A comparative Raman spectroscopic study of cholinesterases

A comparative Raman spectroscopic study of cholinesterases

Biochimie ( 1991 ) 73, 1375-1386 © Soci6t6 franqaise de biochimie et biologie mol6culaire / Elsevier, Paris 1375 A comparative Raman spectroscopic s...

1MB Sizes 0 Downloads 51 Views

Biochimie ( 1991 ) 73, 1375-1386 © Soci6t6 franqaise de biochimie et biologie mol6culaire / Elsevier, Paris

1375

A comparative Raman spectroscopic study of cholinesterases D Aslanian 1, P Grot a*, S Bon 2, P Masson 3, M N6grerie 1, JM Chatel 1, M Balkanski 1, p Taylor 4, j Massouli62 tLaboratoire de physique des solides, Associd au CNRS, Universitd Pet M Curie, 4, place Jussieu, 75252 Cedex 05, Paris; 2Laboratoire de Neurobiologie, ENS. 75005 Paris; 3Unitd de Biochimie, CRSSA, 38702 La Tronche, France; 4Department of Pharmacology, University of California at San Diego, La Joila, CA 92093, USA

(Received 10 July i 99 !; accepted 27 September 1991)

Summary - - We report Raman spectra of various cholinesterases: lytic tetrameric forms (G4) obtained by tryptic digestion of asymmetric acetylcholinesterase (ACHE) from Torpedo californica and Electrophorus electrwuo, a PI-PLC-treated dimeric form (G2) of AChE from T marmorata, and the soluble tetrameric form (G4) of butyrylcholinesterase (BuChE) from human plasma. The contribution of different types of secondary structure was estimated by analyzing the amide I band, using the method of Williams [ l ]. The spectra of cholinesterases in 10 mM Tris-HCl (pH 7.0) indicate the presence of both or-helices (about 50%) and ~sheets (about 25%), together with 15% turns and 10% undefined structures. In 20 mM phosphate buffer (pH 7.0), the spectra indicated a smaller contribution of ix-helical structure (about 35%) and an increased ~-sheet content (from 25 to 35%). This shows that the ionic milieu profoundly affects either the conformation of the protein (ACHE activity is known to be sensitive to ionic strength), or the evaluation of secondary structure, or both. In addition, we analyzed vibrations corresponding to the side chains of aromatic and aliphatic amino acids. In particular, the analyses of the tyresine doublet (830-850 cm-I) and of the tryptophan vibration at 880 cm-I indicated that these residues are predominantly 'exposed' on the surface of the molecules. Raman spectroscopy / acetylcholinesterase / butyrylcholinesterase Introduction Raman spectroscopy provides useful information on the conformation of molecules and particularly proteins in solution [1, 2]. This method has been applied to explore the interactions of acetylcholine with its receptors [3], and a preliminary examination of the R a m a n spectrum of the lytic tetrameric form (l IS) of acetylcholinesterase from Torpedo californica has been presented previously [4]. In the present work we have analyzed Raman spectra of several types of cholinesterases in order to obtain further structural details on the native state of these proteins. Acetylcholinesterase (ACHE, EC 3.1.1.7) and butyrylcholinesterase (BuChE, EC 3.1.1.8) exist as various oligomeric molecular forms [5, 6]. We studied A C h E molecules derived from asymmetric forms or from amphihilic dimers. The asymmetric forms consist of *Permanent address: Institute of Biophysics, Semmelweis Medical University, Budapest, Hungary. Abbreviations. ACHE: acetylcholinesterase (EC 3.1.1.7); BuChE: butyrylcholinesterase (EC 3.1.1.8); Tyr: tyrosine; Trp: tryptophan; Phe: phenylalanine; PI-PLC: glycophosphatidylinositol specific phospholipase C.

catalytic tetramers, attached through disulfide bonds to the strands of a collagenic tail; non-amphiphilic tetramers (G4na) m a y be released from asymmetric form by proteolysis. The amphiphilic dimers (G2a) possess a C-terminal glyco-phosphatidylinositol hydrophobic anchor. The anchor m a y be cleaved by specific phospholipases (PI-PLC), producing nonamphiphilic dimers (G2na). The electric organ of Electrophorus electricus contains almost exclusively asymmetric forms, and that of Torpedo marmorata or T californica contains roughly the same amount of asymmetric forms and glycolip;d-anchored dimers. A C h E catalytic subunits from both Torpedo species have been cloned and show only minor sequence differences (12 amino acids) [7-9]. The subunits of the asymmetric and glycolipid-anchored forms are generated by differential splicing from the same gene. T h e y possess a major c o m m o n catalytic domain (535 amino acids), and differ only in their C-termini. In Torpedo ACHE, the catalytic subunits are linked as dimers by disulfide bonds and each m o n o m e r contains three intra-subunit disulfide loops [ 10]. The major component of human plasma BuChE is a globular tetrameric form (G4"a) organized as a dimer of dimers [ l I, 12]. As in ACHE, the subunits of each

