 48 h and processed as described in the text. Samples were heated a t 100 "C for 3 min in the presence of 1.7% mercaptoethanol. MD1:HSA ratio of 1O:l: lane a, 4 h; lane b, 24 h; lane c, 48 h. Ratio 50:l: lane d, 4 h; lane f, 24 h; lane g, 48 h. Ratio 2001: lane h, 4 h; lane i, 24 h; lane j, 48 h. Lane e, protein standards, values at left.
higher ratios of MDI relative to HSA, the proportion of polymers, relative to monomers, increased. Reaction of MDI with HSA using an initial ratio of 200:l produced several polymeric species (lanes h-j). The appearance by 4 h of at least four products prompted further investigation into the formation and nature of these products. MDI was dissolved in acetone and then added dropwise to a 0.5% solution of HSA which had been purified by elution from an Affi-Gel column. At the times indicated in Figure 4A, aliquots were removed and dispensed into tubes containing a 10-fold molar excess of monoethanolamine. Conjugates were isolated as described previously. Migration of the major band (lower arrowhead) increased with the time of reaction. A second band (top arrowhead) was apparent by 5 min of reaction and also gave increased migration with time. The change in mo-
284
Chem. Res. Toxicol., Vol. 1, No. 5, 1988
Jin and Karol
kD 200
116 92
66
45
BbO uo -
Q.
=--.+?R!HER
A
,304
._.._...._. €l
0
\
0
10
20 30
40
50
60 " 2 4
8
24
HOURS Figure 4. SDS-PAGE of reaction products isolated at 0-24 h from a solution containing MDI and affinity-purified HSA in 0.05 M phosphate buffer, pH 7.4, in a starting molar ratio of 200:l. The reactants were stirred continuously and maintained at 37 "C. (A) Lanes a-k, aliquots isolated at 0-24 h; lane 1, HSA; lane m, protein standards, with corresponding mass at right. Arrowheads indicate initial position of monomeric (lower) and dimeric (upper) MDI-HSA species. (B) Relative mobility of various polymeric species isolated during the 24-h reaction. Equations of lines: monomer, y = 0.2377~+ 65.9144, r = -0.9944; dimer, y = 0.4722~+ 131.9810, r = -0.9767. MINUTES
bility of each of the conjugate species as a function of time is shown in Figure 4B. Migration increased a t a constant rate during the first 2 h of reaction. During the ensuing 22 h, little change in mobility of the products was noted. The polymeric nature of the species was apparent from estimation of the molecular masses. At the start of reaction (time 0, lane a, Figure 4A) the single band had a relative molecular mass of 66 kDa, consistent with that of HSA. By 10 min of reaction, a molecular mass of 130 kDa was readily determined for the second band. The amount of the second band increased with reaction time. Its molecular mass was always close to twice that of the major band. The rates of change of mobility for the two species (Figure 4B) were compared by linear regression analysis. The slope of the "dimer" band was twice that of the monomer band. By 2 h, a third band was detected (Figure
Table IV. Formation of Polymeric Conjugates from Reaction of MDI with Human Serum Albumin MDI hapten:proteino monomer (%) dimer (9%) trimer (%) 96 4 0 101 501 90 10 0 200 1 39 31 29 a
Molar ratios of reactant species. Reactions were performed at
37 "C in 50 mM phosphate buffer, pH 7.4, and terminated 4 h.
Amounts of each species were determined by densitometry tracings of the stained gels.
