THE JOURNAL OF B~OLOCICAL CHEMIWRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 265, No. 25, Issue of September 5, pp. 14911-14916,199O Prmted in U.S.A.

Mechanism-based Inactivation of Leukotriene A4 Hydrolase during Leukotriene B4 Formation by Human Erythrocytes*

(Received for publication, May 22, 1990)

Lars OrningS, David A. Jones, and F. A. Fitzpatrick From the Department of Pharmacology C-236, University of Colorado Health Sciences Center, Denver, Colorado 80262

Evidence is presented in support of a mechanism- based (suicide) inactivation of leukotriene Ad hydro- lyase in intact human erythrocytes by leukotriene A4 and leukotriene A, methyl ester. Loss of enzymatic activity, accompanying leukotriene B4 formation, was proportional to the substrate concentration. Inactiva- tion was directly related to the amount of leukotriene B4 formation: for several, different experimental pro- tocols 50% loss of hydrolase activity corresponded with formation of 10.3 f 2.1 PM leukotriene B4. The time course of inactivation was pseudo-first order and obeyed saturation kinetics. Apparent inactivation (KI) and first-order rate (ki) constants for leukotriene A4 were 28 PM and 0.35 min-‘, respectively. Leukotriene A4 methyl ester was also a site-directed inactivator with a similar KI = 25 MM and a ki = 0.1 min-‘. For single incubations substrate instability limited the ex- tent of inactivation to 50% of the initial enzyme activ- ity. Following multiple, consecutive incubations with leukotriene A4 this increased and approached SO-90%; however, a residual activity of lo-20% suggested that a pool of enzyme was not susceptible to inactivation. Recovery of enzymatic activity, following inactiva- tion, was negligible in intact erythrocytes and isolated enzyme. A single radiolabeled protein, corresponding to leukotriene A., hydrolase, was detected by electro- phoretic analysis of the incubation between [3H]leu- kotriene A4 and erythrocytes, or partially purified en- zyme. Incorporation of [3H]leukotriene A4 methyl ester into enzyme was linearly related to its inactivation: 191 f 5 pmol incorporated corresponded to 10% loss of activity. Results conform to criteria for a mecha- nism-based inactivation, in which leukotriene A4 par- ticipates in two parallel processes, one leading to leu- kotriene B4 formation, the other to “suicide” inactiva- tion of leukotriene A4 hydrolase in intact erythrocytes. The specific, rather than indiscriminate nature of this process has implications for the regulation of cellular leukotriene B4 formation. It may also afford a basis to monitor transcellular biosynthesis of leukotriene B, in vivo.

Leukotriene (LT)’ Aq hydrolase isolated from erythrocytes, neutrophils, or lung is inactivated by its endogenous substrate

* This work was supported by Grant ROlAI26730 awarded by the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC. Section 1734 solely to indicate this fact.

$ Supported, in part, by a grant from the Swedish Medical Research Council.

’ The abbreviations used are: LT, leukotriene; BSA, bovine serum albumin; HPLC, high performance liquid chromatography; PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate.

or substrate analogs (l-8). Although termed “suicide” inacti- vation (l-4), it is uncertain whether the loss of activity originates from a specific, mechanism-based process or indis- criminate reactions between LTA, and nucleophilic substit- uents on the enzyme. This distinction is vital to understand its regulatory or pathological implications for LTB, biosyn- thesis and LTA, disposition. Mechanism-based inactivation is best represented by two competing processes, partitioning substrate between inactivation and turnover (9-11) (Scheme 1). Criteria implicit to this model include: (i) proportionality between inactivation and product formation, (ii) time-de- pendent loss of activity, (iii) saturation kinetics, (iv) irrevers- ibility, (v) stoichiometric proportionality between covalent modification and loss of enzyme activity.

r k,., --f LTB, + Enzyme

LTAI + Enzyme + [LTAI - Enzyme]*

KJKI 1 L k; + Inactive Enzyme

SCHEME 1. Mechanism-based inactivation of LTA, hydrolase

Using this scheme as a basis, we investigated the inactivation of LTA, hydrolase within intact erythrocytes. We chose eryth- rocytes for several reasons. First, this system may be an unusual example of suicide inactivation with constitutive metabolites of human cells. Second, compared with other cells containing LTA, hydrolase, erythrocytes facilitate experi- ments because de nouo enzyme synthesis is negligible. Third, erythrocytes generate LTB, via transcellular biosynthesis, in vitro (5, 12, 13), and an accompanying inactivation could afford an approach to detect this phenomenon in Go. Our results indicate that mechanism-based inactivation of LTA, hydrolase occurs in intact erythrocytes.