D Aslanian et al

1376

d i m e r are linked through a single disulfide bridge and contain three disulfide loops [!1]. The c o m p l e t e a m i n o - a c i d sequence o f the m o n o m e r w a s d e t e r m i n e d by protein s e q u e n c i n g [13], and s u b s e q u e n t l y also deduced from the sequence o f c o r r e s p o n d i n g c D N A s [14, 151. Its C-terminal peptide s e q u e n c e is h o m o l o gous to that o f Torpedo A C h E subunits f r o m a s y m m e tric forms [6]. To date, the locations o f 9 a s p a r a g i n e linked c a r b o h y d r a t e chains and the 3 internal disulfide bridges within the sequence are k n o w n , as well as the location o f the interchain disulfide bridge, .which is located near the C - t e r m i n u s [14]. Since m o n o m e r i c subunits and a disulfide-containing peptide are generated upon action o f trypsin [171, the C - t e r m i n a l peptide m a y be e x p o s e d to the solvent. The cysteines involved in the disulfide bridges are in identical positions within the s e q u e n c e o f Torpedo A C h E [101. Thus, the three disulfide loops h a v e exactly the s a m e length in both e n z y m e s . In addition, h y d r o p a t h y profiles o f h u m a n B u C h E a n d Torpedo A C h E are virtually superimposable. This s u g g e s t s that the secondary and tertiary structures o f the two e n z y m e s are similar. Recently, crystals o f Torpedo A C h E have been obtained [ 18], and the t h r e e - d i m e n sional structure o f this e n z y m e has been d e t e r m i n e d

II91. We p_resent a c o m p a r a t i v e evaluation o f the fractional contributions o f the different types o f s e c o n d a r y structures and an analysis o f the e n v i r o n m e n t o f aromatic and aliphatic side chains, in t w o different buffers.

Materials

Purifications of AChE The lytic I IS form of T californica was purified and assayed according to a previously published procedure 1251. The asymmetric forms of Electrophortls electricus AChE were purified by affinity chromatography, as described previously [26]. They were separated in sucrose gradients, concentrated by isopycnic centrifugation in a CsCI density gradient, lyophilized, and digested by pronase. The resulting G4 form was isolated successively in a gradient and in a moleculac sieve column (Biogel A 1.5 m). It was then concentrated by isopycnic centrifugation in CsCI, to about 100 OD units/min per ml. It was finally concentrated and filtered in a Centricon, to about 2500 OD units/min per ml, in 100 mM NaCI, 40 mM MgC! 2, 50 mM Tris-HCl, pH 7.0. Torpedo electric organs were first homogenized without detergent in 10 volumes of buffer, and the resulting low salt soluble (LSS) extract was discarded. A subsequent detergent soluble (DS) extract, obtained in 3 voJumes of i% Triton X100, 50 mM MgC! 2, 50 mM Tris-HCl, pH 7.0, was diluted with an equal volume of buffer without detergent, and adsorbed in batch with N-methylacridinium Sepharose (5 times 50 ml of extract with 250 ml of gel). The gel was then washed with 4 volumes (1 I) of buffer without detergent, and wit a 3 times 4 velumes of the same buffer with 0.8 M NaCI. AChE was then eluted with 4 volumes of 25 mM edrophonium chloride, 0.1% Triton X-100, 1 M NaCI. If) mM Mg~l_ ~t~ ml~! "r.-~_w~! pH 8.0. The yield was 30 to 40%, relative to the DS extract. The purified amphiphilic G, AChE ~ls concentrated in a Centricon 30. A non-amphiphilic G 2 form of Torpedo AChE was obtained by treatment with Bacillus thuringiensis PI-PLC: 700 lal of AChE (about 2 mg/ml) were incubated with 70 ~ti of PI-PLC (45 min, 37°C). The resulting non-amphiphilic G2 form (7S) represented at least 90% of the enzyme and was separated from the residual aative amphiphilic form (6S) in a sucrose gradient containing 1% Brij 96 (Beckman SW41 rotor, 40 000 rpm, 8°C, 18 h). The 7S form was concentrated by isopycnic centrifugation. The enzyme was further concentrated in a Centricon 30, in I0 mM MgCI,, 50 mM Tris-HC! pH 7.0, to about 700 OD units/min per ml, ie 10 mg/ml. Isopycnic centrifugation was performed in CsC! solution (35 g CsCI for 100 ml of extract) in a SW 60 rotor, at 47 000 rpm, 10°C, for 78 h. ,

~ -m - ~l l U IlIlCILIIILIIU.N . . . . . . . a

Materials All reagents were of analytical grade. Glycophosphatidylinositol specific phospholipase C (PI-PLC) from Bacillus thuringiensis was obtained from Sapporo Breweries Ltd, Funakoshi Pharmaceutical Co (Tokyo, Japan). Edrophonium chloride was a gift from Hoffman-La Roche (Nutley, NJ, USA). Live Torpedo marmorata were obtained from the Marine Biology Laboratory of Arcachon, France. Live Electrophorus electricus were purchased from Worldwide Scientific Animals (Apopka, FL, USA). Human plasma was obtained from the Centre de Transfusion Sanguine de Lyon (France). The N-methylacridinium affinity iigand was synthesized as described by Dudai and Silman [201 and by Vailette et al [21] by Alcanes Entreprise, University Louis Pasteur (Strasbourg, France). Concentrating devices Centricon 10 and 30 were obtained from Amicon (Danvers, MA. USA).