4A, see lanes h-k). Its migration indicated a relative mass three times that of the monomer (see also Figure 4B). By 4 h, a faint fourth band was detected by gel densitometry having a mass of approximately 190 kDa. The relative amounts of polymeric species formed under the above conditions are listed in Table IV. Increasing MDI relative to HSA favored formation of polymeric species. With a 200:l initial ratio, almost equal quantities of the monomer, dimer, and trimer were detected. Tetramers were detected only in trace amounts. Reaction of HSA with Disuccinimidyl Tartarate (DST). An explanation was sought for the observed enhanced mobility of MDI-HSA conjugates in SDS-polyacrylamide gel. Consideration was given to the possibility that the protein shape had been altered by extensive intramolecular cross-linking resulting in modified migration in gel. To address this possibility, HSA was reacted with DST under conditions identical with those employed for reaction of MDI with HSA. DST was selected since it reacts readily with amino groups on proteins (11) and, following reaction, the reagent can be cleaved with sodium periodate, thus destroying the cross-link bonds. Conjugates of DST-HSA were prepared by reacting DST with HSA at a 1801 molar ratio. Following reaction for 24 h at 37 "C, the producs were isolated and evaluated, by using the TNBS reagent, as well as by SDS-PAGE analysis. Reaction with TNBS indicated that 43% of the amino groups on HSA had been substituted with DST. The migration of the DST-HSA conjugates is shown in Figure 5. Two broad bands were observed (lane g) having migration equivalent to 62 and 130 kDa. This result was similar to that obtained with MDI-HSA conjugate (lane d) in that polymeric species were formed and the molecular mass of the monomeric species was less than the 66 kDa of unreacted HSA (compare serum albumin band in lane e with lanes d and g). To confirm that the increased migration reflected altered protein shape and/or reduced flexibility, as a result of extensive intramolecular cross-linking of HSA, the DSTHSA conjugate was cleaved by placing it in a solution containing 15 mM sodium periodate for 4 h. Comparison of the migration of DST-HSA before and following treatment is shown in Figure 5 (lanes g and f, respectively). The mobility of DST-HSA reverted from a molecular mass of 62 kDa before periodate treatment to 66 kDa after treatment. It was concluded therefore that the enhanced migration of the conjugate was the result of intramolecular cross-linking of HSA by the bifunctional DST reagent. Effect of pH and Temperatureon Reaction of MDI with HSA. The effects of increased pH and temperature are summarized in Table V. The most extensive reactions were observed under the following conditions: 200:l initial molar ratio of MDI to HSA, pH of 9.4, and temperature of 37 "C. Under such conditions, haptenic substitution was calculated to be 53 mol of MDI/mol of HSA, and extensive inter- and intramolecular reactions were noted from patterns of migration in gel. Raising the reaction
Chem. Res. Toxicol., Vol. 1, No. 5, 1988 285
MDI-Serum Albumin Conjugates - a .b
c
d
e
f
g
40
200,
116,
92, 66b
HOUR 4 5 r
Figure 5. SDS-PAGE of DST-HSA and MDI-HSA conjugates before and after addition of 0.015 M sodium periodate. Samples of conjugates were treated with sodium periodate for 4 h prior to denaturation and addition to the gel. Lanes a and b, MDI-HSA prepared a t 37 "C using 200:l ratio of MDI to HSA; lanes c and d, 50:l ratio of MDI to HSA; lanes f and g, 180:l ratio of DST to HSA. Lanes a, c, and f, conjugates were incubated with sodium periodate before heat denaturation. Table V. Effect of pH and Temperature on Reaction of MDI with HSA starting reaction mol of MDI migration temperareacted with molar ratio, massb MD1:HSA ture ("C) pH HSA" (kDa) 1O:l 37 7.4 5 66 501 37 7.4 17 56 2m1 37 7.4 37 46 0 7.4 24 52 200:l 200:l 37 9.4 53 44
OReactions were performed for 24 h, and then isocyanate was inactivated by addition of excess monoethanolamine. Values for the isolated conjugate were obtained from the UV absorbance spectra in 50 mM phosphate buffer at pH 7.4. 6Mobility of the monomeric species.
temperature from 0 to 37 "C at pH 7.4, and increasing the pH from 7.4 to 9.4, both contributed to increased haptenic substitution. The relationship between the amount of hapten bound (as assessed by absorption at 250 nm) and the extent of intramolecular cross-linking (as evidenced by enhanced migration in SDS-PAGE) is displayed in Figure 6. The slopes of the two lines were indistinguishable.