EXPERIMENTAL PROCEDURES

Materials-LTA, methyl ester and LTB, (Cayman Co.), (14,15- “HILTA, methyl ester (1.85 TBq/mmol) (Du Pont-New England Nuclear), prostaglandin Bs, BSA, microcrystalline cellulose and LY- cellulose (Sigma), ethyl acetate, acetone, acetonitrile, and methanol of HPLC grade were used. Lithium salts of LTA, and [14,X-“H] LTA, were prepared by saponification of their methyl esters with lithium hydroxide (14, 15). Human erythrocytes in 0.9% (w/v) NaCl were purified by removing neutrophils and platelets as described (16). LTA, hydrolase isolated from lvsed human ervthrocvtes was nurified loo-fold by ammonium sulfateVfractionation ?40-70”%) and Ehroma- tographv on DEAE-Senhacel and Mono Q HR lo/10 (1).

>n&&tions-Soluti&s of LTAs Li-sait were &aio;ated to dry- ness, dissolved in 0.01-0.20 ml of 0.01 M, pH 7.4, phosphate buffer with BSA (5 mg/ml), and incubated for 10 min at 22 “C with eryth- rocytes (2 x 10’ cells/ml, 1 mg of BSA/ml). Incubations were termi- nated by addition of three volumes of buffer and centrifugation at 13,000 x g for 5 s (Eppendorf microcentrifuge). The supernatant fluid conxaining LTB, and non-enzymatic hydration products was re- moved, cells were washed twice with buffer containing BSA, and

14911

14912 Suicide Inactivation of LTB, Formation by Erythrocytes

supernatants were combined. Prostaglandin B, (0.5 pg), a quantitative internal standard, was added and the acidified (pH 3) sample was extracted three times with two volumes of ethyl acetate. The sedi- mented erythrocytes were resuspended to the original volume and their remaining LTA, hydrolase activity was determined by incubat- ing, again, with 20 fiM LTA,, unless specified differently. Substrate concentrations were varied in the initial incubation; in certain exper- iments LTA, methyl ester was substituted for LTA,.

To correlate time dependence of product formation and hydrolase inactivation, erythrocytes were incubated at 22 “C as described, with and without BSA. At intervals from 1 to 30 min, samples (0.10 ml) were quenched with three volumes of BSA-containing buffer. After centrifugation, LTB, in the supernatant and hydrolase activity re- maining in the erythrocytes were determined as described. The ki- netics, from 1 to 30 min, were measured with lo-100 ).LM LTA, and LTA, methyl ester to establish concentration dependence. The inac- tivation rate was determined for each substrate concentration and the first order rate constants (koba) and half-inactivation time (t& of the enzyme were calculated. Data were analyzed by plotting ti/z versus reciprocal substrate concentration and k+ versus substrate concen- tration.

Certain experiments involved multiple, consecutive incubations of LTA, with erythrocytes to establish the maximal extent of inactiva- tion. 20 pM LTA, was mixed with erythrocytes, as above, for 15 successive incubations at lo-min intervals. LTB4 and residual sub- strate were removed by washing prior to each, successive addition of LTA,. Corresponding controls were performed to determine proce- dural losses of enzyme activity derived solely from washing and resuspension. After 15 incubations these were ~15% of the initial activity. Data have been corrected accordingly. Cumulative and in- cremental LTB, formation and hydrolase activity remaining at each incubation have been calculated.

For all experiments LTB, was quantified by reverse phase-HPLC. The ethyl acetate extracts were evaporated, and LTB, was dissolved in 100 ~1 of mobile phase. Chromatographic analyses were performed on Cln columns (4.6 x 25 cm) (Beckman Instruments Inc.) eluted with CH~CN/CH~OH/H~O/CH~COOH (36:24:40:0.1) at 1 ml/min. Solutions of LTA, and LTB, standards were auantified bv UV spec- troscopy using extinction coefficients of 40,006 and 50,000 M-r cm-‘, respectively. Mathematical expressions used in the text were derived by non-linear regression analysis using the program GraphPad. Re- sults are given as mean + S.D.

Irreversible Labeling of Enzyme with [3H]LTA4-Partially purified LTA, hvdrolase (120 ua. 50 ul) was incubated for 10 min with 0 or 100 ;M LTA,. Enzyme was then incubated with 20 PM [3H]LTAs 3.7 kBq, or [3H]LTA, methyl ester for 10 min. Reactions were quenched with 50 ~1 of 0.3% acetic acid, and samples were extracted four times with 0.4 ml of ethyl acetate to remove products. The protein remain- ing in the aqueous phase was mixed with 50 ~1 of 0.05 M Tris, pH 6.8, containing 10% alvcerol, 4% (w/v) SDS, 0.15 M B-mercaptoethanol. Samples were boiled for 2 min.and analyzed by SDS-PAGE on a 10% gel with a 5% stacking gel (17). Radiolabeled proteins were detected by autofluorography with Kodak XAR-5 film.

In a separate experiment, isolated enzyme was incubated for 10 min with O-400 pM [3H]LTA, methyl ester. Partially purified enzyme (100 ua) dissolved in 0.01 M Tris (50 ~1) was added to [3H]LTA, I methyl ester (0.3-2 TBq/mol). After 10 min at room temperature, a part of the sample was quenched with 150 ~1 of 12% trichloroacetic acid, held on ice for 10 min, centrifuged at 13,000 x g for 15 min at 4 ‘C; and washed twice with acetone. A corresponding sample was assayed for remaining hydrolase activity as described above. Washed, precipitated protein was suspended in a liquid scintillation mixture and incorporation of radioactivity was determined by p scintillation spectroscopy.

Erythrocytes were incubated with 16 $tM 13H]LTA4 (3.7 kBq) and washed twice with 0.9% (w/v) NaCl containing BSA 5 mg/ml, then three times with 0.9% NaCl alone to remove BSA. The sedimented erythrocytes were lysed hypotonically; samples were suspended in 100 ~1 of polyacrylamide gel loading buffer and analyzed by SDS- PAGE and autofluorography as described above. Control erythrocytes were compared with erythrocytes which had been preincubated three times with 16 pM unlabeled LTA, to detect competitive inhibition of [3H]LTA4 binding.

RESULTS

LTB, Formation and Inactivation Using Intact and Lysed Erythrocytes-Erythrocytes were incubated twice with O-30

pM LTA,. LTB, production during the second incubation declined, relative to the initial incubation, suggesting inacti- vation of LTA, hydrolase within the erythrocytes. The decline was proportional to the substrate concentration (Table I, upper panel). Results using lysed cells were similar to intact cells (Table I, lower panel), indicating that decreased LTB, formation during the second incubation originated from in- activation of LTA, hydrolase and not impaired LTA, uptake. The uptake of LTA, is not a rate-limiting step in LTB, production by erythrocytes. In either case there was no inhi- bition of enzymatic activity by LTB,.

Similar experiments were performed using a constant 20 pM LTA, in the second incubation to quantify the remaining LTA, hydrolase activity. Inactivation accompanied LTB, for- mation and both were proportional to the LTAl concentra- tion. The hydrolase activity declined exponentially, approach- ing an asymptote at 60% of the initial activity (Fig. 1). There was a linear relationship (r = 0.98, Y = -4.6X + 99.0) between the formation of LTB, (catalytic turnover) and loss of enzyme activity (Fig. 1, inset), consistent with a mechanism-based inactivation (Scheme 1). From this relationship, formation of 10.6 pM LTB, corresponds to 50% inactivation of 2 X IO9 cells/ml.

Kinetics of Inactivation-Time course experiments con- firmed that inactivation accompanied LTB, production with half-maximal values for each occurring within 2 min of catal- ysis (Fig. 2, lower panel). A plot of the logarithm of inactiva- tion versus time (Fig. 2, upper right) showed a rapid pseudo- first order phase. Consistent with data from Fig. 1 there was a linear relationship between loss of hydrolase activity and LTB, formation (r = 0.99, Y = -6.5X + 100.5). From this equation, formation of 7.8 PM LTB, corresponds to 50% inactivation (Fig. 2, upper left). Inactivation by LTA, methyl ester was similar (data not shown). The pseudo-first order

TABLE I Inactivation of LTB, formation: comparison between intact

and lysed erythrocytes Inactivation of LTB, formation in intact and lysed erythrocytes.

Intact erythrocytes (0.5 ml, 2 x log/ml) were incubated at 37 “C for 15 min with O-30 pM LTA, in pH 7.4 phosphate buffer, 0.9% (w/v) NaCl, BSA (5 mg/ml). Cells were centrifuged 5 s at 13,000 X g, and the supernatant containing LTB, was removed. Cells were washed twice, resuspended, and incubated again with O-30 pM LTAr. The upper panel depicts LTB, formed during each successive incubation. The experiment was repeated using hypotonically lysed erythrocytes.

LTB, formation by intact erythrocytes

LTA, cont. LT.4, hydrolase Initial Second remaining % Initial

incubation incubation

MM PM PM 0 0 0 100 2.5 2.0 1.0 50 5 3.3 1.5 45

10 4.8 1.9 40 20 5.1 2.0 39 30 5.2 1.8 35

LTBI formation by lysed erythrocytes

Initial Second incubation incubation

% Initial

w PM PM

0 0 0 100 2.5 0.9 0.6 66 5 1.3 0.6 46

10 1.8 0.8 44 20 2.2 0.7 32 30 2.4 0.7 29

Suicide Inactivation of LTB, Formation by Erythrocytes 14913

1 . i

-( 0 IO 20

Substrate Concentration

uM LTA4

30

FIG. 1. Relationship between LTAl concentration, LTB, production, and inactivation of LTA, hydrolase in erythro- cytes. Erythrocytes (0.5 ml, 2 x 109/ml) were incubated with O-30 pM LTA,. LTB, formation (M) and hydrolase activity remaining after the initial incubation (O---0) were determined. Enzyme in- activation accompanied product formation. LTA, hydrolase activity declined exponentially (Y = 38.6e-0.09X + 60.8, r = 0.98) as a function of LTA, concentration, and linearly (Y = -4.6X + 99, r = 0.98) as a function of LTB, formation during the initial incubation (inset, U---U).

uM LT64

60.

FIG. 2. Time dependence of inactivation. Erythrocytes were incubated with 20 gM LTA,; aliquots (100 ~1) were removed from 0 to 30 min for determination of LTB, formation (M) and hydro- lase activity remaining (o--O). Enzyme inactivation accompanied product formation (lower panel). LTA, hydrolase activity declined exponentially ( Y = 29.6e-0,“X + 67.9, r = 0.96) as a function of time (upper right panel), and linearly (Y = -6.5X + 100.5, r = 0.99) as a function of LTB, production (upper left panel).

phase of inactivation was concentration dependent for lo- 100 pM LTA, or LTA, methyl ester. Plots of t1j2 uersus the reciprocal of substrate concentration were linear (Fig. 3, upper panel), and plots of kObs versus substrate concentration (Fig. 3, lower panel) were hyperbolic, demonstrating that inactiva- tion was a saturable process. Apparent binding constants for inactivation (KI) and first order rate constants (ki) were 28 pM and 0.35 min-’ for LTA, and 25 pM and 0.1 min-’ for LTA, methyl ester. In the same protocol, the apparent K,,, for LTB, formation was 29 PM LTA,. Similar values for K, suggest that LTA, and LTA, methyl ester bind at a common site; differences in ki suggest that LTA, methyl ester inacti- vates more slowly. LTB, methyl ester formation was unde- tectable, indicating that k, for this substance is large in relation to k,.,. For erythrocytes from several donors K, and K,,, values were always similar and within 10% of each other. K, values of 38, 15, 29, and 9 pM were paired with K, values of 37,13,28, and 8 pM.

Multipte, Consecutive Additions of LTA,-Inactivation of erythrocyte LTA, hydrolase by a single incubation with LTA, approached, but seldom exceeded 50% of the initial activity (Fig. 1). To determine if this was a limit of residual activity, erythrocytes were incubated with consecutive doses of 20 FM LTA, at lo-min intervals. After each incubation LTB, was removed and quantified. Cumulative LTB, production and

-0.04 0 0.04 0.00

“M-1

FIG. 3. Concentration dependence of inactivation kinetics. The time for half-inactivation (t& is a linear function of the recip- rocal of concentration for lo-100 pM LTA, (U) or LTA, methyl ester (M) (upperpanel). The observed first order rate constants for inactivation (&,J are a hyperbolic function of LTA, or LTA, methyl ester concentration (lower panel). These data refer to the pseudo-first order phase of inactivation. K, = 28 PM and k, = 0.35 min-’ for LTA,; K, = 25 pM and k, = 0.1 min-’ for LTA, methyl ester. The K,,, was 29 PM in this experiment.

14914 Suicide Inactivation of LTB,

hydrolase activity remaining after each incubation are de- picted in Fig. 4, upper panel. Enzyme activity decreased ex- ponentially (r = 0.99, Y = 81.9e-“““X + 17.6), reaching an asymptote by the eighth incubation. The maximum inactiva- tion after 15 additions was typically 85%. Procedural losses of LTA, hydrolase activity due to the washing and resuspen- sion steps were less than lo-15% after 15 additions, and results have been corrected accordingly. Plots of remaining hydrolase activity uersus LTB, formation were biphasic (Fig. 4, lower panel). Consistent with results from Figs. 1 and 2, inactivation was linearly related to LTB, formation during the first eight incubations (r = 0.97, Y = -4.1X + 91.2); 9.9 FM LTB, corresponds to 50% loss of activity. After eight incubations, there was no further change. Results from a separate experiment show that the inactivation phase is par- allel for 20 or 100 pM LTA,. Data plotted as incremental LTB, production uersus cumulative LTB, production depict this (Fig. 4, inset, lower panel). The slopes for these lines are -0.24 +- 0.2 and -0.25 f 0.2, indicating that LTB, production by 2 x 10’ cells is reduced by 1 nmol for every 4 nmol formed. A 50% loss of LTA, hydrolase activity corresponds to forma-

30-

d - E

; 20- z .-

2 - c

d s IO_ 3 “0 Ii

-100

0 5 IO 15 20 25

Product Farmatm uM LTS 4

FIG. 4. Inactivation of LTA, hydrolase during multiple, consecutive incubations of erythrocytes with LTA,. Cumula- tive LTB, formation (M) and enzyme activity remaining after each incubation (o---O) were determined for 15 consecutive incu- bations between erythrocytes and 20 pM LTA,. The remaining en- zyme activity declined exponentially (Y = 81.9e-“JfiX + 17.6, r = 0.99) as a function of the number of incubations, reaching an asymptote at 17% of the initial activity (upper panel). Hydrolase activity declined linearly ( Y = -4.3X + 92.5, r = 0.98) as a function of LTBr formation, for the first eight incubations. No further inactivation was evident after this (a--O, lower panel). The inset depicts incremental LTB, formation as a function of cumulative LTB, formation during incu- bations with 20 pM (A-A) and 100 pM (U) LTA,. The slopes for the decline in incremental LTB, formation were parallel.

Formation by Erythrocytes

tion of 9.6 and 13.6 pM LTB, with LTA, concentrations of 20 and 100 PM, respectively.

Reversibility-Erythrocytes were incubated with three con- secutive doses of 20 pM [“HILTA and their hydrolase activity declined to 34% of the initial value. These cells were washed and retained at 4°C. Aliquots were removed daily for 3 days to determine if the inactivation was reversible. The hydrolase activity increased from 34 to 40, 50, and 47% of the original activity after 1, 2, and 3 days, respectively. The hydrolase activity of control cells was 82, 79, and 84%, respectively. Likewise, partially purified enzyme was inactivated to 24% of its initial activity. After gel filtration on PD-10 columns the value was 19% of initial LTA, hydrolase activity, indicating that no reversal had occurred.