Cholinesterase assay, protein concentration AChE et al iodide buffer

tivity producing an increase in optical density of 1 OD unit per rain in I ml of assay medium (1 cm pathlength), and corresponds to the hydrolysis of 75 mmol of substrate per min. The activity of 1 mg of pure enzyme corresponds to approximately 50 Ellman units in the case of Torpedo AChE and 130 EIIman units in the case of Electrophorus AChE [23]. Protein concentrations were determined by the Pierce BCA (bicinchoninic acid) protein assay (Pierce Chemical Co Rockford, IL, USA), or by the method of Lowry et al [ 19], with bovine serum albumin as the standard.

and BuChE activities were assayed according to Ellman 122], using acetylthiocholine and butyrylthiocholine (1 mM) as substrates, respectively, in 0.1 M phosphate (pH 7.0) at 25°C. One Ellman unit is defined as the ac-

-

.

.

.

.

.

.

.

.

.

.

.

.

2 .

.

.

.

.

S i l l

/

11t~

I

i

~

i

Purification of BuChE The tetrameric form of BuChE was purified from plasma obtained from the blood bank. The two-step pt, rification method of Lockridge and La Du [27] was used; ie chromatography on DEAE-cellulose (Whatman) at pH 4.0 and 4°C followed by affinity chromatography with procainamide as ligand. The experimental conditions were essentially as described, except that lS-mercaptoethanol was omitted in the buffers.

Raman spectra of cholinesterases The BuChE preparation we obtained had a specific activity of 400 mmol butyrylthiocholine hydrolyzed rain -I mg -I. The concentration of active sites was 10 laM. In non-denaturing electrophoresis [281, only one band corresponding to the tetramer was observed after activity staining with butyrylthiocholine as substrate [291. Silver staining of these gels showed the tetramer and several minor fast-migrating bands. SDS-gels [30] showed the enzyme monomer (85 kDa), a covalent (nonreducible) dimer which represents about 10% of the enzyme, and several minor contaminants of low molecular weight. The titration and densitometric analysis of gels indicated a purity of 80%. This preparation was extensively dialyzed either against 20 mM phosphate buffer, pH 7.0, containing I mM EDTA or against 10 mM Tris-HCl, pH 7.0, containing 1 mM EDTA, using a Centricon 10. The enzyme was !hen concentrated 10- to 40-fold in a Centricon 10. Titration of the active enzyme was carried out using the carbamyl ester reagent N-methyl (7-dimethyl carbamoxy) quinolinium iodide (Molecular Probes, Eugene, OR, USA) according to Mooser et al [261. 7-Hydroxyl-N-methylquinolinium, prepared according to Prince [321, was used as the standard. A comparison of the concentration of active sites with the concentration of protein indicated that the preparation consisted of at least 95% active BuChE.

Raman spectroscopy Raman spectra were recorded with a Jobin-Yvon Ramanor U 1000 double monochromator, under the control of an IBM XT 286 computer. Aliquots (10 lal) of concentrated enzymes (10-15 mg protein/ml, corresponding to about I-2 10-4 M monomers) were introduced into capillary tubes of I mm internal diameter. After sealing, the tubes were put vertically into a brass holder, thermostated at 10°C by a thermoelectric cooling system. The samples were excited with an argon laser (Coherent Innova 90-3) tuned at 488 rim, with an output of 140 mW at the sample level. The spectral slits were 500/320 rim. The recorded spectra were computer averaged over 25-30 scans. Scan speed was 1 cm-J/s. All averaged spectra were corrected for solvent background by subtraction of the buffer (and water) spectrum as well as the fluorescence background. The spectra were smoothed by Fourier transforms. The secondary structures of the enzymes were quantitatively estimated by the method of Williams l I, 2]. In this method, the solvent spectrum and aromatic side chain bands are first subtracted from the amide I region. The procedure of constrained iterative deconvolution [33, 34] improves considerably the resolution of the aromatic amino acid ring vibrations which had to be subtracted. For the subtraction, either decomposition [361 or deconvolution [33, 341 were applied. In the latter case, the lines coming from vibrations of aromatic rings were deleted from the deconvoluted spectra. Thereafter, spectra were reconvoluted with the same linewidth as used for deconvolution. Comparing the results obtained by the two methods, we found that there is less than !% difference in secondary structure contributions. The amide I spectrum is then fitted with a linear combination of amide I spectra of standard proteins whose structure is known from X-ray diffraction. The contribution of the various types of secondary structures are deduced from the fitting coefficients. The region of the amide i band also contains contributions from other vibrations. For example the v(C=O) vibration of the carboxylate groups of glutamic and aspartic acids appears at frequencies of 1690-1695 cm -I. These vibrations overlap with the amide I region (1600-1700 cm-I), and thus interfere with

1377

the determination of the secondar'q structure Bec:~use of this limitation, the spectral contribution:, corresponding, to different secondary structures, according t,9 the method of Williams 121. are evaluated with an accuracy of + 5%. The band located at 1450 + 5 cm -t, which is due predominantly to the CH_, scissoring mode from methylene groups of the side chains, was used as an internal standard to normalize the intensities of other lines [35-371. Due to the very high and similar number of methylene groups in the samples, peak height and band area are !east affected by perturbations of structure.

