THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 247, No. 12, Issue of June 25, pp. 3810-3821, 1972

Printedin U.S.A.

Covalent Labeling of the Active Site of Human Carbonic

Anhydrase B with IV-Bromoacetylacetazolamide*

(Received for publication, November 26, 1971)

SHOW-CHU C. WONG,~ STEPHEN I. KANDEL,~ MARIANNE KANDEL, AND ALLAN G. GORNALL

From the Faculty of Pharmacy and Department of Pathological Chemistry, University of Toronto, Toronto i81, Canada

SUMMARY

Two or possibly three equivalents of N-bromoacetyl- acetazolamide interact with two histidine residues located in the active site region of human carbonic anhydrase B. The most rapid reaction occurs at the l-nitrogen of a histidine residue and this is followed by a slower alkylation of the 3-nitrogen of the same histidine and the 3-nitrogen of a second histidine.

From a reaction mixture of N-bromo[14C]acetylacetazol- amide and the enzyme, a species was isolated that contained approximately 1 eq of reagent covalently bound to the l- nitrogen of a histidine (or histidines). This alkylated enzyme is active toward p-nitrophenyl acetate, but shows a pH dependence which is different from that of the native enzyme. At pH 6 the esterase activity is 9.5%, and levels off at pH 7.6 to about 35% relative to the native enzyme. The K,, of the alkylated enzyme is almost identical with that of the native enzyme. These observations indicate that the histidine that reacted fastest with the reagent does not participate in the catalysis.

Like the native enzyme alkylated human carbonic an- hydrase B reacts with bromoacetate at the j-nitrogen position of a histidine in the active site region. The partially purified dialkylated enzyme has a residual esterase activity at pH 7.6 of less than 0.8%. This raises the possibility that the histidine that reacts with bromoacetate plays a role in the catalysis.

Human carbonic anhydrase I< (carbonate hydro-lynse, EC 4.2.1 . 1) is one of t.hree major isoenzymex isolated from human erythrocytes. It is a single chain protein containing 1 zinc ion which is essential for t’he catalysis and the binding of inhibitors

* This work was supported by Grant, MA3057 and Grant MT369 from the Medical Iicsearch Council of Canada.

2 This work comprises part of a thesis submitted to the Urli- versity of Toronto, Toronto, Canada, for the degree of Doctor of Philosophy. Present, address, Deparbmemt of Biology, Brook- haven National Laboratory, I!pton, Long Island, New York 11973.

8 To whom inquiries may be addressed at the Faculty of Phar- macy, University of Toronto.

(1, 2). It catalyses the reversible hydration of CO2 (3) and certain carbonyl compounds (4), as well as the hydrolysis of esters (5) and of fluorodinitrobenzene (6). Although the rate of the catalysis is markedly different for the different substrates, the pH rate profiles for most of the known catalytic hydration or hydrolysis reactions are sigmoid with an inflection point around neutrality (4-7). This indicates that a group in the enzyme with a pK around 7 is required in the basic form for catalysis.

Several attempts have been made recently to identify this group. It is believed that the catalysis of CO:! hydration is dependent upou a water molecule bound to the zinc ion at, t,he active site (8-13). The ioni&ion of t,his water molecule and the subsequent nucleophilic attack by the resulting OH group on the carbon dioxide molecule, along with an accompanying proton transfer, would be the basis of the cat,alytic mechanism and the observed pH dependence of the reaction. It seems obvious, however, that this simple mechanism cannot account for the extremely high ca.talytic rate of COZ hydration. The side chain of a basic amino acid was postulated by Wang (14) to be hydro- gen bonded to the zinc-bound OH ion to enhanre its nucle- ophilicit,p a.nd to facilitate the protou transfer. The imidazole group of hi&dine seemed to be the most likely candidate for such a role since its pK is closest of that of all amino acids to the pK of the group found to be essential for catalysis. A histidine was also implicated in t’he catalytic mechanism of rocker and Stone (15).

-Wini& labeling of human tnrbouic anhydrase B by Bradbury (10) and Whitney et al. (16, 17) led to the alkylntion at the 3- nitrogen position of two sequentially remote histidines. Sim- larly, we have found that bromoacetazolamide specifically and covulently interacts with the 3-nitrogen of a histidine residue of bovine carbonic auhydrase B (18). The enzymes alkylated by bromoacet.ate and bromoacet,azolamide, however, retain 5 to 15% of t’he original e&erase activity of the native enzyme. This observation was taken as evidence that the hi&dines that reacted with t.hese reagents, although located in the active site region, are not essential for the catalysis. On the other hand, the enzyme alkylated by N-chloroacetylchlorothiazide is com- pletely inactive (16). It is believed, however, that this is not the result of the modification of a hi&dine, but rather is due to the presence of the bulky group covalently bound at the active site (13). The modification of all lysine residues in the bovine enzyme B (19) and in the human enzyme B (17) did not lead to

3810

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 25, 1972 S.-C. C. Wonq, S. I. Kandel, 31. Karulel, and A. G. Gornull 3811

any significant loss in the activity of these proteins. Similar results were obtained also when tetranitromethane was used to modify one tyrosine residue in the bovine enzyme and three in the human enzyme (20). Consequently, any critical role for these residues can be ruled out. Khalifah (13) recently pro- posed a mechanism for the CO2 hydrtltion reaction that does not involve any amino acid side chain.

In order to map the active site of carbonic anhydrase further, with the aim of identifying an amino acid that is directly or indirectly involved in the catalysis, we prepared N-bromoace- tylacetazolamide and studied its reaction with human carbonic anhydrase B. The results of these experiments are reported here.

EXPERIMENTAL PROCEDURE

Jf aterials and Reagents-Acetazolamide (Lot CL-6063) was supplied by Lederle Laboratories through the eourt.esp of Mr. S. B. Davis. Bromoacetic acid (Eastman) and thioglycolic acid (British Drug Houses, Toronto) were distilled under reduced pressure prior to use. Urea (British Drug Houses, Toronto) was recrystallized twice from 95% ethanol and p-nitrophenyl ace- tate (Eastman) three times from acetone-water (9:l v/r). Bromo[l-14C]acetic acid was obtained from Xew England Nuclear Corporation (23 mCi per mmole). DEAE-cellulose (Celles-D, high capacity) was purchased from Bio-Rad, Richmond, Cali- fornia. Peroxide-free tetrahydrofuran was obtained by distilling tetrahydrofuran (Fisher Reagent) from sodium borohydride. Acetone and ethyl acetate (Fisher Reagent) were kept with anhydrous potassium carbonate overnight. and then dist.illed at their respective boiling points. Other solvents were reagent grade and redistilled prior to use. Crude mixture of bovine carbonic anhydrnses was purchased from Worthington and horse blood from Woodlgn Farms, Guelph, Ontario. Human red blood cells were supplied through the courtesy of Dr. J. 1-I. Crookston from the blood bank at Toronto General Hospit,al.

Preparation of Carbonic Bnh$rasesHuman, bovine, and horse carbonic a.nhydrases were sepa,rated as previously de- scribed (18).

Zinc-free and Carbcxymethyl Human Carbonic Anhydrase B- These enzyme species were prepared according to Lindskog and blalmstriim (21) and Whit’uey et al. (17), respect’ively. Zinc analysis was carried out by atomic absorption spect,roscopy (Perkin Elmer, model 303).

Thin Layer Chrolnatography--,liinlgtical and preparative thin layer chromatography was done on precoated silica gel plates with fluorescent indicator F-254 (0.25 mm and 2 mm layer t’hickness, respectively, Merck Brinkmann Instrument (Canada) Resdale, Ontario). The solvent system was ethyl acetate- acetone-water (60:40:4). Aromatic sulfonamides were detected by short wave length ultraviolet light.

Synthesis oJ N-Acetylacetazolamide (Fig. 2)--1~.X(:et~lacetazol- amide was prepared by acetylating acetazolamide (300 mg, 1.4 mmoles) with acetic anhydride (1 ml) in the presence of roncen- trated sulfuric acid (3 drops) as a cat,alyst,. The reaction mix- ture was kept at 96-100” for 60 min, then estract.ed wit,h carbon tetrachloride (2 x 10 ml) and triturated with cold water (2 x 5 ml). The solid residue was recrystallized twice from 50yo tet’rahydrofuran. N-Acetylaretazolamide was obtained in a 47 y0 yield, m. p. 230-231”.

R-HN SozNH-R’

FIG. 1. Structure of acetazolamide (R = CH-C, R’ = H),

ic, deacetylacetazolnmide (R, R’ = H) : iv-bromoacetylacetazolamide (R = CH,--C, R’ = Br-CH-C); N-bromoacetylbromoacetazol-

! II

0 amide (R,R’ = Br-CH-C).