Discussion Diisocyanates are recognized for their reactivity with numerous functional groups on proteins and other biomacromolecules (12). It is this chemical reactivity that undoubtedly is responsible for their multiple toxic effects, which include irritation, sensitization, and pneumonitis (3). Diisocyanates have been used historically as bifunctional reagents to create synthetic protein dimers. Classically,
Figure 6. Haptenic content and migration in SDS-PAGE of MDI-HSA conjugates formed during 24 h of reaction. At the indicated times, aliquots were removed from the reaction vessel, reacted with excess monoethanolamine, and then isolated by dialysis and lyophilization. MDI content of conjugates was determined from absorbance a t 250 nm (after subtraction of contribution from HSA) by using an E, of 35 873. Migration position was determined from densitometer tracing of the Coomassie Blue stained gel (see Figure 4A) and comparison with a standard curve. Equations of the lines were as follows: ( 0 )molar ratio MDLHSA, y = 0.1604.~ + 5.558,r = 0.9840;(0) monomeric band, y = -0.1559~ + 64.1509, r = -0.9476.
tolylene diisocyanate has been employed to link antigens with specific y-globulins and for coupling of small molecules to protein for use as synthetic antigens (13). Our interest in the reaction of MDI with human serum albumin arose from the need to prepare standardized MDI-protein reagents for use in screening human sera for MDI-specific reaginic antibodies (Le., those associated with allergic sensitization). The high reactivity of the chemical made it necessary to define parameters such as temperature, pH, and reaction time that would produce optimal binding of MDI with the carrier protein and evaluate the consequences of variation of these parameters on the reaction products. Routinely, the products from reaction of diisocyanates with protein have been characterized solely by determining the average number of isocyanate haptenic groups bound per mole of protein (6,14,15).This calculation has been frequently made by assuming reaction of isocyanates with lysine residues and accordingly from use of trinitrobenzenesulfonic acid for determination of the number of amino groups reacted on the protein. In the current study, we used this procedure but, additionally, compared results with those obtained from ultraviolet absorption of the conjugate. In the latter procedure, it was necessary to determine the molar extinction value for MDI. Employing excess glycine to assure nucleophilic reaction of the amino acid with isocyanate groups, the molar extinction of MDI a t 250 nm was found to be 35873. Conjugates were then evaluated by both methods (Table 11). Use of the TNBS reagent to assess the number of amino groups on HSA that had reacted with MDI always gave values considerably greater than those obtained from the absorbance spectra of the conjugates. Concern that the extinction coefficient, determined by reaction of MDI with excess free glycine, is equivalent to that in proteins was addressed by determination of the extinction coefficient of the monofunctional isocyanate pT. In this way, cross-linking reactions were prevented and haptenic substitution calculated from the UV absorbance could be compared with that obtained by titration of lysine
286
Chem. Res. Toxicol., Vol. 1, No. 5, 1988
residues in the protein using TNBS reagents. Values listed in Table I11 indicated good agreement between results from the two methods. Taken together, results in Tables I1 and I11 are consistent with the interpretation that MDI reacted as a bifunctional reagent, with each of the isocyanate groups having the potential to react with a lysine residue. An alternative explanation is that some of the 59 lysine residues in human serum albumin (16) may have become unavailable for reaction with MDI as a consequence of isocyanate-induced change in protein shape, resulting in lysine residues becoming buried within a highly crosslinked structure. Occurrence of intermolecular cross-linking was readily apparent from migration patterns of the