Labeling of Enzyme-Erythrocytes incubated with 16 PM [“HILTA,, 3.7 kBq, for 10 min contained a single, radioactive protein, M, = 68,000 by SDS gel electrophoresis (Fig. 5). Pretreatment by three consecutive incubations with 16 PM unlabeled LTA, decreased the hydrolase activity to 40% of its initial level and also decreased incorporation of tritium into the protein. Results were similar for purified LTA, hydrolase (120 pg) incubated with 20 pM [“HILTA, or its methyl ester.

200 F I

92 - 7 69 - -+

0 46 - X

r' 30-

22

t

a b FIG. 5. Labeling of erythrocytes with [3H]LTA4. Autofluo-

rography of SDS-PAGE of proteins from human erythrocytes incu- bated with [“HILTA, (lane b). Preincubation of erythrocytes three times with 16 pM unlabeled LTA, inhibited the incorporation of radiolabeled ligand (lane a). Autofluorography of SDS-PAGE of partially purified LTA, hydrolase incubated with [“HILTA, or [“HI LTA, methyl ester gave similar results (data not shown). Data on the quantitative relationship between incorporation of [“HILTA, methyl ester and inactivation of enzyme are presented in Table II.

TABLE II

Relationship between covalent incorporation of [“HILTA, methyl ester (Me) and inactivation of isolated LTA, hydrolase

Data represent the picomoles of [“HILTA, methyl ester incorpo- rated into 120 pg of isolated LTA, hydrolase. From 0 to 200 pM there was a linear relationship (r = 0.99) between inactivation and amount incorporated, with a slope of 18.9 pmol incorporated/l% inactivation. At 400 pM, the greater incorporation, 31.9 pmol/l% inactivation, indicates that indiscriminate binding occurs when the concentration greatly exceeds the K,.

[‘HILTAd me EtU.plle [“HI LTA, me

inactivation incorporated

PM X inactioatron pmol

50 16.9 329 100 23.0 427 200 36.0 686 400 42.2 1343

Suicide Inactivation of LTB, Formation by Erythrocytes 14915

Furthermore, the incorporation of radioactivity into the pro- tein was quantitatively related to the degree of inactivation (Table II). For 50, 100, and 200 pM [3H]LTA, methyl ester the enzyme incorporated 195, 186, and 191 pmol/lO% inacti- vation. At 400 pM [3H]LTA4 methyl ester incorporation in- creased to 318 pmol/lO% inactivation, suggesting indiscrimi- nate binding to nucleophilic substituents of the protein.

DISCUSSION

Our results conform to criteria defining a mechanism-based inactivation process (9-11). These criteria include (i) loss of activity proportional to substrate (LTA,) concentration and product (LTB,) formation (catalysis dependent), (ii) time- dependent loss of activity, corresponding to catalysis, (iii) saturation kinetics, (iv) pseudo-first order reaction rate, (v) irreversibility, (vi) stoichiometric relationship between en- zyme modification and loss of activity.

The distinction between a specific, mechanism-based proc- ess and non-specific inactivation is notable. For instance, LTA, is an electrophile which might react with any available nucleophile, typified by glutathione (18) and methanol or water (19, 20). There is no a priori reason to assume that inactivation of LTA, hydrolase, within erythrocytes, must be specific. Isolated LTA, hydrolase from several sources is inactivated (l-8) and covalently modified (2) by its substrate; however, these investigations did not characterize the process in sufficient detail to permit conclusions about its mechanism. In one case, Ohishi et al. (6) showed that several LTA, analogs with reactive 1,2-oxido-3,5-hexadiene substituents inacti- vated LTA, hydrolase from human lung. They concluded that the inactivation proceeded via non-specific, denaturation re- actions because no enzymatic products were detected. How- ever, a conventional scheme for mechanism-based inactiva- tion is equally consistent with their results (Scheme 1). For instance, their data could be interpreted in terms of a partition ratio favoring inactivation rather than catalysis. In other words, k, > k,,, for certain substrate analogs.

With isolated enzymes and excess, reactive substrates non- specific processes may occur or even predominate (6). How- ever, SDS-PAGE confirmed that LTA+ at concentrations near its K,,,, does not react indiscriminately with cellular constituents. Only one labeled protein was detected. The quantitative relationship between inactivation and incorpo- ration of [3H]LTA4 methyl ester, from 0 to 200 FM, into purified enzymes substantiates and extends the report by Evans et al (2) who first demonstrated a covalent modification of LTA, hydrolase. Since our enzyme preparation was not homogenous the stoichiometry of the reaction was not calcu- lated. SDS-PAGE resolution of erythrocyte LTA, hydrolase at different purification stages showed that its molecular weight was 68,000-70,000. The enzyme degraded, even at 4 “C, to a component with molecular weight of 54,000-57,000. This has been observed with neutrophil LTA, hydrolase.’ This proteolysis explains the molecular weight for erythrocyte LTA, hydrolase previously reported (1).

Inactivation approached but did not exceed 50% in certain protocols (Figs. 1 and 2). Plausible explanations for this apparent limit are the instability of LTA, or product-inhibi- tion of the catalytic process. Another possibility is the exist- ence of two pools of enzyme, one susceptible to inactivation, the other not. To resolve this matter, erythrocytes were in- cubated with consecutive doses of LTA,. Since inactivation increased t.o 85% with continued replenishment of LTA, it is probable that its instability accounts for the limit to inacti-

’ 0. Radmark and J. Haeggstrom, personal communication.

vation during single incubations, even with LTA, concentra- tions (200-400 pM) which exceed the K,. However, instability of LTA, does not account for the 15% of activity persisting after 15 consecutive incubations. These data suggest the ex- istence of two pools of LTA, hydrolase, a major pool that is susceptible to mechanism-based inactivation and a minor pool that is refractory. Whether these consist of two discrete proteins or a residual activity, uniformly distributed, is under investigation. Spontaneous recovery, or de nova protein syn- thesis, is an unlikely source of the residual enzyme activity. Inactivation of LTA, hydrolase in erythrocytes was essentially irreversible with an insubstantial recovery after 1-2 days at 4 “C. Furthermore, inactivated, isolated enzyme did not re- cover after gel filtration on PD-10 columns or standing on ice for 24 h. Product inhibition occurs with many mechanism- based inactivators; however, we and others (6) have demon- strated that there is no product inhibition by LTB,.

For mechanism-based inactivation Michaelis-Menten ki- netics, if applicable, apply to both catalysis and inactivation (9-11). The pseudo-first order rate and saturable kinetics (Fig. 3) indicate that LTA, hydrolase conforms to this model. The equivalent KI values of 28 and 25 pM suggest that LTA, and LTA, methyl ester bind to a common site. These binding constants were also similar to the apparent KM = 29 pM,

obtained using the same protocol and to K,,, = 7-36 pM

reported for purified LTA, hydrolase from various sources (1, 3, 6, 7, 21, 22). Collectively, the data are consistent with, but not proof for, a common binding site for turnover and inac- tivation. Unambiguous proof of active-site specificity will require additional experiments which would be facilitated by a competitive inhibitor of the enzyme. It is interesting to note that the LTA, hydrolase can be adequately characterized within erythrocytes because the uptake of LTA, is not a rate- limiting step which would complicate the experiments.

Inactivation was rapid and showed a pseudo-first order time dependence which corresponded to the equally rapid turnover. Results were similar when BSA was included to stabilize LTA, (14); however, there was a more pronounced, slower phase of inactivation from 10 to 30 min. This phase represented the redistribution of LTA, between different binding sites on BSA and LTA, hydrolase.

Based on the correlation between inactivation and product formation, the time dependence and kinetics of inactivation, and the linear correspondence between inactivation and co- valent binding of LTA, methyl ester we conclude that a specific, mechanism-based (suicide) inactivation accompanies catalysis by LTA, hydrolase in erythrocytes. Results also suggest the existence of two pools of enzyme, one susceptible to inactivation, the other not. Due to the reactivity of LTA, and LTA, methyl ester indiscriminate attachment to various nucleophilic groups cannot be excluded as a cause for inacti- vation in experiments using high concentrations of LTA, (> K, or KI) with purified enzyme. Our results emphasize the value of characterizing enzymological phenomenon, not only with purified enzymes, but also in cellular systems. Certain points are notable in this context. First, examples of suicide inactivation involving physiologically relevant constituents of mammalian cells are rare. Most examples involve eicosanoid metabolism (23-25); few have been comprehensively charac- terized with intact cells. Second, the detection of enzymes covalently modified during the catalytic process may offer a novel approach to monitor eicosanoid formation in vivo. At- tractive features of erythrocytes in this context are their long half-life in vivo, their limited capacity for de novo protein synthesis, and their potential involvement in transcellular biosynthesis of LTA, (5). Isolation of blood containing eryth-

14916 Suicide Inactivation of LTB4 Formation by Erythrocytes

rocytes with suicide inactivated LTA, hydrolase would imply that transcellular biosynthesis occurs naturally during inflam- mation or other disorders.

REFERENCES 1. McGee, J., and Fitzpatrick, F. (1985) J. Biol. Chem. 260, 12832-

12837 2. Evans, J. F., Nathaniel, D., Zamboni, R., and Ford-Hutchinson,

A. W. (1985) J. Biol. Chem. 260, 1096610970 3. Evans, J., Dupuis, P., and Ford-Hutchinson, A. W. (1985)

Biochem. Biophys. Acta 840, 43-50 4. Nathaniel, D., Evans, J., Leblanc, Y., Leveille, C., Fitzsimmons,

B., and Ford-Hutchinson, A. W. (1985) Biochem. Biophys. Res. Commun. 131,827-835

5. McGee, J. E., and Fitzpatrick, F. A. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 1349-1353

6. Ohishi. N., Izumi, T., Minami, M., Kitamura, S., Seyama, Y., Ohkawa, S., Terao, S., Yotsumoto, H., Takaku, F., and Shimizu, T. (1987) J. Biol. Chem. 262. 10200-10205

7. Haeggstrom, J., Bergman, T.,‘Jornvall, H., and Radmark, 0. (1988) Eur. J. Biochem. 174, 717-724

8. Prescott, S. (1984) J. Biol. Chem. 259, 7615-7621 9. Walsh, C. (1984) Annu. Rev. Biochem. 53,493-535

10. Rando, R. (1984) Pharmacol. Reu. 36, 111-142 11. Fersht, A. (1985) Enzyme Structure and Function, 2nd ed., W. H.

Freeman, San Francisco 12. Marcus, A. (1988) in Inflammation: Basic Principles and Clinical

13.

14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

24. 25.

Correlates (Gallin, J. I., Goldstein, I. M., and Snyderman, R., eds) pp. 129-137, Raven Press, New York

Murphy, R., Fitzpatrick, F. A., and Maclouf, J. (1989) in Media- tors of the Inflammatory Process (Henson, P., and Murphy, R., eds) pp. 147-157, Elsevier Science Publishers B. V., Amsterdam

Fitzpatrick, F. A., Morton, D., and Wynalda, M. A. (1982) J. Biol. Chem.257,4680-4683

Carrier, D., Bogri, T., Cosentino, G. P., Guse, I., Rakhit, S., and Singh, K. (1988) Prostaglandins Leukotrienes Essent. Fatty Acids 34,27-30

Fitzpatrick, F., Liggett, W., McGee, J., Bunting, S., Morton, D., and Samuelsson, B. (1984) J. Biol. Chem. 259,11403-11407

Laemmli, U. K. (1970) Nature 227,680-685 Murphy, R., Hammarstrom, S., and Samuelsson, B. (1979) Proc.

Natl. Acad. Sci. U. S. A. 76, 4275-4279 Borgeat, P., and Samuelsson, B. (1979) J. Biol. Chem. 254,2643-

2646 Borgeat, P.. and Samuelsson, B. (1979) Proc. N&l. Acad. Sci. ZJ.

S.-A. +6;3213-3217 Radmark. 0.. Shimizu. T.. Jornvall. H.. and Samuelsson. B.

(1984) i. Blol. Ckem. ‘259, 12339-12345 Bito, H., Ohishi, N., Miki, I., Minami, M., Tanabe, T., Shimizu,

T., and Seyama, Y. (1989) J. Biochem. (Tokyo) 105,262-264 Smith, W. L., and Lands, W. E. M. (1972) Biochemistry 11,

3276-3285 Dewitt, D., and Smith, W. (1983) J. Biol. Chem. 258,3285-3293 Egan, R., Paxton, J., and Kuehl, F. (1976) J. Biol. Chem. 251,

7329-7335