II CsH,O,N,Sz (264.3)

Calculated: C 27.27, H 3.05, N 21.20% Found : C 27.34, H 3.17, N 21.25%

Preparation of Bromoacetic Anhydride-Bromoacetie anhy- dride was prepared according to the method of Brauns (22). Pure bromoacetic anhydride was distilled between 145 and 148” at 20 mm, m.p. 37-39” (literature m.p. 41’) (22).

Synthesis of N-Bromoacetylacetazolamide (Fig. l)-N-Bromo- acetylacetazolamide was prepared by bromoacetylsting acet- azolamide (1.11 g, 5 mmoles) w&h bromoacetic anhydride (6.5 g, 25 mmoles) in the presence of concentrated sulfuric acid (5 drops) as a catalyst. The reaction mixture was kept at 96-100” for 50 min with occasional swirling. A semisolid mixture was pro- duced and extracted with carbon tetrachloride (3 x 10 ml) which was subsequently removed with a Pasteur pipette. The remaining precipitat.e was suspended and triturated with water at 0” (2 x 5 ml), then filtered through a fine sintered glass fun- nel. The residue was dissolved in a minimum amount of 90% tetrahydrofuran and vacuum filtered. Part of the solvent was removed on a rotary evaporator under reduced pressure until crystals formed and carbon tetrachloride (30 ml) was added to complete the precipit,ation. Aft.er filtration the precipitate was dried in a desiccator over PZ05 under vacuum prior to further purification.

Purification of N-Bromoacetylacetazolamide by Preparative Thin Layer Chromatography-The above reaction misture (80 to 100 mg) was dissolved in a minimum amount of 95y0 tetra- hydrofuran and applied to a preparative precoated silica g 1 p1at.e. The plate was developed by allowing the solvent to run to the top. Approximately 4 hours were needed for each run and 2 to 4 plates were used simultaneously. The band rorre- sponding to N-bromoacetylaretazolamide was scraped off and extracted with 95% acetone (4 x 15 ml). After centrifugat,ion the supernatant was collected and evaporated to dryness on a rotary evaporator. The residue collected from 4 plates (400 mg of reaction mixture) was suspended in 1-butanol (8 ml) and ethyl acetate (40 ml) and extracted with 0.01 N IICl (3 X 6 ml). The solvent, was evaporated on the rotary evaporator urltll heavy crystals precipitated and then 2 volumes of chloroform were added to complete the precipitation. Crystals were collected by filtrat.ion, dried and dissolved in peroxide-free tetrahydrofuran (30 ml), and filtered through a fine sintered gIass funnel to remove any insoluble contaminants. I-Butanol (5 ml) was added to the filtrate and the mixture was evaporated again to 4 to 5 ml. The crystalline mixture so obtained was kept in a freezer (-20”) for 20 min then filtered and washed with cold anhydrous acetone

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

3812 His&line Involvement in the Activity of Carbonic Anhydrase Vol. 247, No. 12

(3 ml). The crystals were dried under vacuum in a desiccator plus 4 X lop4 M p-nitrophenyl acetate in 0.075 M Tris sulfate over PZOc (141 mg, 35% yield, m.p. 224-225”, with decomposi- buffer at pH 7.6, plus 0.8% acetone, unless otherwise indicated tion). in the legends to the various tables and figures.

CGH704N&Rr (343.2)

Calculated: C 21.00, H 2.06, N 16.33, Br 23.29y0 Found : C 21.04, H 2.16, N 16.29, Br 23.13%

S@hesis of N-Bronzo[‘4C]acetyZucetu~~Zumide-Bromo[14C]ace- tic anhydride was prepared by t’he method used for unlabeled bromoncetic anhydride. Bromo[14C]acetic acid (3 mCi) was diluted with unlabeled bromoacetic acid (25 g, 180 mmoles) to give a specific activity of approximately 16.7 PCi per mmole. N-Bromo[14C]acetylacetazolamide was prepared in the same manner as unlabeled N-bromoacetylacetazolamide. The reac- tion mixture contained 550 mg (2.47 mmoles) of acetazolamide and 3.23 g (12.42 mmoles) of bromo[14C]acetic anhydride. The final product had a specific activity of 16.1 PCi per mmole and t,his was unchanged after three crystallizations.

Synthesis of N-BromoacetylbromoacetazoEamide (Fig. l)-This compound was prepared by the procedure used for N-bromo- acetylacetazolamide. A mixture of deacetylacetazolamide (Compound I, Reference 23) (180 mg, 1.0 mmole), bromoacetic anhydride (2.0 g, 7.7 mmoles), and concentrated sulfuric acid (3 drops) was kept at 966100’ for 60 min. The cold semisolid mixture was extracted with carbon tetrachloride (3 X 5 ml) to remove unreacted bromoacetic anhydride, then cold water (5 ml) was added to initiate crystal formation. The crystals so formed were filtered and washed successively with cold water (5 ml) and carbon tetrachloride (5 ml), then dissolved in ethyl acetate (15 ml). This solution was estraoted with 0.1 N HCl (2 X 3 ml) to remove unreact.ed deacetylacet.azolamide and any sulfuric acid, then evaporated to dryness under reduced pressure. Purifica- tion was achieved by dissolving the residue in tetrahydrofuran (2 ml) followed by addition of carbon tetrachloride (8 ml). This procedure was repeated twice. A yield of 170 mg of N-bromo- acetglbromoacetazolamide (407,) was obtained (m.p. 204-205’, with decomposition).

Preparation and Separation of Human Carbonic Anhydrase B Alkylated with N-Bromoacetylacetazolamide (Alkylated Enzyme)- Human enzyme B (300 mg, 10 pmoles) was reacted with lo-fold molar excess (34.32 mg, 100 pmoles) N-bromoacetylacetazola- mide in 0.1 M phosphate buffer (300 ml) at pH 7.2 for 7 hours at 24”. The reaction mixture was dialyzed against distilled water and freeze-dried.

A portion of the dried mat.erial (200 to 300 mg) was then applied to a DEAE-cellulose column (2.5 X 50 cm) previously equilibrated with 0.02 M Tris chloride at pH 8.4. Elution was begun with this buffer at a flow rate of 210 ml per hour. After the first peak had emerged completely, the elution was continued with sodium chloride in linear gradient from 0 to 0.5 M in 0.02 M

Tris chloride buffer at pH 8.4 using the same flow rate. The reservoir contained 1000 ml of buffer and 0.5 M sodium chloride and the mixing chamber contained 1000 ml of buffer. The effluent was monitored at 280 nm (Isco, model UA 2).

Another portion of the reaction mixture (150 mg) was applied to a preparative isoelectric focusing column (model 8102, 440-ml internal volume LKB Producter AB, Stockholm) and run ac- cording to the procedure of Svensson (25). Carrier ampholytes (Ampholine, LKB Producter AB, Stockholm) in the pH range 5 to 8 were used. The separation was achieved at 5” for 72 hours using a constant voltage power supply at 350 to 1000 volts depending on the current. At no time did the power exceed 5 watts. The column was emptied at a rate of 2 ml per minute and 5-ml fractions were collected. The absorbance at 280 nm and the pH were measured for each fraction prior to the com- bining of appropriate fractions and remova of Ampholine and sucrose by dialysis. Proteins were recovered by freeze drying.

Preparation and Separation of N-Bromoacetylacetazolamide-

C6H604NJ&Br2 (422.1)

Calculated: C 17.07, H 1.43, N 13.27, Br 37.87% Found : C 17.06, H 1.26, N 13.40, Br 37.90%

Stability of N-Bromoacetylacetuzolamide in Different Bu$ers- N-Bromoacetylacetazolamide (4 mg) was dissolved in the fol- lowing buffers (15 ml each): 0.1 M acetate, pH 5.2 and 6; 0.1 M

phosphate, pH 5.2, 7.2, and 8.2; 0.1 M Tris sulfate pH 7.6 and 8.7; 0.1 RI veronal pH 7.8 and 8.7; and 0.1 M ammonium sulfate, pH 9. The solutions were kept at 24” and aliquots (3 ml) removed at different times (5 min, 1, 2, 24, and 48 hours) were freeze-dried and then triturated with 1 ml of peroxide-free tetrahydrofuran to dissolve the inhibitor and its degradation product. After centrifugation, + of the supernatant was applied separately to silica gel plates.

Alkylated Ifuman Carbonic Anliydrase B Alkylated with Bromo- acetic Acid (Dialkylated Enzyme)-Human enzyme B alkylated with N-bromoacetylacetazoIamide and purified by DEAE-cellu- lose chromatography (78 mg, 2.6 pmoles) was further reacted with 40-fold molar excess of bromoacetate (14.3 mg, 104 /Imoles) in Tris sulfate buffer (8.7 ml) at pH 7.6 for 24 hours at 24”. Buffer and excess reagent were removed by dialysis. The freeze- dried protein (30 mg) was applied to an LKB isoelectric focusing apparat,us (440ml internal volume). The separation was carried out as described above with the exception t.hat a pH gradient be- tween 5 and 7 was used.