conjugates in SDS-PAGE. Results, depicted in Figure 4,clearly indicated the formation of polymeric species, thus confirming intermolecular cross-linking reactions. The kinetics of these reactions indicated rapid incorporation of 5-6 mol of MDI/mol of HSA within 1min and then a steady reaction rate of 0.16 mol of MDI/min for the next 2 h. Intramolecular cross-linking was also apparent from SDS-PAGE profiles. A time-dependent enhanced mobility of the conjugate species was observed. Although such an effect has not been described previously in the literature, three factors have been reported to influence the migration of proteins in SDS-PAGE, i.e., molecular weight, charge, and size (17). It was clear that the molecular weight of the conjugates was not the factor causing increased mobility of conjugates with time. Smaller peptides were never observed in the gels. Additionally, such an explanation would imply identical hydrolytic reaction of monomeric and polymeric conjugates with time since these species displayed identical migration kinetics (see Figure 4B). The contribution of the net charge of conjugates to their migration characteristics was investigated. In experiments not detailed here, twice the amount of SDS was added to one set of MDI-HSA conjugate samples prior to their application to the gel. This procedure assured that all protein samples (including HSA and conjugates) carried a strong net negative charge. The excess SDS was observed to cause no change in the migration pattern, indicating that an increased net negative charge on hapten-dense MDIHSA conjugates was not likely the cause of their enhanced migration in SDS-PAGE. The possibility that the altered migration of conjugates resulted from constrained size and shape of the protein caused by extensive intramolecular cross-linking was tested. DST, a classical bifunctional reagent, was selected because of its uic-glycol bond. These cross-linking bonds can be broken by simple treatment with sodium periodate. The result of periodate treatment would be a reversion of the gel migration pattern to that of the monomeric form. Reaction of DST with HSA under conditions identical with those used for reaction of MDI with HSA yielded products with somewhat enhanced migration in SDS-PAGE (Figure 5). Following treatment with sodium periodate, the DST-HSA conjugates no longer showed this altered migration. Rather, they displayed a molecular mass consistent with that of HSA, Le., 66 kDa. For control purposes, the MDI-HSA conjugates were also incubated with sodium periodate. As expected, this treatment had no effect on the migration properties of the MDI-HSA species (Figure 5). This series of experiments supported the impression of extensive intramolecular cross-linking of MDI-HSA conjugates. It was of interest to compare the molecular masses of DST-HSA and MDI-HSA conjugates which contained
J i n and Karol
approximately equivalent amounts of bound hapten. MDI-HSA conjugates showed greater change in SDSPAGE migration per haptenic substituent. The reason for this difference in SDS-PAGE migration between the conjugates is not clear but may reflect a greater bifunctional reactivity of MDI compared with DST, or a more favorable distance between functional ends in MDI (11, 18). The reaction of MDI with HSA was found to be dependent upon several reaction conditions including relative concentrations of reactants. In the industrial situation, exposure to a spill, spray, or splash of diisocyanate would be expected to result in in vivo conjugation of the diisocyanate with nucleophilic sites on biomolecules. These conjugates would function as immunologic stimulants to induce production of MDI-specific antibodies. Use of the MDI-HSA conjugates characterized in this report to detect MDI-specific antibodies in humans and an animal model is described in the following paper (21).
Acknowledgment. We thank Dr. William E. Brown, Carnegie-Mellon University, for many helpful discussions. The study was supported by NIEHS Grant ES01532. Registry No. MDI, 101-688;DST,62069-75-4;glycine, 56-40-6.