Results We studied the f o l l o w i n g e n z y m e s : lytic tetrameric f o r m s (G~) obtained by tryptic digestion o f the a s y m metric f o r m s o f Torpedo californica (fig l a, b) and Electrophorus electricus A C h E (fig 2a, b): a P I - P L C treated dimeric form (G2) o f A C h E from T marmorata (fig 3a, b) and the tetrameric form (G4) o f B u C h E f r o m h u m a n p l a s m a (fig 4a, b). The solutions contained 10-15 m g / m l o f protein, c o r r e s p o n d i n g to a c o n c e n t r a t i o n o f 1-2 10 -4 M in catalytic subunits, in 10 m M Tris-HCl, pH 7.0 (figs 1--4), and in the case o f B u C h E and of the a m p h i p h i l i c dimeric f o r m (G2) from T marmorata, also in 20 m M p h o s p h a t e buffer, p H 7.0 (figs 5 - 6 ) . The R a m a n spectra are plotted after subtraction o f the b u f f e r spectrum and fluorescence b a c k g r o u n d , and s m o o t h i n g . A s s i g n m e n t s o f freq u e n c i e s are given in figure 1.

Quantitative estimation of the secondary structure of "holin esterases T h e a m i d e I region is g e n e r a l l y used for the quantitative evaluation o f the s e c o n d a r y structure o f proteins. R a m a n spectra (figs I a - 6 a ) s h o w the strong and broad a m i d e I b a n d c e n t e r e d b e t w e e n 1653 and 1663 cm -~. D e c o n v o l u t i o n o f spectra in the region 1 5 0 0 1800 c m -~ (separate p a p e r in preparation) i m p r o v e s the resolution o f the a r o m a t i c a m i n o acid vibrations n e a r 1550, 1580, 1602 and 1616 cm -~. These aromatic side chain vibrations overlap the a m i d e I band ( 1 6 4 0 - 1 7 0 0 cm-~), w h i c h m a y be a n a l y z e d after their subtraction (dotted lines in figures l a - 6 a ) . The s a m e region al~o contains v ( C = O ) vibrations o f the c a r b o x y l a t e groups o f acidic a m i n o - a c i d s , w h i c h interfere with the analysis o f the amide I vibrations. As s h o w n in table I, cholinesterases o f different origins in 10 m M Tris-HCl pH 7.0 contain on a v e r a g e 4 9 % total or-helical structure, about 3 6 % o f which represent ordered (o~0) a n d 14% disordered (Ctd) helix. T h e total ]I-sheet content has been estimated to be on a v e r a g e 23%, a l m o s t entirely anti-parallel (l~p). Further, we f o u n d a b o u t 14% reverse turns (T) and 10% undefined structure (U).

i 378

D Aslanian et al

AI Phe

!

0 o

E} E)

AI

,,CH2 OH31 I AIII

T.rp Tyr I

"

'=



Phe

Trp



Tyr

to

1 '

I1 O

7(JO

~ rn

'

~

tD

m

Phe (u to

Lq

Pmm



/ (13

960

or]

~.tbo

OD

I

113 0 I

: .5~001

I

'

15100 . . . .

16b

;

I

I I i 17 0 V cm "1

Fig 1. a. Raman spectrum of AChE (G4) from T californica in 10 mM Tris-HCl, pH 7.0. b. Amide I region after subtraction of aromatic amino acid ring vibrations ( ...... ). In 20 m M phosphate buffer pH 7.0, the Ga form of BuChE (fig 5a) and the amphiphilic G 2 form of AChE from T marmorata (fig 6a) present a smaller s-helical structure contribution: 39 and 34% respectively. Conversely, the ~-sheet contribution is higher, 31 and 36%. For the enzymes studied, we found that, in general, the ordered o~-helix ((:to) is more prevalent than the disordered structure (s~). The only exception is the amphiphilic (G2) AChE which shows equal fractions for so and s 0 structures in phosphate buffer (table I). Characteristic side chain vibrations Amino acid side chain environment Tyrosine. Raman spectra are particularly informative about H-bonding of the tyrosine OH group. The intensity ratio (I850/1830) of the tyrosine doublet near

8 3 0 - 8 5 0 cm -~ indicates the state of tyrosine residues in proteins: 'exposed' on the surface of the molecule, or 'buried' in the hydrophobic regions [36--39]. This ratio is 2.5 for 'exposed' tyrosines (the phenol ring acts as an acceptor with strong H-bonds); it is 0.3 for 'buried' tyrosines (the phenol ring act~ as a donor with strong H-bonds) [38]. The ratio is 0.9-1.3 for normally 'exposed' acceptor/donor tyrosines with moderate H-bonds. The intensity ratios of tyrosine doublets in the cholinesterase spectra are given in table II. The values obtained, between 0.9 and 1.2, indicate that the tyrosine residues mostly reside on the surface of the enzymes, 'exposed' to the aqueous medium. Tryptophan. The state of tryptophan residues can be analyzed from vibrations near 1360 cm -l and 880 cm -1, which are sensitive to the immediate environment [35, 39-42]. The vibration near 1360 cm -~