Starch Gel Electrophoresis-This was performed according to the procedure of Rickli et al. (26).

Enzyme Assays-The COY hydration activity was measured at 4” according to the method of Wilbur and Anderson (24). The esterase activity was det.ermined at 24” using p-nitrophenyl acet,ate as the substrate and by following the optical density change at 348 nm on a Gilford recording spectrophotometer (16). The observed value was corrected for the rate of hydrol- ysis of pnitrophenyl acetate in the absence of the enzyme. The reaction mixture cont,ained 1 X lo-6 M to 3 X lo+ IVX enzyme,

Amino Acid AnalysesProtein samples (1 to 2 mg) were hy- drolyzed in constant boiling glass-distilled hydrochloric acid (0.5 to 1 ml) in evacuated, sealed tubes for 22 hours at 110”. Alkyl- ated protein samples were oxidized prior to acidic hydrolysis (18, 27). Tryptophan was determined according to the method of Matsubara and Sasaki (28). Amino acid analyses were per- formed with a Beckman-Spinco automatic amino acid analyzer according to the method of Spackman et al. (29).

Measurement of Radioactivity-Radioactive counting was car- ried out in a Packard Tri-Carb liquid scintillation spectrometer (model 3375). Dried protein samples were dissolved in 1 ml of Soluene (Packard Instrument Co.). Counts were corrected for quenching by means of an external st,andard.

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of Juno 25, 1972 X.-C. C. TVonu, S. I. Kandel, M. Kandel, and A. G. Gornall 3513

RESULlS TABLE I

Isolation of Pure N-Brow~oacefylacetazolamide-The reaction of acet,azolamide with bromoacetic anhydride gave N-bromoacetyl- acetazolamide in good yield but the product contained, in addi- tion t.o unreact.ed aeetazolamide, some N-bromoacetylbromo- acetazolamide and N-acetylacetazolamide. These last two derivatives probably resulted from an intermolecular transacyla- tion and they were prepared separately in order to facilitate their identification as by-products of this reaction. The reaction mixture was separated by preparative thin layer (2 mm) chro- matography; when re-run on thin layer (0.25 mm) N-bromoace- tylacetazolamide moved as a single spot even when 70 pg of the final product were applied. In this way acetazolamide impurity, even as little as 0.01 %, would have been detected in the purified N-bromoacetylaceta~zolamide.

Apparent ionization constants of M-acetylacetazolamide and av-bromoacety7acetazolamide

Acetazolamide derivatives (50 pmoles) were dissolved in 1 ml of acetone and diluted wit.h 0.05 M KC1 solution to 25 ml. Of the solutions, 5 ml were stirred under NS until the pH was constant and titrated under Nz with 0.025 N NaOH at 24” in the Radiometer automatic titration apparatus.

Compound PK’

C-CO-NH-) ( (--SO%-XH-)

iv-Acetylacetazolamide. . 8.16 +I 0.02 3.28 f 0.01 N-Bromoacetylacetaaolamide. . . 8.10 f 0.02 3.08 f 0.02

Stability of N-Bromoacetylacetazolamide in Various Bu$ers- It was shown by silica gel thin layer chromat.ography that N- bromoacetylacetazolamide is rapidly converted to acetazolamide in 0.1 M Tris sulfate buffer (Tham, Fisher Scientific or British Drug Houses AnalaR) at pH 7.2 and 8.7 and room temperature. After 1 hour no detectable amount of N-bromoacetylacetazola- mide could be found in the incubation mixture. The decom- position of the reagent was confirmed by showing that human carbonic anhydrase B (3.3 X lop5 M), when reacted for 8 hours with N-bromoacetylacetazolamide (3.3 X 10v4 M) that had been kept in 0.1 M Tris sulfate buffer at pH 7.2 for 23 hours, formed only trace amounts of carboxymethylhistidine derivative. Un- der identical conditions, in 0.1 RI phosphate buffer, 0.55 eq of His@Cm) was found in the hydrolysate.

TABLE II

Specificity of the reaction of human carbonic anhydrase B with AT-bromoacetylacetazolamide

The enzyme (3.3 X 10e5 M) and N-bromoacetylacetazolamide (3.3 X 10-d M) were kept in 0.1 M phosphate buffer at pH 7.2 and room temperature for 24 hours. When used, urea (8 M) and aceta- zolamide (3.3 X low4 M) were previously incubated with the en- zyme for 30 min prior to the addition of N-bromoacetylacetazol- amide. Following the reaction, buffer and unreacted reagent were removed by dialysis against distilled water and the carboxymethyl- histidine content of the lyophilized protein was determined.

Experiment

No decomposition of N-bromoacetylacetazolamide could be observed in thin layer chromatograms when exposed to 0.1 M

ammonium sulfate buffer at pH 9.2 for as long as 7 hours, or in 0.1 M phosphate buffer in the pH range 5.2 to 8.2 for 48 hours.

In 0.1 M acetate buffer at pH 5.2 a different type of decomposi- tion was observed. On thin layer chromatograms a gradual in- crease in N-acetylacetazolamide corresponded to a decrease in N-bromoacetylacetazolamide, a conversion that could be ex- plained by an intermolecular transacylation mechanism.

The formation of acetazolamide in Tris sulfate buffer might be due to some catalytic contaminant in this buffer. Hydrolytic or ammonolytic cleavage cannot be the cause, since the compound appears to be stable in ammonium sulfate and phosphate buffer even at alkaline pH. The fast decomposition of N-bromoacetyl- acetazolamide in Tris and acetate buffer prompted us to use phosphate buffer in these studies, whenever it seemed necessary, in spite of the known inhibitory effect of this buffer (10).

Native enzyme plus A’-bromoacetyl- acetazolamide...................... 0.72 0.13

Zinc-free enzyme” plus N-bromoace- tylacetazolamide. . . . . . 0 Trace

Native enzyme + urea (8 M) plus N-bromoacetylacetazolamide. . Trace Trace

Native enzyme + acetazolamide plus N-bromoacetylacetazolamide . 0 Trace 0

Carboxymethyl human enzyme B.. . . . Trace 1.02 0 Carboxymethyl human enzyme B plus

N-bromoacetylacetazolamide . 1 Trace 1 1.06 1 0

a This enzyme preparation contained less than 0.05 g atom of zinc per molecule of enzyme.

Apparent Ionization Constants of Acetazolamide Derivatives- Potentiometric titrations of N-acetylacetazolamide and N-bro- moacetylacetazolamide yielded two pK values for each com- pound (Table I). The two dissociabIe protons are located in each case at the acetamido and sulfonamido groups, respectively. The assignment of the apparent pK values is based on both structural considerations and literature data (30).

Specificity of Interaction of N-Bromoacetylacetazolamide with Human Carbonic Anhydrase B-The results in Table II show that a lo-fold molar excess of N-bromoacetylacetazolamide, when reacted with human enzyme B at pH 7.2 for 24 hours, yielded histidines alkylated at the l-nitrogen, the a-nitrogen, and the 1,3-nitrogen positions. Under identical conditions, neither the native enzyme in the presence of 8 Rx urea nor the zinc-free

enzyme reacted, thereby indicating that the native structure of the active site is necessary for the reaction. The observation that acetazolamide, a potent reversible inhibitor of carbonic anhydrases, protects against these reactions provides further evidence that the interactions take place at or near the active site. a-Haloacetates are known to interact with human carbonic anhydrase B with the formation of 1 eq of S-carboxymethylhis- tidine in the active site region (10, 17). The carboxymethyl human enzyme B so formed does not react with N-bromoacetyl- acetazolamide. Results shown later indicate that this observa- tion cannot be taken as evidence that the two reagents react with the same histidine. Nevertheless, it seems that the car- boxymethyl group hinders the approach of the reagent toward the active site histidines either sterically or by charge repulsion.

Finally, the fact that human enzyme C does not react under the same conditions (Table VIII) can also be considered evidence for the specificity of the reaction with human enzyme B.

0.18

0

Trace

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

3814 Histidine Involvement in the Activity of Carbonic Anhydrase Vol. 247, No. 12

TABLE III

Time course of inactivation of human carbonic anhydrase B with N-bromoacetylacetazolamide

The reaction mixture contained 3.3 X 1O-5 M enzyme and 3.3 X W4 M N-bromoacetylacetazolamide in 0.1 M phosphate buffer at pH 7.2 and room temperature. Aliquots (2 ml) were withdrawn at appropriate time intervals and diluted to 25 ml with 0.025 M phosphate buffer, pH 7.6 and the esterase activity was measured. Simultaneously, 4.ml aliquots were withdrawn and immediately dialyzed against distilled water. Following lyoph- ilization, these samples were used to determine the extent of carboxymet.hylhistidine formation.

Time

hrr

1 2 4 8

11 24 48

Alkylated histidines formed

0.12 None None 0.20 None None 0.31 None None 0.55 Trace Trace 0.61 Trace 0.06 0.72 0.13 0.14 0.65 0.20 0.29

:nactivation observed

% 8

12 20 38 48 65 83 1 -

Inactivation/ eq of

His(l-Cm)

66.7 60.0 64.5 69.0 76.0 90.0

127.6

TABLE Iv

Formation of alitylated histidines in human enzyme B at various concentrations of N-bromoacetylacetazolamide

The reaction mixture contained 3.3 X 1O-6 M enzyme and various concentrations of inhibitor in 0.1 M phosphate buffer at pH 7.2 and room temperature. After 24 hours the reaction mix- ture was dialyzed against distilled water, freeze-dried, and an- alyzed.

Concentration of Alkylated histidines formed

N-bromoacetyl- acetazolamide His(l-Cm) H&(3-Cm)

i 1

His&B- di-Cm)

moles /mole cnayme

1 2

10 20

100

residuelmolecule enzyme residue/molecule enzyme

Trace Nil Nil Trace 0.35 Nil Nil 0.35 0.72 0.13 0.14 0.86 0.69 0.18 0.26 0.95 0.64 0.21 0.32 0.96

The sum of His,&Cm) and H~o&3h-Cm)

Time Course of Inactivation of Iiuman Carbonic Anhydrase B with N-Bromoacetylacetazolanzide-N-Rromoacetylacetazolamide slowly inactivates the enzyme and the acid hydrolysate of the reaction mixture contains varying amounts of His&Cm) His- (3-Cm) and His(l,3-di-Cm) depending on the reaction time

(Table III). In the first 8 hours, during which the alkylation is confined to the l-nitrogen position, the ratio of percentage in- hibition to the number of equivalents of His(l-Cm) varies be- tween values of 60 and 69 indicating that the formation of the enzyme species alkylated at the l-nitrogen position of a histidine exceeds the rate of inactivation. -4s the reaction proceeds fur- ther this ratio increases, reaching a value of 127 after 48 hours

with the simultaneous formation of two other alkylated histidine species. It appears that while a relatively fast alkylation at the l-nitrogen position leads to a partial inactivation the two slower secondary alkylations cause a much greater degree of in-

01;’ 1 1 I I I I I 5 6 7 8

PH

FIG. 2. His@-Cm) formation in the reaction of human carbonic anhydrase B with N-bromoacetylacetazolamide as a function of pH. The reaction mixture contained 3.3 X lo+ M enzyme and 3.3 X 10e4 M N-bromoacetylacetazolamide in 0.1 M phosphate buffer at values of pH between 5.6 and 8. After 4 hours the re- action mixture was dialyzed against distilled water, freeze-dried, and analyzed for the extent of His(l-Cm) formation.

activation. However, the relative contribution of these two latter reactions is not known.

In Table IV we have tabulated the number of equivalents of alkylated histidines that formed in 24 hours with different con- centrations of the reagent. When this concentration was in- creased from IO-fold to 20- and loo-fold molar excess, the amount of l-alkylated enzyme gradually decreased with a parallel in- crease in 1,3-dialkylated enzyme. The total of the two derivatives approximates and does not exceed 1 eq per molecule of enzyme. Based on these observations it seems likely that the reagent at first reacted with the l-nitrogen of a histidine residue, followed by a second alkylation of the same histidine at the 3- nitrogen position to form dialkylated histidine. Further evi- dence for the sequential formation of a dialkylated histidine residue is presented in Table VII.

pH Dependence of Reaction-Fig. 2 shows the amount of His- (l-Cm) formed in 4 hours as a function of pH. The rate was maximal at a pH near neutrality and decreased at both lower and higher pH. In order for the reaction to proceed, the interacting histidine should be present in unprotonated form since the re- action can be considered to take place via the nucleophilic attack of the l-nitrogen of the histidine on the electrophilic carbon atom of the N-bromoacetyl group of the reagent. Based on this con- cept one might expect that the rate of alkylation would increase with an increase in pH up to a limiting value after which it would remain constant. The increase in formation of alkylated enzyme between pH 5.6 and 7 probably reflects the deprotona- tion of the histidine that reacted. The reason for the decrease in the rate above pH 7.0 is less obvious. We can rule out the pos- sibility that the decomposition of N-bromoacetylacetazolamide at alkaline pH causes the decreased rate, since the reagent after being kept at pH 8 for 24 hours is still equally capable of reacting at pH 7.2. N-Bromoacetylacetazolamide possesses two ioniz- able hydrogens. While dissociation of the acetamido proton (pK = 8.1, Table I) is unlikely to contribute to the electrophilic- ity of the methylene carbon of the N-bromoacetyl group, the dissociation of the sulfonamido proton might. However, the

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 25, 1972 S.-C. C. Wong, S. I. Kandel, M. Kandel, and A. G. Gornall 3815

pK of this group was found to be 3.1 (Table I) which is far too low to affect the reaction between pH 7 and 8.

Since the rate of the reaction is likely controlled by the con- centration of the reversible enzyme-inhibitor complex, a rapid decrease in the affinity of the enzyme for the reagent at alkaline pH would considerably decrease the rate of the reaction. A decreased binding of sulfonamides and anionic inhibitors at alkaline pH values has been reported (7, 15, 31). However, the decrease in rate at alkaline pH might also be due to the influence of a residue located close to the reactive histidine such that the former should be present in protonated form to facilitate the reaction by enhancing the nucleophilicity of the latter. Our data are not sufficient to distinguish between these two possi- bilities.

Separation of Alkylated Human Enzyme B (Alkylafed Enzyme) -In view of the effects of alkylation shown in Tables II and III,

VOLUME (ml)

FIG. 3. Elution pat,tern from the column chromatography of N-bromoacetylacetazolamide-alkylated human carbonic an- hydrase B on DEAE-cellulose. See under “Experimental Pro- cedure” for details. The bars indicate the fractions that were pooled.

‘; 6.0 J -8

6-

c.

-5 = a

-4

” .

9 17 25 33 41 49 57 65

FRACTIONS

FIG. 4. Isoelectric focusing of the products of a ‘i-hour reaction mixture of human carbonic anhydrase B and N-bromoacetyl- acetazolamide. After the removal of buffer and excess reagent by dialysis, 160 mg of freeze-dried protein were applied to an LKB electrofocusing apparatus. The pH gradient was established between 5 and 8 and equilibration allowed to proceed over a period of 72 hours after which 5-ml fractions were collected. The bars indicate the fractions that were pooled.

it was of interest to study the properties of this form of the enzyme. To avoid multiple alkylation, the reaction mixture contained a lo-fold molar excess of N-bromoacetylacetazolamide and the incubation period was limited to 7 hours at room temper- ature. The result of the chromatographic separation of this mixture on DEAE-cellulose is shown in Fig. 3. Fraction I was found to be identical with the unreacted native enzyme. Frac- tion III was not characterized although its acidic hydrol- ysat,es contained His(l-Cm), His(3-Cm), and His(1 ,&di-Cm) indicating that even under the controlled conditions specified above the reaction is not confined to the l-nitrogen position of one histidine. After acid hydrolysis, Fraction II yielded 0.91 eq of His(l-Cm) with a corresponding loss of 0.89 eq of histidine. In all other respects the amino acid content (including trypto- phan) of this hydrolysate was identical with that of the native enzyme. When both the reaction and the separation were repeated under identical conditions using W-labeled N-bromo- acetylacetazolamide, Fraction II contained 0.98 eq of covalently bound reagent. Therefore, Fraction II is human carbonic anhydrase B alkylated with this reagent at the l-nitrogen posi- tion of one or more histidine residues. During electrophoresis on starch gel at pH 8.9 (Fig. 5) this alkylated enzyme migrated

FIG. 5. Starch gel electrophoresis of alkylated and dialkylated human carbonic anhydrase B. Well 1, native enzyme. Well 2, N-bromoaeetylacetazolamide-alkylated enzyme separated by column chromatography (Fig. 3, Fraction II). Well S, N-bromo- acetylacetazolamide-alkylated enzyme separated by isoelectric focusing (Fig. 4, Fraction I). Well 4, N-bromoacetylacetazol- amide-alkylated enzyme alkylated further with bromoacetate (dialkylated enzyme) separated by isoelectric focusing (Fig. 9, Fraction I). The electrophoresis was run in Tris chloride buffer at pH 8.9. Migration is toward the anode from top to bottom.

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

3816 Histidirle Involvement in the Activity of Carbonic Anhydrase Vol. 247, No. 12

-K i z-9.2 mM (a) =-5.3 mM (b) [BROMOACETATEI $ mM

FIG. 6. Dixon plots for nat.ive and alkylated human enzyme B catalyzed hydrolysis of p-nitrophenyl acetate in the presence of bromoacetate. The inhibition was determined in 0.025 M phos- phate buffer containing 0.8% acetone at pH 7.6 assuming non- competitive kinetics. The enzyme concentration in the reaction mixture was 2.66 X lo-” M. The p-nitrophenyl acetate concen- tration was 4 X 10-t M. The assay was carried out as described under ‘LExperiment’al Procedure.” Plot a, alkylated human enzyme B; Plot b, native human enzyme B. The velocity is expressed as the change in optical density at 348 nm per min. Lines were drawn from a least squares analysis of the data points.

toward the anode ahead of the native enzyme. However, the electrophoretogram indicated that it contained minute amounts of unidentified protein mat~eriuls.

The alkylated enzyme was obtained by electrofocusing (Fig. 4) in a form that appeared homogeneous on starch gel elec- trophoresis (Fig. 5). Only Fractions I and II were isolated. The ot.her peaks when combined did not exceed 5% of the alkyl- ated enzyme. As expected, the alkylated enzyme (Fraction I) has a lower isoelectric point (pH 6.12) than does the native enzyme (Fraction II) siuce the former possesses two extra nega- tive charges. Our value of 6.67 for the native enzyme agrees well with that of 6.57 reported by Funakoshi and Deutsch (32) who also employed the electrofocusing technique. It deviates, however, by almost one pH unit from the value of 5.7 determined by Rickli et a2. (26) using boundary electrophoresis in a system of buffer solutions. Unlike Rickli and his coworkers, Funakoshi and Deutsch removed dissolved COY from their electrofocusing system and they have suggested that this is the possible cause of the observed difference between their value and that of Rickli et al. However, since CO2 was not removed from our system this rationalization appears unlikely. On the other hand, our enzyme as well as that of Funakoshi and Deutsch (32) had been treated with rhloroform-ethanol solutions during the initial removal of hemoglobin from the hemolysates, whereas Rickli et al. (26) had employed the more gent,le gel filtration method for this purpose. Thus, the difference in the values obtained for the isoelectric point of the native enzyme could reflect either the different isolation procedures or the electro- focusing versus t.he zone electrophoresis technique.

Binding of Substrate and Inhibitors to Native and Alkylated Enzymes-Like native human enzyme 13, the alkylated enzyme binds bromoacetate and acetazolamide.’ The inhibition con-

1 We attempted to determine the Ki of N-bromoacetylaceta- eolamide for human enzyme B using the Dixon plot. This result.ed in a curve that, was concave upward instead of the expected straight line, indicating that the KJ value decreases with an in-

Inhibition of esterase activity of native and alkylated human carbonic anhydrase B by acela~olamir~e

The esterase activity was determined at pH 76 in 0.025 M phos- phate buffer. Concentrations of the native and alkylated en- zymes were 2.66 X lo-” M and 5.33 X lo+ M, respectively.

Concentration of inhibitor

Native enzyme

2.64 0.25 5.28 0.35 7.92 0.39

10.56 0.41 13.20 0.44

0.04

I - stant’s (KJ for these derivatives were determined on the assump- tion that the inhibition mechanism is noncompetitive (1). K; values of bromoacetate for both the native and alkylated en- zymes were found by the Dixon plot (Fig. 6) to be 5.3 and 9.2 m&r, respectively when measured in 0.025 M phosphabe buffer at pTI 7.6. The inhibition constant of acetazolamide was cal- culated from Equation 1.

k’, = tEt - EI)(zt - ET) z

@I) (1)

in which Et and It represent the total concent,ration of enzyme and inhibitor, respectively and EI the concentration of the enzyme-inhibitor complex. The latter value was estimated from t,he ratio of the initial velocities in the presence (Vi) aud in the absence (V,) of inhibitor using the following relationship:

(Ef)=L-E,; (2)

Acetazolamide combines almost stoichiometrically with car- bonic anhydrases, thereby determining the lower limit of the concentration of acetazolamide in the reaction mixture. If the enzyme concentration exceeds the inhibitor concentration, the value of (E1) obtained from Equation 2 will approximate the value of It. Consequently, the quantity It - (KI) may fall within the range of experimental error and in this way zero or even negative values may be obt,ained for the Xi. On the other hand, the sensitivity of the spectrophotometer determines the upper limit of the concentration of the inhibitor, since the value of Vi rapidly approaches zero as t.he inhibitor concen- tration exceeds that of the enzyme. Thus, the nature of the interaction between acetazolamide and carbonic anhydrase severely restricts the useful range of inhibitor concentration. In Table V are list.ed the K; values calculated from Equation 1 for the native and alkylated enzymes at pl1 7.6 in 0.025 M phos-

phate buffer. We found, for example, that at equimolar con- centrations of the native enzyme and acet,azolamide, this value

crease in the inhibitor concentration. The Kg at 0.21 mM inhibitor concentration, as calculated from Equation 1, was found to be 0.3 mM for the native and 0.19 m&f for the alkylated enzyme. At an inhibitor concentration of 0.96 mM, these values decreased to 0.11 mM and 0.05 mM, respectively. The concentrat,ion dependence of the inhibition const.ant is possibly due to the presence of a minute amount of acetazolamide in our N-bromoacetylacetazolamide preparation.

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 25, 1972 S.-C. C. TVong, 8. I. Kandel, &I. Kandel, and A. G. Gornall 3517

was somewhat lower than that obtained at higher concentra- tions of the inhibibor. Our value for the nut,ive enzyme, 0.35 pM and that of Armstrong et al. 0.25 PM (5) agree reasonably well, although the two are not directly comparable due to differences in the reaction conditions. The value obtained for the alkylated enzyme at equimolar concent.rations of acetazol- amide and enzyme is 0.04 PM, which is approximately one-ninth that for the native enzyme. Even at this inhibitor concentra- tion, the a&i&y remaining was only 3.1 yG, and therefore we did not att.empt t,o e&mate Ki values at higher inhibitor concen- trations.

In Table VI we present the Michaelis constants (Km) and catalytic constants (kont = JTm,, /&) which were determined by the Lineweaver-Burk plot at three 11~I values. A typical plot for the native and alkylated enzymes at pH 7.6 is shown in Fig. 7. The values of K, for the native enzyme at pH 6, 6.8, and 7.6 vary between 5.6 and 6.9 mM. These values are close to the value (6 mu) obt,ained by Verpoorte et al. (33), but considerably smaller t,han that (20 nlM) obtained by Whitney (34). The corresponding ralues for the alkylated enzyme are 6.1 rni\l and 6.4 m&f, respectively. These variations probably do not indicate a variation of K, with pH but rather reflect the experimental uncertainties involved in the use of p-nitrophengl acetate as substrate (5). A pH-independent binding of p-nitrophenyl acetate to the n:tt.ive enzyme has been report.ed (33, 34).

Esterase and CO2 Ilydration Activity of Alkylafed Enzyme-The a.lkylated enzyme possesses residual esterase and CO2 hydration activity, although the values show some variation depending on the method of separation. Samples sellarnted on a DEhE- cellulose column possess 32 to 35yA residual esterase activity, while those isolated by electrofocusing retain only 25 to 28% residual esterase activity when assayed at pI1 7.6. Values for COY hydration activity are 40% and 345;, respectively. The lower value of the electrophoretically purified sample did not increase a,fter filt.ering t,hrough a Sephades G-50 column to remove traces of ampholine which might still be present even after exhaustive dialysis. The observed discrepancy could be due to the presence of some native enzyme in the chromato- graphically purified sample. A partial loss of enzymatic activ- ity during electrofocusing appears t.o be ruled out since the specific catalytic activity of the unreacted enzyme reisolated from t’he react,ion misture by elect’rofocusing (Fig. 4, Fraction II) is identical with that of the native enzyme.

Fig. 8 illustrates the variation with pH of t,he esterase activity of both the native and alkylated enzymes. The observed activity at pH 9.2 was taken as lOO’$&. The resulting plot for the native enzyme is sigmoid and the pK as estimated from the half maximal velocity is approximately 7.6. Our value for the pK, although slightly higher than earlier reported values (7.3 to 7.4) (17, 33) agrees reasonably well. The same plot for the alkylated enzyme is not sigmoid, indicating that a more complex relationship exists between pH and activity. This complexity is readily seen in the plot of the residual enzyme activity (mod- ified/native) versus pH, which is also included in Fig. 8. The activity of the alkylated enzyme at pH 6 is 95% that of the native species. It t,hen decreases sharply with the increase in pH, leveling off at around pH 7.6. This observation indicates that although the activity of the alkylated enzyme, like that of the native enzyme, increases with increasing pH, this increase is greater for the native than for the alkylat.ed enzyme. Obvi-

TABLE VI Kinetic conslants of native and alkylated human

carbonic anhydrase B Kinetic constants were determined as described in Fig. 7 except

that, at pH 6 and 6.8, 0.075 M 2,2-bis(hydroxymethyl)-2,2’,2”- nitriIoethano1 (his-tris) suIfat,e was used.

Enzyme PH

Native ............. 6.0 AIkylated. ......... Native. ............ G.8 Alkylated. ........ Native ............. 7.6 Alkylated. .........

m‘u

6.2 f 0.4 6.1 f 0.2 5.6 f 0.6 6.4 i 0.7 cl.9 * 0.8 6.3 f 0.4

&at

min-’

32.8 f 2.4 30.9 f 1.5 77.7 f F.6 40.4 f 4.0

173 f 18 62 z!z 2

I 1 I

80

I

I I 60-

1 v /--

/

.

40-

,& (mM-‘1

FIG. 7. Double reciprocal plot of initial velocity as a function of p-nitrophenyl acetate concentration. Errzymic activity was measured in 0.075 M Tris sulfate at PI-I 7.6 as described under “Experimental Procedure.” The enzyme concentration in the reaction mixt’ure was 3 X 1OW M and the substrate concentration was varied from 0.59 X lo+ M to 2.38 X lo-” .M. Lines were dra\vn from a least squares analysis of the data points. Velocities are expressed in mmolar concentration of p-nit’rophenol plus p-ni- trophenolate ion per min, a, nat.ive human enzyme U; b, alkylated enzyme B.

ously from t~his plot we cannot obtain nrcurately the pK of the catalytically essential group in the alkylated enzyme, although it seems that t,he value has shift.ed downward relative t.o t.hat of the native enzyme.

Covalent Interaction of Alkylated Enzyme with Blomoacetate and N-Bromoacetylacetazolamiok-Tile alkylated enzyme react.s with bromoacetate and N-bromoacetylacetazolamide in a com- plex manner (Table VII). The acid hydrolysate of the purified enzyme alkylated with N-bromoacetylacetazolamide cont.ained 0.91 eq of His(l-Cm) and after a period of 24 hours in the presence of bromoacetate this value decreased to 0.52 eq with the simul- taneous formation of 0.39 eq of His(l,3-di-Cm). The latter value almost completely accounts for the loss of His(l-Cm) initially present, indicat,ing that the same hist.idine reacted with

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

3818 H&id&e Involvement in the Activity of Carbonic Anhydrase Vol. 247, No. 12

PH

FIG. 8. pH dependence of the esterase activity of native and alkylat.ed human carbonic anhydrase B. Assays were carried out in 0.025 M bis-Tris sulfate (from pH 6 to 6.8) and Tris sulfate (from pH 7.2 to 9.2) buffer as described under “Experimental Procedure.” The concentration of the native enzyme in the reaction mixture was 6 X lo-$ M between pH 6.0 and 6.8 and 3 X 10e6 M between 7.2 and 9.2. The concentration of the alkylated enzyme was S X 10-O M between pH 6.0 and 6.8 and 4 X lo-” M be- tn-een pH 7.2 and 9.2. The observed activities of both enzymes at pH 9.2 were taken as 100% and other values are related to this. 0, esterase activity of native enzyme; 0, est.erase activity of alkylated enzyme; n , residual esterase act,ivity of alkylated en- zyme

the two reagents. In addition to this, 1.05 eq His(3-Cm) was also formed. The native enzyme under the same conditions yields 0.94 eq of His(3-Cm). Acetazolamide prevented the formation of a dialkylated histidine since when the reaction was carried out in the presence of this inhibitor only a trace of His(l,3-di-Cm) was found in the hydrolysate and the value of His@Cm) remained unchanged relative to that of the alkylated enzyme. Acetazolamide also considerably reduced the car- bosymethylation of the histidine at the 3-nitrogen position in that this value dropped from the initial 1.05 eq to 0.14 equiva- lents.

To rationalize these observations, we postulate that the bromo- acetate at first reacted with the 3-nitrogen of a histidine of the alkylsted enzyme which is likely identical with the reactive histidine of the native enzyme known to yield His(Y-Cm) with bromoacetate (17). This faster reaction would be followed by a slow carbosymethylation at the S-nitrogen position of the histidine that had originally reacted with N-bromoacetylacet- azolamide at the l-nitrogen position.

As expected, N-bromoacetylacetazolamide also reacted with the alkylated enzyme (Table VII). The 0.91 eq of His(l-Cm) initially present dropped to 0.72 eq with the concurrent appear- ance of 0.22 eq of His(l,3-di-Cm). The total of t.hese two derivatives is identical with the amount of His(l-Cm) found in

TABLE VII Covalent interaction of alkylated enzyme with bromoacetate and

N-bromoacetylacetazolamide

The reaction mixture contained native, carboxymethyl, and N-bromoaeetylacetazolamide alkylated human carbonic anhy- drase B and various reagents at the concentrations indicated. The interaction with bromoacetate was carried out in 0.1 M Tris sulfate buffer at pH 7.6 and that with N-bromoacetylacetazol- amide in 0.1 M phosphate buffer at pH 7.2. The concentration of bromoacetate in the reaction with the alkylated enzyme was increased relative to that with the native enzyme to compensate for its decreased affinity (Fig. 6). After keeping the mixture for 24 hours at room temperature, the buffer and unreacted reagent were removed by dialysis against distilled water and the protein was then freeze-dried and analyzed for its alkylated histidine content.

Experiment

illkylated enzyme. . , Native enzyme. . .

Plus bromoacetic acid. . Alkylated enzyme.. . . .

Plus bromoacetic acid. Alkylated enzyme.. . . .

Plus acetazolamide.. . . Plus bromoacetic acid.

Alkylated enzyme. Plus Xbromoacetylace-

tazolamide. . Alkylated enzyme..

Plus acetazolamide.. Plus N-bromoacetylace-

tazolamide. . .

:oncen - ( .ration of the :nzyme

llZM

0.3

0.3

0.3

0.03

0.03

-

:oncen- tration of the reagent

?nM

18

31.2

6 31.2

0.33

0.33

0.33

‘ -

I -

4lkylated histidines formed

His(l- Cm)

I !

H$isi’- H+O, 3- dl-Cm)

residue/molecule ensyme

0.91 None None

None 0.94 None

0.53 1.05 0.39

0.92 0.14 Trace

0.72 0.20 0.22

0.89 Trace None

the alkylated enzyme, which confirms the multiple interaction of N-bromoacetylacetazolamide with the native enzyme (Tables III and IV) as well as the ability of this reagent to alkylate both nitrogens of the same histidine residue. We do not know whether the presence of 0.2 eq of His(3-Cm) (Table VII) arises from the alkylation of the histidine that reacted with bromo- acetate or whether it is a result of the alkylation of yet another histidine. Our observation that the carboxymethyl human enzyme B does not react with N-bromoacetylacetazolamide (Table II) strengthens though it does not prove the hypothesis that the same histidine is involved in both cases. Acetazol- amide protects against these alkylat.ions (Table VII) support- ing our earlier suggestion that both reactive histidines are in the active site region.

IIum.cl?z Carbonic Anhydrase B Alkylated with N-Bromoacetyl- acetadamide and Browacetate (Dialkylated Enzyme&The dialkylated enzyme from a reaction mixture of N-bromoacetyl- ncetazolamide alkylated human enzyme B and bromoacetate was separated in order to test its esterase activity. Our attempt by several chromatographic techniques to obtain the desired enzyme species in pure homogeneous form was unsuccessful. Isoelectric focusing between the pH gradient of 5 and 8 led to a single peak (Fig. 9), and the ascending limb of the peak (Frac- tion I) appears to be almost homogeneous on starch gel (Fig. 5). However, the descending limb (Fraction II) in addition to

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 25, 1972

1.6

S.-C. C. Wong, S. I. Kandel,

1

FRACTIONS

FIG. 9. Isoelectric focnsing of the reaction products of AT- bromoacetylaeetazolamide-alkylated human carbonic anhydrase B with bromoacet.ate. After the removal of buffer and excess reagent by diaIysis 25 mg of freeze-dried protein were applied to an LKB electrofocusing column. Conditions were t,he same as in Fig. 4 with the except,ion that a pH gradient between 5 and 7 was used. The bars indicate the fractions that were pooled.

TABLE VIII Covalent interaction of N-bromoacetylacetazolamide with horse

carbonic anhvdrase B, huntan carbonic anhydrase C, and bovine carbonic anhydrase B

The reaction conditions were the same as those described in the legend of Table III with the exception that the enzyme species indicated in the first column were used. Reaction times of 8 and 24 hours were employed. The data for human enzyme B is in- cluded in this table for purposes of comparison.

Enzyme

HumanB.... HorseB..... Human C.... Bovine B. Human B.... HorseB..... Human C . . Bovine B.

.......

.......

.......

....... ....... ....... .......

1

-

Reaction time

hrs

8

8

8 8

24 24 24 24

Alkylated histidine formed

His His His (l-Cm) (3.Cm) (1,3-di-Cm)

I

midxe/?iwlecule en*yme

0.55 Trace Trace None 0.51 None None None None None None None 0.72 0.13 0.14 None 0.91 IGone None Trace None None 0.08 None

the main product, contained some other unidentified protein species.

After acid hydrolysis of Fraction I, 0.92 eq of His(l-Cm) and 1.04 eq of His(bCm) were found with a simultaneous loss of 2.1 eq of histidine. The esterase activity of the dialkylated enzyme at pH 7.6 was less than 0.8% relative to the native enzyme.

Covalent Interaction of Horse Carbonic Anhydrase B, Human Carbonic Anhydrase C, and Bovine Carbonic Anhydrase B with I\i-BronwacetylaeetazolamioG-VVe have reported earlier (18) that bromoacetate preferentially alkylates human enzyme B and horse enzyme 13, which represent low activity forms of this enzyme. By contrast, bromoacetazolamide was found to alkylate preferentially human enzyme C and bovine enzyme B, which represent high activity forms (18). As Table VIII shows,

M. Kandel, and A. G. Gomall 3819

the effect of N-bromoacetylacetazolamide on the two groups of isoenzymes is similar to that of bromoacetic acid. A lo-fold molar excess of this reagent reacts only with the two low activ- ity forms to yield carboxymethylhistidine residues, whereas no change in amino acid content is found after reaction with the high activity forms. Although the amount of alkylated histi- dine found in the two low activity forms is nearly identical, the position of the alkylation is not. Horse enzyme B reacted at the 3-nitrogen and not the l-nitrogen position. This observation demonstrates further that a structural difference exists between the active sites of the two groups of isoenzymes. Due to uncer- tainties in the affinity of N-bromoacetylacetazolamide~ we do not know whether this selectivity is absolute, indicating that there is no histidine sterically available for reaction, or whether it is relative, indicating that the histidine, although sterically available in the high activity forms, possesses a markedly different reactivity.

DISCUSSION

The complex nature of the interaction of carbonic anhydrases with inhibitors is well documented. The active site of bovine enzyme B is capable of interacting with 2 molecules of bromo- acetazolamide (23). Human enzyme B binds I molecule of bromoacetate, which reacts with a single histidine and in so doing shift,s away from the original binding site of the non- covalent enzyme-inhibitor complex, thereby making the recon- stituted site available for binding another molecule of bromo- acetate (17, 34).

The covalent interaction of N-bromoacetylacetazolamide with human carbonic anhydrase B is confined to the active site region, although in the over-all reaction two histidine residues and 2 or possibly 3 molecules of reagent are involved. If a mechanism similar to that for bromoacetate operates for N- bromoacetylacetazolamide the observed differences could be explained by differences in the structure of the two inhibitors. A second molecule of bromoacetate may fail to react simply because the distance between the binding point and a nucle- ophilic amino acid side chain is longer than the reagent itself. N-Bromoacetylacetazolamide is a considerabIy Iarger moIecule and as a consequence the second molecule, when bound, is capable of bridging to a reactive nucleophilic side chain.

Although a third molecule of N-bromoacetylacetazolamide can apparently be accommodated in the active site region, it is possible that the alkylated enzyme reacts with an additional equivalent of N-bromoacetylacetazolamide to form either His- (1,3-di-Cm) or His(3-Cm) but not both. Thus, a dialkylated but not a trialkylated protein would be formed. If either the reaction time or the reagent concentration could be increased to the point where the total amount of covalently linked acetyl- acetazolamide exceeded the limit of 2 eq per enzyme molecule, it would help to rule out this possibility. Alternatively, the purification and characterization of the various reaction products would also permit a final decision to be made. Neither of these approaches is within the scope of this investigation.

In contrast the reaction of the purified Wbromoacetylacetazol- amide alkylated enzyme with bromoacetate led t.o a reaction mix- ture in which the sum of the covalently bound reagent.s did exceed 2 eq, clearly demonstrating that a trialkylated species can be formed. The protection against alkylation at both posi- tions by acetazolamide suggests that these alkylations occurred in the active site region.

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

3520 HisticGne Involvement in the Activity of Carbmic Anh,y&ase Vol. 247, No. 12

From the foregoing it would appear that A-bromoacetylacet- azolamide interacts in the same way wit,h both the native and alkylated enzymes. Uromoacetate 011 the other hand reacts only once with the native enzyme (17) but twice wit,h the alkyl- ated species to form dialkylated and trinlkylated enzymes. To reconcile this apparent contradiction, we suggest that N-bromo- acetylacetazolarnide induces a local conformational change at the active site of the enzyme making a second histidine sterically available for the subsequent reaction with bromoacetate. Aro- matic sulfonamide and iodide-induced local conformational changes have been postulated recently by King and Robert’s (35) on the basis of nuclear magnetic resonance spectroscopic data.

The presence of two sequentially removed histidines in t,he active site region has been revealed by a.ffinit.y labeling with CL- haloacetates and N-chloroacetylchlorothiazide. cY-Haloacetates reacted with histidine residue number 204 from the amino terminus (36, 37), and the histidine that reacted wit,11 N-chloro- acetylchlorothiazide was located between residues numbers 65 and 75 (37).

Bromoacetate at pH 7.6 binds to the alkylated enzyme with a Ki of 9.2 IIIM, which is of the same order of magnitude as 6he Ki of t.his inhibitor for t,he nat’ive enzyme (5.3 m&r). Furthermore, the reaction with both the native and alkylat’ed enzymes occurs at the 3-nitrogen position of a histidine residue. These observa- tions suggest that bromoacetate reacted with the same histidine in both native and alkylat,ed enzymes and consequently the histidiue that reacted fastest with N-bromoacetylacetazolamide and is modified in the monoalkylated enzyme is not histidine201. The assignment of the position of t,he fastest reacting histidine in t,lre primary structure must wait mltil the sequence around it is known,

The properties of the nlkylated enzyme resemble in many respecm those of the native species. It still catalyses the hydrat,ion of CO? and t.he hydrolysis of p-nitrophenyl acetate. The pH profile of the ester hydrolysis catalysed by the modified enzyme is qualitatively similar to that of the native enzyme, but the apparent pK of the act,ivity linked group is shifted to a lower pII value. Bt pH 6.0 the residual esterase activity of the modified enzyme is 95q;, of that of the native enzyme. -4s the pH is increased the residual activit.y decreases rapidly, leveling off around neutrality at about 35:&. This demonstrates that, the decrease in t,he pK of t,he catalytically essent,ial group accounts at least in part for t’lie decrease in residual activity at neutral and alkaline pH after modification of the histidine.

Like the native enzyme, the alkylated species still binds bromoacet,ate or acetazolamide. While the affinit.y of bromo- acetate ~YZLS decreased, the affinity of acet,azolamide has increased as a result of the alkylation. Such a decrease and increase in the affinity of inhibitors has been reported previously for carbonic anhydrases that have been modified at the histidine?od position (34, 38). .4t three pH values (6, 6.8, and 7.6) the Ii,, of the alkylated enzyme is almost identical to that of the native enzyme, indicating that the former enzyme binds substrate with the same affinity as does the lat,ter? The modified enzyme reacts with bromoacetate to alkylate a second histidine in the active site

* Kinet,ic investignt,ions demonstrated that Zi, is the true dis- socintiou constant for CO* (39). No similar studies are available for p-nitrophenyl acetate, although in our discussion we assume t.hnt, similar relationshins exist between I<, and substrate constant (Z&) for the ester substrate as well.

region, this time at the 3-nitrogen position. This second histidine residue is probably identical with hist.idinezo4 in the native enzyme that reacts with bromoacetat,e (36, 37). The dialkylsted enzyme so formed is almost completely inactive.

911 of these observations indicat,e that t,he histidine residue that reacted with N-bromoacetylacetazolamide, although it must be located in the active site region, is not a catalytically essential amino acid. By discounting this histidine as an essen- tia.1 group, we lend support to the hypothesis that the group in the enzyme that controls t,he catalysis and ionizes around neutrality is not a histidine but likely the zinc-bound water molecule itself (8, 10, 12, 13).

An observation that must not be overlooked, however, is that there are at least, t,wo hist.idines in the active site region. While the chemical modification of one does not result in a com- plete loss of a.ctivity the chemical modification of two histidines does. This suggests that one or the other of these histidines does participate in the cat.alyais. IYpon t.he alkylat,ion of one histi- dine the enzyme can compensate for it by using a sterically neighboring histidine. For the following reasons we prefer to assign this participating role in the native enzyme to histi- dinesol which reacted with bromoacetate. The residual act.ivity of t,he bromoacetate alkylated enzyme never exceeds 15% (17, 34). Furthermore, this modified enzyme binds p-nitrophenyl acetate wit’h the same affinity as does the native enzyme, thereby indi- cating that the decreased activity is not due to a decreased biud- ing of the substrate, but &her to a decrease in V,,, (3-Q.

The psrticipation of histidineZo., in the catalysis is not neces- sarily functional, but rather it, could well be structural. X-ray crystallography of human enzyme C has revealed t.hat there are two histidine residues in t’he active site region that are probably bonded to 2 water molecules, which in turn are hydrogen bonded to several other water molecules, including the one that is liganded to the zinc ion (12). Khalifah (13) has proposed that one of t.he water molecules in such an ordered water structure would function both as a hydrogen donor and acceptor, thus catalyzing the CO2 hydration reaction. Sssuming a similar water structure in human enzyme B, it is postulated that the alkylation of histidine?04 would, either direct.ly or t,hrough an induced local conformational change, interfere wit.11 t.his struc- ture, resulting in a less favorable orieutation of t,he participating water molecules. The enzyme would then contiuue t,o function but with a reduced activit,y.

No functional role can be assigned to the histidine that reacted with N-bromoacetylacetazolamide, since at pH 6 the native and alkylated enzymes are kinetically almost indistinguishable. However, in the enzyme alkylated at llistidine204, it could be responsible for holding t,he participating water molecules ia the less favorable orientation suggested above. On further alkyla- tion to form a dialkylated species the wat’er structure would collapse completely, resnlt’ing in an almost nonfuiictioning enzyme.

While this interpretation of our results is att,ractive iu that it assigns a role to hist,idinenor, the almost complete lack of activity of t.he dialkylated enzyme could be due to a different type of conformational change, or could be the result simply of steric blocking of the active site region and the resultant prevention of substrate binding. We still lack sufficient information to exclude these alternatives.

Acknowledgments-We thank Mrs. L. Beysovec and Mr. A.

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Issue of June 25, 1972 S.-C. C. Wang, X. I. Kandel, 39. Kandel, anrl A. G. Gornalb 3821

Naay for t,lleir excellent and enthusiastic contributions to this 21.

research project. REFEl:EUCI+‘S 22.

i i 23.

1. ?ckREh-, T. H. (1967) Physiol. Rev. 47, 591 2. EDSALL, J. T. (1968) Harrvr/ &ct. 62, 191-230 24. 3. ItOTG&N, F. j. W. (193.i).Ph@oZ. Rev. 15, 241 4. I'OCKEH. Y.. AND ME.LNT. J. E. (1965) Biochemistw 4, 2535 5. ARYSTIL~ND; J. McD., RITmts, D. ‘V., ?ERP~~~~TF, J.1 A:, .UXD

25, 26,

I:DY.ILL, J. T. (1966) J. Bier. Chem. 241, 5137-5149 6. HENK.~RT,P., GUIDOTTI, G., .~ND EDSALL, J. T. (1968) J. Biol. 27.

Chem . 243, 2447-2449 28. 7. KEILNOH.XN, J. C. (1965) Biochenc. Biophys. Acta 96, 304 8. D.kv~s, R. P. (1961) in ‘I’hr Entumes, Vol. 5, BOYWR, I?. D., 29.

L~I~DY, H., AND MYIiI+XCK, K., cds. 2nd Ed, p. 543, Aca- demic Press, New York 30.

9. COLEMIIN, J. E. (1965) Biochemistry 4, 2614 10. BR.IDI~IJRY, S. L. (1969) J. Biol. Chrm. 244, 2002, 2010 31. 11. F.~BRY (RIBIv;),M. IC., KOIXIG, S. H., END SCHILLINGER, W.

E. (1970) J. Riol. Chem. 245, -K236-4262 32. 12. LILJAS, A., K.IXNIN, K. K., BERGSTEN, P.-C., WAARA, I.,

FRIDBORG, K., STRINDBERG, B., CARLDOM, U., JBILUP, L., LBvGRICN, S., AND PETX,F, M. (1972) Sature New Biol. 235,

33,

131 13. KHILIF.UI, 1~. 0. (1971) ?I. nior. Chem. 246, 2361-2573

34.

14. W.\NO, J. H. (1968) Science 161, 328 35.

15. POCKLCB, Y., AND STONE, J. T. (1968) Biochemistry ‘7, 2936 16. WHITNEY, P. L.,Fij~,sc~r, (>.,NTu.~N, P. O., .IND M.~LMSTR~M,

36

1~. G. (1967) J. BioZ. Chem. 242, 4206-4211 37'

17. WHITNI~:Y,P. L., NYM.IN,P.O., .IR‘D M.~LMSTIL~~M, B. G. (1967) J. Biol. Chem. 242, 4212-4280

18. KISDEL, M., GORN.\LL, A. C;., WONG, S.-C. C., AND KANDEL, 9. I. (1970) J. Bid. Chcn~. 245, ‘Sk-2430

19. NILSSON, A., -4~0 I,INI)RE;OG, S. (196i) Eur. J. Biochem. 2, 309 38. HENIUILT, P., ~NI) DC~RNER, F. (1971) J. Riol. Chem. 246,

20. VI.;RPOORTE. J. A.. .uvu LINDI~LOX~-, C. (1968) J. Biol. Chem.. 2714

243, .5993&00 ’ 39.

LISDSKOG, S. ASD MGMSTRiiM, B. G. (1962) J. Rid. Chem. 237, 1129-1137

BRAUNS, D. II. (1925) J. Amer. Chem. Sot. 47, 1297 KANIXL, S. I., WONG, S.-C. C., K~NDEL, M., AND GORN.U,L,

A. G. (1968) J. Biol. Chem. 243, 2437-2449 WILrxu~, K. M., .~ND ANDERSON, N. G. (1948) J. Bid. Ch,em.

176, 147 QVENSSON, IT. (1961) Acta Chem. Scud. 16, 325 RICKLI, E. R., C:HU.iNFaR, S. A. S., GIBBONS, 13. H., END

EDS.~LL, J. T. (1964) J. Biol. Chem. 239, 1065-1078 Ihs, C. H. W. (1956) J. Biol. Chem. 219, 611 M~TSUB~R~, H., AND SX+XKI, R. M. (1969) Biochem. Biophys.

Res. Commun. 35, 175 SPACXM.IN, D. H., STEIN, W.H., AND MOORE,~. (1958),4nu!.

Chem. 30, 1190 LIXDSICOG, S., dsz) THORSLUND, $. (1968)Eur. J. Biochem. 3,

453 LINDSKOG, S. (1963) J. Biol. Chem. 238, 945; (1966) Biochem-

istry 5, 2641 FUN.ZKOSHI, S., ASD DEUTSCH, H. F. (1969) J. Biol. Chem. 244,

3438-3446 VERPOORTE, J.A., MICHTA,S.,.IND EDS.ILL, J.T. (1967) J.Biol.

Chem. 242, 4221-4229 WIIITNEY, P. L. (1970) Eur. J. Biochem. 16, 126 KING, R. W., .~ND ROBERTS, G. C. K. (1971) Biochemistry 10,

558 BR.IDBURY, S. L. (1969) J. BioE. Chem. 244, 2010 ANDERSSON, B., NYXIN, P. O., .~ND STRID, L. (1969) in CO:

Chemical, Biochemical and Physiological Aspects, (FORSTICR, II. E., EDS.~LL, J. T., OTIS, A.B., AND ROUGHTON, F. J. W., eds) National Aeronautics and Space Administ,ration, Washington, D. C.

KEILNOEUN, J. C. (1964) Biochim. Bioph~~.~Acta 81, 346

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from

Show-Chu C. Wong, Stephen I. Kandel, Marianne Kandel and Allan G. Gornall-Bromoacetylacetazolamide

NCovalent Labeling of the Active Site of Human Carbonic Anhydrase B with

1972, 247:3810-3821.J. Biol. Chem. 

  http://www.jbc.org/content/247/12/3810Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/247/12/3810.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on January 19, 2020http://w

ww

.jbc.org/D

ownloaded from