References (1) Chadwick, D. H., and Cleveland, T. H. (1981)“Isocyanates, organic”. In Encyclopedia of Chemical Technology (Grayson, M., Ed.) 3rd ed., pp 789-818,Wiley, New York, NY. (2) Bernstein, I. L. (1982)“Isocyanate-induced pulmonary diseases: a current perspective”. J. Allergy Clin. Zmmunol. 70,24-31. (3) Karol, M.H.(1986)“Respiratory effects of inhaled isocyanates”. CRC Crit. Reo. Toxicol. 16,349-379. (4) Karol, M.H. (1981)“Survey of industrial workers for antibodies to toluene diisocyanate”. JOM, J . Occup. Med. 23,741-747. (5) Chang, K. C., and Karol, M. H. (1984)‘Diphenylmethane diisocyanate (MD1)-induced asthma: evaluation of the immunologic responses and application of an animal model of isocyanate sensitivity”. Clin. Allergy 14,329-339. (6) Baur, X.,Dewair, M., and Fruhmann, G. (1984)”Detection of immunologically sensitized isocyanate workers by RAST and intracutaneous skin testing”. J.Allergy Clin. Immunol. 73,610-618. (7) Karol, M. H., Thorne, P. S., and Hillebrand, J. A. (1987)“The immune response as a biological indicator of exposure”. In Occupational and Enuironmental Chemical Hazards (Foa, V., Emmett, E. A., Maroni, M., and Colombi, A., Eds.) pp 86-90,Horwood, Chichester, England. (8) Bascom, R., Kennedy, T. P., Levitz, D., and Zeiss, C. R. (1985) “Specific bronchoalveolar lavage IgG antibody in hypersensitivity pneumonitis from diphenylmethane diisocyanate”. Am. Reu. Respir. Dis. 131,463-465. (9) Laemmli, U. K.(1970)“Cleavage of structural proteins during the assembly of the head of bacteriophage T,”. Nature (London) 227,680-688. (10) Snyder, S. L., and Sobocinski, P. Z. (1975)“An improved 2,4,6-trinitrobenzenesulfonicacid method for the determination of amines”. Anal. Biochem. 64,284-288. (11) Smith, R. J.,Capaldi, R. A., Muchmore, D., and Dahlquist, F. (1978) “Cross-linking of ubiquinone cytochrome c reductase (complex 111) with periodate-cleavable bifunctional reagents”. Biochemistry 17,3719-3723. (12) Brown, W. E., Green, A. H., Cedel, T. E., and Cairns, J. (1987) “Biochemistry of protein-isocyanate interactions: a comparison of the effects of aryl vs alkyl isocyanates”. EHP, Enuiron. Health Perspect. 72,5-11. (13) Likhite, V.,and Sehon, A. (1967)“Antigens: conjugated and synthetic antigens”. In Methods in Immunology and Zmmunochemistry (Williams. C. A.. and Chase, M. W., Eds.) -DD_ 150-166, Academic, New York. (14) Karol. M.H.. and Hauth. B. A. (1982)“Use of hexvl isocvanate antigen to detect antibodies’to hexamethylene diisocianate(HD1) in sensitized euinea Dies and in a sensitized worker”. Fundam. Appl. TozicoT. 2, 108-:13. (15) Tse. K. S.. Johnson. A.. Chan. H.. and Chan-Yeuna. M. (1985) ;A study of serum antibody activity’in workers with occupational exposure to diphenylmethane diisocyanate”. Allergy 40,314-320.
MDI-Serum Albumin Conjugates (16) Meloun, B., Moravek, L., and Kostka, V. (1975) “Complete amino acid sequence of human serum albumin”. FEBS Lett. 58, 134-137. (17) Altschul, A. M., and Evans, N. J. (1967) “Zone electrophoresis with polyacrylamide gel”. Methods Enzymol. 11,179-194. (18) Lockhart, W. L., and Smith, D. B. (1975) “Cross-linking of hemoglobin, haptoglobin, and hemoglobin-haptoglobin complex with bifunctional imidoesters”. Can. J. Biochem. 53, 861-867. (19) Brown, W. E., and Wold, F. (1973) “Alkyl isocyanates as ac-
Chem. Res. Toxicol., Vol. 1, No. 5, 1988 287 tive-site specific reagents for serine proteases. Identification of the active-site serine as the site of reaction”. Biochemistry 12, 835-837. (20) Karol, M. H., Ioset, H. H., Riley, E. J., and Alarie, Y. C. (1978) “Hapten-specificrespiratory hypersensitivity in guinea pigs”. Am. Ind. Hyg. Assoc. J. 39, 546-556. (21) Jin, R., and Karol, M. H. (1988) ‘Diisocyanate antigens that detect specific antibodies in exposed workers and guinea pigs”. Chem. Res. Tozicol. (following paper in this issue).