Raman ,,;pectra of cholinesterases

1379

a

f

g,. 03

I'~

R

70

IIo

A

~

I Io 1

I%.

o,

113

.,

o~

"

o,

i",-

O)

o~

o.9, ~

90

II 0

i30

1171o0

II . , , ' . . . . 1500

,

I J""i 16bo J

, '

I

1171o ° I

V cm "1

Fig 2. a. Raman spectrum of AChE (G 4) from E electricus in 10 mM Tris-HCl, pH 7.0. b. Amide I region after subtraction of aromatic amino acid ring vibrations ( ...... ).

forms a sharp peak when the tryptophans are 'buried', and its intensity is lower when they are ' e x p o s e d ' [40]. The tryptophan residues in cholinesterases (fig l a-6a) seem to be i aostly ' e x p o s e d ' because the vibration near 1360 cm -~ appears as a very weak shoulder. This contrasts with the fact that this vibration stands out as a sharp peak in the spectrum of ot-chymotrypsin in which the tryptophans are buried (data not shown). Kitagawa et al [35] have shown that the vibration near 880 cm -I is also conformationally sensitive and can serve as a probe to the environment of tryptophan residues: the ratio 1875//~445 is low when tryptophan residues are exposed. This ratio varies between 0.85 (BuChE in phosphate buffer) and 0.51 (amphiphilic A C h E in phosphate buffer). Phenylalanine. The most intense peak near 1003 cm -~ (fig l a - 6 a ) arises from phenylalanine [39] and indi-

cates a high proportion of these aromatic residues. This vibration of phenylalanine is not sensitive to the conformation of the protein [39-421.

Vibrations of aliphatic side chains Frequencies of the skeletal (C-C) and (C-N) stretching vibrations (800-1150 cm-'). Although a detailed assignment is very difficult in this region, Thomas et al [46] showed that the pair of lines near 943 and 960 cm -I correlates rather well with the total valine plus leucine content in the viral proteins fd, Ifl, IKe, Pfl and Xf. Pezolet et al [47] reached the same conclusion in Raman investigations of human immunoglobulin G. They interpret the band at 939 cm -I as a vibration which arises from the (C-C) stretching mode of the side chains. The (C-C) and (C-N) stretching modes of other side chains are also assigned in this region [46]. Thus, the vibration near 905 cm -1 is

D Aslanian et al

! 380

b

a

113 (13 i

m

ftl O1 £kl ('LI ~ m

mu.ktOOo

12o t'rj

O3 0 0 ol

('tl

0

q I 7

o

I

I 9"o II0

I

I 1 I I 1 lbo

I

I

13bo

I

I

15bo"

I

I 17 ~)0 I

1500

16100

17 V cm -1

Fig 3. a. Raman spectrum of G4 na form of AChE T marmorata in 10 mM Tris-HCI, pH 7,0~ b~ A_m_ide! region after subtraction of aromatic amino acid ring vibrations ( ...... ).

correlated with alanine, 931 cm -~ with lysine, glutamic acid and serine, 1057 cm -~ with alanine. The vibration near 1120 cm -~ is most probably a v(C-C) mode related to the optical skeletal mode of polymethylenes [481. The contribution of the CH 3, CH.~ and NH:* rocking vibrations in this region is not expected to be detectable because their Raman intensities are extremely weak [39]. The skeletal frequencies in a protein are expected to be conformationdependent [49]. The spectra of cholinesterases show similar skeletal regions with six frequencies situated near 897, 932, 950, 1078, 1103 and 1124 cm -~. Thus, the exposition and environment of the (C-C) and (C-N) bonds of the side chains of these enzymes appear to be similar, especially the tetrameric forms of T marmorata and T californica AChEs.

Methylere (CH2) and methyl (CHO vibrations. The band in the 1300-1340 cm -I interval is mainly due to the CH2 bending mode. All n-alkanes (n >> 3) yield an intense Raman vibration near 1300 cm -I, assigned to CH2 wagging or twisting motions [50]. The frequency and intensity of this band are sensitive to gauchetrans isomerization, chain branching and other substitutions [51, 52]. The CH z twisting/wagging vibration appears as a very intense band at 1337 cm-~ in the spectra of AChEs from T californica and Electrophorus electricus and BuChE, and at 1329 cm -~ in the spectrum of amphiphilic ACHE. Its intensity is lower, and it is accompanied by a new intense peak at 1293 cm -~ in the spectrum of the dimeric (G2) form of A C h E from T marmorata. The stretching vibrations of CH2 and CH 3 groups appear in the region 2700--3120 cm -i. They present an

Raman spectra of cholinesterases

a

138 !

b

t'M 0 0

I.~ I1") (C)

fr• I",.

Lt)

O

O

113

.

'

/t

0rl

I~

trl

.

C

V c m "1

Fig 4. a. Raman spectrum of BuChE (G4) from human plasma in Tris-HCI (10 mM, pH 7.0). b. Amide I region after subtraction of aromatic amino acid ring vibrations ( ...... ).

intense band with a maximum at 2931 cm -~ (a mixture of CH3 symmetric and CH2 asymmetric vibrations) and a shoulder near 2872-2880 cm -! (CH 2 symmetric vibrations), and appear to be similar in all spectra (data not shown).

Discussion The known primary structures of several cholinesterases [7-9, 13-15], as well as biochemical studies on their association as oligomers and interactions with membranes and other cellular components (reviewed in [6]) have provided impetus to investigate the submolecular architecture of these molecules, using such sensitive methods as Raman spectroscopy. The recent improvements of data acquisition and data handling in Raman spectroscopy allow for a more

detailed evaluation of the Raman spectra. This method may thus reveal similarities and differences among different cholinesterases of similar overall sequence, and may lead to an investigatton of interactions of the enzymes with their substrates or inhibitors, to study the structural effects of chemical modifications (eg at the active center by organophosphate inhibitors) and of amino acid substitutions produced by site-directed mutagenesis. The stability of cholinesterases depends on environmental conditions, not only on pH and temperature but also on the composition of the medium [53-58]. In the present study our goal was to compare different cholinesterases in the same buffer and to study possible differences in the spectra of the same enzyme in two different buffers. The primary sequences of cholinesterases are clearly homologous, with a large degree of amino acid

! 382

D Aslanian et al

b

a

Od tO I/3 (13

t'kl .tD

O

11'3 ~r oB

7(~0

'

'

9(~0

'

~ ~I lbo I

/

1300

1500

17DO V em-1

Fig ~ o ~ . . . . . . . Ir ~ , m .~" n,,,",hc: ~ . . . . I~..~41x c.~-lilt.pill k H U. I.I .l a.l .i . p i a, ..~.l l.l .a subtraction of aromatic amino acid ring vibrations ( ...... ).

identity, suggesting that all these proteins assume very similar secondary and tertiary conformations [5c~]. Only a small C-terminal peptide region, determined by differential splicing, differs between the catalytic subunits of glycolipid-anchored molecules and the heteropolymeric forms (eg asymmetric forms). This represents less than a tenth of the length of the poly.peptide chain, aiad should have little effect on the Raman spectrum. Indeed, the evaluation of the participating secondary structures by the method of Williams [l, 2] showed similar values for the four cholinesterases studied in l0 m M Tris-HCl-buffer at pH 7.0 (table I). However, in 20 m M phosphate buffer pH 7.0 (table I), the spectra indicated a lower contribution of helical structure and a higher one of 13sheets, for the two e,,zymes studied. Recently, the three-dimensional structure of Tot'pedo AChE has been determined by X-ray crystallo-

:_ ...I.U..b..l ... l.l .I gll|C . ._ . I l l -,n / U raM .1.,1,

UI.II 1~1

17

(pH .t.w). . . . . . .b.. . .~ m m e i re= ion after

graphy [ 19]: the proportion of residues participating in t~-helices and in 13-sheets is 32 and 15%, respectively. These values are consistent with the estimates obtained from circular dichroism for a lytic G4 form and the G2a form of A C h E from T californica in phosphate buffet [60]. The estimates obtained in the present study are somewhat higher (about 50 and 25% in Tris buffer), but there is a general agreement that both types of secondary structures coexist in cholinesterases, with an ~13 ratio of about 2. The differences observed between the spectra obtained in Tris-HCl and phosphate buffers were unexpected. This difference is reproducible; it was similar for Torpedo A C h E (G2a) and human BuChE (Gana). According to current ideas about protein structure and function, the three-dimensional structure of an active protein is not expected to present such a flexibility as a function of the buffer. In particular, the

Raman spectra of cholinesterases a

1383

b (13 tO tO

,q. 13)

'q"

or)

tO

8 •~ tO tO

O tD A

~

I

I

t

I

~o

I loo' I ' .Joo 'l

:"

16~)0

17D0

V cm -1

LIIIILI~ z l t ~ .. l ... ~ u.L r"-" Fig 6. a. Raman spectrum of T m a r m o r a t a AChE (G2a) in 20 mM phosphate . . . . . . (pH " n, b • / ""_:.4~ I -~-:~" ("~" traction of aromatic amino acid ring vibrations ( ...... ).

Raman spectra of reference proteins used by Williams to define the contributions of the various types of secondary structures were obtained in a number of different media [ 1, 2]. However, it has been known for a long time that the activity of cholinesterases varies significantly with ionic strength [61]: the rate of hydrolysis, with a saturating concentration of substrate (3 mM acetylthiocholine), was about 10% higher in the 20 mM phosphate buffer than in the 10 mM Tris-HCl buffer, and the difference was probably due to ionic strength since it could be more than reversed by addition of 40 mM NaCI to the Tris-HCl buffer (data not shown). A possible explanation for the difference observed may be that the enzymes become partially denatured during the recording of the spectra, in the laser beam, and this would be expected to reduce the contribution of oc helix. Denaturation, however, would most probably be more extensive in the Tris-HCl than in the phosphate buffer, since

BuChE is more thermally stable in the latter (P Masson, unpublished results), in contradiction with the observed effect (table I). It is thus likely that the Raman spectra reflect not only structural changes in the active protein, but also differences in their interactions with the solvent. If this is so, the solvent conditions must be taken into account precisely to evaluate secondary structures, and this would question the validity of the method used here. The influence of ionic conditions on Raman spectra will be investigated more systematically in a subsequent study. In addition to the information obtained on the secondary structures, the Raman spectra provided information about the environment of aromatic amino acids, the (C-C), and (C-N) bonds, and the (CH,_) and (CH3) groups of the aliphatic side chains. Three regions reflect alterations in the environment of side chains corresponding to vibrations of tyrosine and

1384

D Aslanian et al

Table I. A quantitative estimation of structural components of the peptide chain of the cholinesterase enzymes, t" total; T-turns; U: undetermined. G2na: non-amphiphilic dimers, obtained after PI-PLC digestion; G2a: amphiphilic glycolipid anchored dimer. Structure (%) Enzyme

ott ( oto or,i)

[3 ( [3,~t,[3r)

T

U

49 49 51 47

23 25 23 26

I! 14 14 16

16 10 11 10

39 34

31 36

17 18

11 13

10 mM Tris-HCl, pH 7.0 T ca/ifornica AChE (G4) Electrophorus AChE (G4) T marmorata AChE (G~ha) Human plasma BuChE (G4)

20 mM phosphate buffer, pH 7.0 Human plasrna BuChE (G 4) T marmorata AChE (G2a)

Table !I. Relative Ram,'m intensities of 830 cm -I and 875 cm -a vibrations. I. non-amphiphilic; .-. ~' amphiphilic form.

!~5J1,~'3o

l,~7s/l t445

1.2 I. i !.0 1.0

0.6 0.8 0.6 0.5

1.0 0.9

0.9 0.5

(Tyr)

(Trp)

10 mM Tris-HCi, pH 7.0 T californica AChE (fig la) Electrophorus AChE (fig 2a) T marmorata AChE I (fig 3a) Human plasma BuChE (fig 4a)

20 mM phosphate but'fer pH 7.0 Human plasma BuChE (fig 5a) T marmorata AChE 2 (tig 6a)

tryptophan (800-890 cm-t), skeletal optical region (900-1150 cm-~), and methylene and methyl groups (1300-1340 cm-:). Our spectra indicate that tyrosines and tryptophans are largely exposed in cholinesterases, although it is not possible to quantify the proportions of exposed and buried residues. This is quite consistent with the three-dimensional structure of Torpedo ACHE, which shows that aromatic side chains line the wall of a deep gorge in which the active site serine is located, as well as a region of the surface adjacent to the opening of this gorge. These exposed ,aromatic residues are thought to constitute the predominant element of the acetylcholine-binding site [ 19]. The conformation of the backbone or the side chains of a protein are sensitive to the internal and

external forces acting on the molecule, Thus, any change of the ionic milieu, small differences in the primary structure or the presence of an interacting molecule can result in modifications of side chains and/or skeletal vibrations. As shown with 132-glycoprotein [62], carbohydrate units also act upon the secondary structure of glycoproteins. Human plasma BuChE contains 24% of carbohydrates, with 9 asparagine-linked carbohydrate chains [ 13], while Torpedo AChE has 4 glycosylation sites 1631. In addition, reaction with an anti-carbohydrate antibody, Elec-39, has shown that Electrophorus ACHE, asymmetric and G2 forms of Torpedo AChE possess distinct carbohydrate epitopes [641. Thus, some of the observed differences may originate from differences in the carbohydrate moiety of the enz~,mes. In conclusion, the Raman spectra reveal similarities of secondary structures among cholinesterases of different origins. The ionic milieu, however, seems to influence the contributions of different structures as well as other sensitive vibrations, eg skeletal and aromatic amino acid ring vibrations. Such vibrations may be expected to be sensitive also to interactions of cholinesterases with their substrates or inhibitors, a question which will be examined later.

Acknowledgments We thank Dr O Lockridge (University of Nebraska Medical Center, Omaha, NE) for the gift of purified human plasma cholinesterase and Dr Israel Silman for useful discussions. This work was sup,ported by research grants from the Direction des Recherches, Etudes et Techniques (DRET no 88/199; 89/037; 89/257), from the Centre National de la Recherche Scientifique, and from the Institut National de la Sant6 et de la Recherche M6dicale.

Raman spectra of cholinesterases

References 1 Williams RW (1983) Estimation of protein secondary structure from the laser Raman amide I spectrum. J Mol Biol 166, 581-603 2 Williams RW (1986) Protein secondary structure analysis using Raman amide I and amide III spectra. Methods Enzymoi 130, 311-331 3 Aslanian D (1989) Raman and infrared study of acetylcholine and postsynaptic membranes. In: Molecules in physics, chemistry and biology (Jean Maruani, ed) Kluwer Academic Publishers, Vol IV, 233--280 4 Aslanian D, Grof P, N6grerie M, Balkanski M, Taylor P (1987) Raman spectroscopic study on the conformation of I1S form acetylcholinesterase from Torpedo californica. FEBS Letters 219, 202-206 5 Massouli6 J, Bon S (1982) The molecular forms of cholinesterases in vertebrates. Annu Rev Neurosci 5, 57-106 6 Massouli6 J, Toutant JP (1988) Globular and asymmetric forms of cholinesterases in vertebrates, h~: Cholinergic Synapse. Handbook of Experimental Pharmacology (Whittaker J, ed) Springer-Verlag, Vol 86, 167-224 7 Sehumacher M, Camp S, Maulet Y, Newton M, MacPheeQuigley K, Taylor SS, Friedman T, Taylor P (1986) Primary structure of Torpedo californica acetylcholinesterase deduced from eDNA sequence. Nature 319, 407--409 8 Sikorav JL, Krejci E, Massouli6 J (1987) eDNA sequences of Torpedo marmorata acetylcholinesterase: primary structure of the precursor of a catalytic subunit; existence of multiple 5'-untranslated regions. EMBO J 6, 1865-1873 9 Sikorav JL, Dural N, Anselmet A, Bon S, Krejci E, Legay C, Osterlund M, Reimund B, Massouli6 J (1988) Complex alternative splicing of acetylcholinesterase transcript in Torpedo electric organ; primary structure of the precursor of the glycolipid-anchored dimeric form. EMBO J 7, 2983-2993 10 MacPhee-Quigley K, Vedvick TS, Taylor P, Taylor SS (1986) Profile of the disulfide bonds in acetylcholinesterase. J Biol Chem 261, 13565-13570 11 Lockridge O, Eckerson HW, La Du BN (1979) Interchain disulfide bonds and subunit organization in human serum cholinesterase. J Biol Chem 254, 8324-8330 12 Masson P (1979) Formes mol6culaires multiples de la butyrylcholinest6rase du plasma humain. Biochem Biophys Acta 578, 493-504 13 Lockridge O, Bartels CF, Vaughan TA, Wong CK, Norton SE, Johnson LL (1987) Complete amino-acid sequence of human serum cholinesterase. J Biol Chem 262, 549-557 14 Prody CA, Zevin-Sonkin D, Gnatt A, Goldberg O, Soreq H (1987) Isolation and characterization of fulllength cDNA clones coding for cholinesterase from fetal human tissues. Proc Natl Acad Sci USA 84, 3555-3559 15 MacTiernan C, Adkins S, Chatonnet A, Vaughan T, Bartels C, Kott M, Rosenberry T, La Du BN, Lockridge O (1987) Brain cDNA clone from human cholinesterase. Proc Natl Acad Sci USA 84, 6682--6686 16 Lockridge O, Adkins S, La Du BN (1987) Location of disulfide bonds within the sequence of human serum cholinesterase. J Biol Chem 262, 12945-12952 17 Loekridge O, La Du BN (1982) Loss of the interchain disulfide peptide and dissociation of the tetramer following limited proteolysis of native human serum cholinesterase. J Biol Chem 257, 12012-12018

18

19

20 21

22

23

24 25

26 27 28 29 30 31 32 33 34 35

36

1385

Sussman JL, Harel M, Frolow F, Varon L, Toker L, Futerman AH, Silman I (1988) Purification and crystallization of a dimeric form of acetylcholinesterase from Torpedo californica subsequent to solubilization with phosphatidylinositol-specific phospholipase C. J Mol Biol 203, 821-823 Sussman JL, Harel M, Frolow F, Oefner C, Goldman A, Toker L, Silman I (1991) Atomic structure of acetylcholinesterase from Torpedo californica: a prototypic acetylcholine-binding protein. Science 283, 872-879 Dudai Y, Silman I (1974) Acetylcholinesterase. In: Research Methods in Neurochemisny Vol 3, 209-251 Vallette FM, March DM, Muller F, Massouli6 J, Marqot B, Viel C (1983) Comparative affinity chromatography of AChE from five vertebrate species acetylcholinesterases. J Chromatog 257, 285-296 EIlman GL, Courtney KD, Andres V, Featherstone RM (1961) A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharmacol 7, 88-95 Vigny M, Bon S, Massouli6 J, Leterrier F (1978) Activesite catalytic efficiency of acetylcholinesterase molecular forms in Electrophorus, Torpedo, rat and chicken. Em'J Biochem 85,317-323 Lowry OH, Rosebrough NJ, Fan" AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193,365-375 Taylor P, Jones JW, Jacobs NM (1974) Acetylcholinesterase from Torpedo califotv~ica: characterization of an enzyme species isolated by lytic procedures. Mol Pharmacol 10, 78-92 Massouli6 J, Bon S (1976) Affinity chromatography of acetylcholinesterase. Eur J Biochem 68, 531-539 Lockridge O, La Du BN (1978) Comparison of atypical and usual human serum cholinesterase. J Biol Chem 253, 361-366 Davis BJ (1964) Disc electrophoresis. I1. Method and application to human serum proteins. Ann NY Acad Sci 121,404--427 Karnovsky MS, Roots L (1964) A direct coloring thiocholine method for chohnesterase. J Histochem Cytochem 12, 219-222 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685 Mooser G, Schulman H, Sigman DC (1972) Fluorescent probes of acetylcholinesterase. Biochem J 11, 1595-1602 Prince AK (1966) Spectrophotometric study of the ACHEcatalyzed hydrolysis of l-methyl-acetoxyquinolinium iodides. Arch Biochem Biophys 113, 195-204 Agard DA, Steinberg RA, Stroud RM (1981) Quantitative analysis of electrophoretograms: a mathematical approach to super-resolution. Analyt Biochemistry 111,257-268 Thomas GI Jr, Agard DA (1984) