Bio,'himica ¢'t Bioph.vs'ica Act,t. 1121(It~92)325-330

325

~ Iqt~2Elsevier Science Pt,blishcrs, B.V. All rights rcse~'cd 01h7-4838/92/$(15.00

BBAPRO 34222

Methionine oxidation and inactivation of a -proteinase inhibitor by Cu-'÷ and glucose Philip K. Hall and Ronald C. Roberts Unict'rsity ,~/"li iscon.~in Cemcrx and flu' Marxhfidd Mcdic, d Rt:~earch Foun,hltitm. ,$hzrshth'hl. II7 f US.4)

(Received 22 Januat3' lot)2)

Key words: aI-Antitwpsin: Free radical: Diabetes: Copper(ll) ion The effect of g l u c o . ~ / C u : " incubation on (a) pure mcthioninc oxidation. (b) the oxidation of active-site methionine in al-proteinase inhibitor (atPl) and (c) the resulting activity and structural changes of this inhibitor was investigated. While no methioninc was oxidized during a 24 day, 37°C incubation with 0.01 M EDTA and lf~l mM glucose. 64.2~'~ oxidation occurred in 6 days when t|.(Jl mM Cu z' was added to the 100 mM glucose. The first-order rate constant for oxidation in l0 mM glucose. 0.01 mM Cu '~ was 0.02iS day l . Oxidation was inhibited by catala.~, but accelerated by ascorbatc ion. The active-site mcthionyl residue of ~lPI was oxidized 71.3c/- after a 4 day incubation in 100 mM glucose, 0.01 mM Cu:' (pH 7.45), 0.1 M phosphate buffer. The elasta.~ and twpsin inhibiting activities were lowered to 3. I and I.SC~- of control samples during this incubation. The inclusion of I mM DETAPAC, a transition metal chelator, rcsultcd in a 98 + % retention of activity. Intrinsic fluorescence (350 nm excitation, 415 nm emission) of aiPi increased 576r~ over control for the sample incubated in 100 mM glucose, 0.01 mM Cu:' and SDS-PAGE revealed protein fragment molecular wcight,~ of 44.4 and 39.,'q kDa. These studies suggest that both mcthionine oxidation and free radical induced fragmentation contribute to loss of a tPl activity during glucosc/Cu :÷ incubations.

Introduction at-Proteinase inhibitor ( a t P l . also known as at-antitrypsin). MW 51 000. is a major serine proteinase inhibitor of human plasma [l]. At a concentration of 25 ,,,tM it is responsible for over 90% of the plasma neutrophii elastase inhibiting capacity [2]. a t P l is particularly ~nsitive to oxidation. It can be easily inactivated by oxidation of a single methionine residue. Met-358. at the PI position of the proteinase binding site [3] by various chemical oxidants: iV-chlorosuccinimide. ozone [2] or hydrogen peroxide with Cu -'+ ions [4]. In a recent report by Dean [5] free radicals generated by a Fcnton reaction, H : O . _ / C u "-~. were shown to inactivate both elastase and a'~Pl, in addition, phago-

Abbreviations: a tPl. t~l-proteinase inhibitor: SDS-PAGE. sodium dode~.3'l sulfate polyacD'lamide gel eleclrophoresis: DETAPAC. diethylcnetriaminepentaacetic acid: cb--DDP, cis-dichlorodiamnfineplatinum(ll): EIA. elastase inhibitoD"activitv; "rlA. tD'psin inhibitoD" activity: OH °. hydrox3l radical: tlO.:d, h.vdtoperox3'lradicaL. Correspondence: P.K. Hall. Marshfield Medical Research Foundation. 10110North Oak Avenue. Marshfield. WI 5444t). USA.

cytes upon stimulatior have been shown to oxidize at Pl by releasing myeloperoxidase and reactive free radical oxTgen species such as superoxide [6]. Scislowski [7] has demonstrated the oxidation of methioninc to methionine sulfoxidc by chemically or enzymatically generated ox.'ygen free radicals. Recent studies have described superoxide free radical generation during non-end, marie glycation of proteins. Mullarkey et al. demonstrated a 50-fold increase in the rate of free radical production following a 2 week incubation of ribonuclease A in 500 mM glucose. p h o s p h a t e / E D T A buffer (pH 7.8) [8]. The presence of superoxide was confirmed by electron paramagnetic resonance. Earlier. GilleD' [9] and Sakurai [10] utilized the reducticm of ferri~'tochrome c to quantitate superoxide production during the glycation of human serum albumin, in all three of these studies free radicals were generated during protein glycation in the absence of added Cu -'÷ ions. In a series of papers by Wolff and Hunt. however, there is evidence that the more reactive hydrox.'yl radical is formed by a C u ' * catalyzed autoxidatior, of glucose [ l l - 1 3 ] . In addition, fragmentation of bovine serum albumin was detected following an 8 day incubation with 25 mM glucose. 0.1 mM Cu :"

326 (pH 7.4) phosphate buffer. The inclusion of 1 mM DETAPAC, a (2 + ) metal chelator, prevented hydroxyl formation and protein fragmentation. Several investigations have found lower serum a~P! activity in persons with diabetes suggesting an inverse correlation between elevated serum glucose and a~Pi activity [14-16]. The present study was undertaken to address the following questions regarding a~ PI modifications during glucose incubation: (a) can methionine be oxidized by free radicals generated during glucose autoxidation; (b) is the activity of atPI affected by glucose and Cu 2+ ion incubation; (c) is the active-site methionine of a tPl oxidized; and (d) do a tPI molecules undergo fragmentation? Materials and Methods

Human a~P! (purity greater than 95% by SDSPAGE), was obtained from Athens Research and Technology (Athens, GAl. D-methionine. catalase (thymol-free), porcine elastase and porcine trypsin were purchased from Sigma (St. Louis, MO). Methionine sulfoxide, DETAPAC and cis-DDP were purchased from Aldrich (Milwaukee, WI). Methionine (50 nmol/ml) and alP! (0.5 mg/ml) were incubated from 3 to 24 days at 37°C with 0.1 M Na2HPO 4 or 0.1 M Mops (pH 7.45) buffer containing 0.02% NaN 3 and 0 to 100 mM glucose a n d / o r 0 to 0.1 mM copper (II) sulfate. No evidence of insoluble Cu3(PO4) 2 was observed for buffers containing a maximum of 0.01 mM Cu z+ in the phosphate buffer at 37°C. The concentration of Cu 2÷ was confirmed by atomic absorption. In certain experiments zinc sulfate, iron(Ill sulfate or 1 mM DETAPAC were substituted for Cu-' ÷ ions. All incubations were performed in sealed polystyrene tubes. The tubes were weighed before and after the incubation period to check for vapor loss. For timed studies aliquots were withdrawn at specified intervals and immediately analyzed. Following incubation SDS-PAGE was performed using a Phast System (Pharmacia LKB, Uppsala, Sweden) with 20% gels to detect protein fragmentation or evidence of enzyme proteolysis during incubation. Gels were scanned with a Shimadzu CS-9000 scanner (Shimadzu, Kyoto, Japan). Methionine and methionine sulfoxide were measured with a Beckman 6300 amino acid analyzer.

0.001 M HCi, and 100 #1 of a i P l (0.036 m g / m i in 0.2 M Tris) (pH 8.0). After a 1 min incubation in a 25.ff'C water bath, 20 #i of elastase substrate (Suc-Ala-AlaAla-p-nitroanilide, 30 m g / m l in dimethyisulfoxide, Sigma) or trypsin substrate (benzoyI-Arg-p-nitroanilide, Sigma) was added and the absorbance at 405 nm was recorded continuously for 2 min. The a~ Pi activity was expressed as the difference in the slopes of the linear recordings (r = 0.999) of the elastase or trypsin alone and that with the inhibitor added. Comparison activities are expressed as a percent of control samples incubated without glucose or Cu 2+ added. The quantity of control sample was selected to achieve a 50% decrease in slope, in separate experiments, the activities of elastase and trypsin alone were round to be uninhibited by the presence of 1 mM DETAPAC, 0.01 mM Cu 2+ or 100 mM glucose. Determination of actit'e-site methionine To assess whether the active-site methionine of a~ P! ha5 been oxidized by free radicals generated during glucose autoxidation, the procedure of Gonias et a!. [ 18] was employed using cis-DDP to selectively protect the active-site methionine. Samples containing 10.0 nmol a~Pi which had been incubated 4 days at 37°(2 in 0.1 M sodium phosphate, 0.02% sodium azide, (pH 7.45) buffer (control) or with added 100 mM glucose or 100 mM glucose and 0.01 mM Cu z÷ were dialyzed overnight at 4°C against 20 mM Tris, 0.1 M NaCI, 0.01% Triton-Xl00 (pH 7.45) to remove glucose and Cu 2÷ ion and then incubated 4 h at 37°C with 1.67 mM cis-DDP to protect the active-site methionine which had not been oxidized during the initial incubation. Following overnight dialysis against the same Tris buffer the unprotected methionyl residues were oxidized with 6% H 2 0 z for 15 rain at room temperature and the samples were transferred to polypropylene tubes and freeze-dried, l'he dried samples were dissolved in 200 p.l of 4 M KOH, flushed with Nz, and hydrolyzed for 16 h at 105°C. Methionine, which is not converted to the suifoxide by alkaline hydrolysis, was determined with a Beckman 6300 amino acid analyzer after lowering the pH to a value of 2.5. An internal standard, S-2-aminoethyl-t.-cysteine, was used to correct for any sample losses during transfers. Results

Measurement of t~t PI actit'ity The elastase and trypsin inhibitory activities of incubated alPl samples were determined at 25.0°C in triplicate by a modification of Beatty [17] using a thermostatically controlled Varian DMS300 spectrophotometer equipped with a kinetics module. To a 1.5 ml polystyrene euvette was added 1.00 ml of 0.05 M Tris/0.05 M NaCI (pH 8.0) buffer, 100 #! of porcine elastase 0.05 mg/ml or porcine trypsin 0.13 mg/ml in

Methionine oxidation Significant oxidation of methionine to methionine sulfoxide occurred with pure methionine samples incubated 4 days at 37°C in the presence of glucose and 0.01 mM Cu 2+ (Fig. 1). With 50 mM glucose 51.5% of the methionine was oxidized in this time period. A control sample incubated under identical conditions without glucose yielded no methionine sulfoxide. Also,

327 18-m

r8

i

7

16"

so

W

6~ m 14"

,.J

s~

Z

9. .v-

t~d Z

-4~_ -]

-3 ,.P,

o 10"

~ z~_ m

3o -.

20

-1

~\

0 0

3

6

' g

. . . . .

12

15

";~ .......... , - ~ - -

18

21

--

24

DAYS INCUBATED mM GLUCOSE

Fig. !. Methionine oxidation at various glucose concentrations. Samples were incubated 4 days at 37"C with (}.1 M .sodium phosphate. 0.02,~. NaN a (pH 7.45) buffer containing 0.01 m M C u z* and the indicated molarity of glucose. • repre~nts nmol methionine (left axis); • indicates nmol methionine sulfoxide (right axis).

no oxidation was observed with the inclusion of 1 mM DETAPAC in the incubation mixture with 50 mM glucose. The substitution of 0.01 mM Zn -'+ or W.01 mM Fe z+ for Cu :+ in the 50 mM glucose solution resulted in methionine values of 90.6 and 9W.8% of the control. In a separate experiment conducted to determine whether the presence of (I.02% NaN 3 affected the rate of oxidation, 50.0 nmoi/ml of methionine was incubated 7 days at 37°C in 0.1 M Na_,HPO 4 (pH 7.45), 0.01 M Cu z+ buffer containing (a) 0.02% N a N 3 (control); (b) W.02% NaN 3 and 5W.() mM glucose; (c) IWW U / m l penicillin/streptomycin and 5().0 mM glucose. Following incubation the control contained 48.1 + 1.6 nmol/rnl methionine while the concentration of methionine in solutions (b) and (c) was 25.6 + 1.1 nmol/ml (53.2% control) and 28.8 + 1.5 nmol/ml (59.8% control), respectively. Since the 0.02% NaN 3 did not inhibit the rate of oxidation, it was used in all buffers to prevent bacteria growth. As illustrated in Fig. 2, the oxidation was markedly dependent on Cu z+ concentration for methionine samples incubated up to 24 days with 100 mM glucose. Although a control sample containing 0.WI M E D T A and 10W mM glucose showed no methionine oxidation during this period, the methionine sample containing 0.01 mM Cu z+ was 64.2% oxidized in 6 days. A slight difference in the rate of oxidation over a 14 day period (Table I) was detectable for methionine samples incubated at 37°C with phosphate vs. Mops buffers (pH 7.45) in the presence of 0.01 mM Cu -'÷ and low glucose molarities. With 10 mM glucose the pseudo-first-order rate constant for methionine oxidation in 0.1 M phosphate buffer, k =0.0218 da -s, was 38.8% higher than in 0.1 M Mops. At 25 mM glucose

Fig. 2. Methionine oxidation at various Cu-" * concentrations. Methionine samples. 50 n m o l / m i , were incubated 24 days al 37°C wilh 0.1 M Mops, 0.02¢~ N a N 3 ( p l l 7.45) buffer containing i(X) m M glucose and the following: • {l.01 M E D T A , z~ ~.dashed line) 0.0ell m M Ca-' ". o (dashed line) 0.01 m M Cu -'+. o (dotted line) 0.1 m M Cu 2 +.

the phosphate-buffered rate constant more than doubled k = 0.0498 da - ~. The results of methionine incubations in the presence of antioxidants are given in Table 1. Although the control sample containing 50.0 nmol methionine, but no glucose or Cu -'+ showed no oxidative loss over 18 days, only 65.9% methionine remained in the sample incubated with 10 mM glucose. The addition of 0.01

TABLE I Methionine oxidation Rates in varkms buffers " Buffer Glucose 0.1 M (pH 7.45), 0.01 m M Cu 2'

(mM)

Pseudo-first-order rate constant (da - i )

Correlation coefficient, r

MOPS N a 2H P O 4 Na 2 HPO.~

10 10 25

0,0157 0,0218 0,0498

0.990 0.996 0.999

Incubation with inhibitors h Glucose Ca" (mM)

(mM)

10

0.01 0.01 0.01 0.01

Ill 10 10 10

Inhibitor

Methionine (,% control)

5(111 U catalase 50 m M cy.steine 50 m M a.~orbate

65.9 ± 2.6 38.6 ± 1.9 611.0±3.1 32.4±3.6 0.0 _+3.2

~' Rates of methionine oxidation in 0.1 M Mops or Na211PO4 (pH 7.45) buffers containing 0.01 m M Cu 2 . . 0.02% NaN~ and the indicated concentration of glbcose. Methioninc samples. ~)

n m o l / m L were incubated 14 days at 37°C. Aliquots were withdrawn at 48 h intervals and analyzed for methionine. h Methionine samples. 50 nmol/ml, were incubated 18 days at 37~C in 0.1 M N a 2 H P O 4 (p 7.45). 0.02f~ NaN 3 buffers containing 0 m M (control) or l0 m M glucose and the indicated concentration of Cu z• and inhibitor.

328 mM Cu -'÷ to the incubation mixture reduced the available methionine to 38.6% of the control. Some protection, 60% of the control, resulted from the inclusion of 500 U catalase (a scavenger of H2Oz) in the incubation mixture. However, the inclusion of 50 mM cysteine or sodium ascorbate did not prevent oxidation.

10C",~

i?....~=_ ..... _-. 90 ~ "'~. . . . . . . . ".,.

80 ~ •3 O

70

z

60

o

The result of active-site methionine analysis in atPl following a 4 day incubation is shown in Table 1I. With no added Cu -'+ the sample incubated in 100 mM glucose had 71% of the active-site methionine of the control. The sample incubated with 0.01 mM Cu -~÷ and 100 mM glucose contained only 28% of the control active-site methionine.

alP! acticiO' Both the elastase (EIA) and trypsin (TIA) inhibiting activities of oetP! are shown in Fig. 3 as a function of glucose molarity following a 4 day incubation in 0.1 M phosphate, 0.01% NaN.~, (pH 7.45) buffer. The activities are presented as a percent of control samples incubated without glucose. In the absence of added Cu z÷, 89.9 + 0.9% of the EIA and 79.5 + 1.8% of the TIA are retained after incubation in 100 mM glucose. These activities are reduced to 3.1 5: 0.5% and 1.5 + 2.7%, rcspectively, with the inclusion of 0.01 mM Cu 2÷ in the 100 mM glucose buffer. Also, considerable loss of activity occurred to 61.4 + 0,8% and 57.0 + 0.9% of control, for samples incubated in the presence of 10 mM glucose 0.01 mM Cu -'+. The influence of Fe z+, DETAPAC and antioxidants

'.-..,. ............................................

.-

't, ,;~

50.

ActiL'e-site methionine #z at Pi

,.i

\-,,

t-- 40 -~

~ , \, "-,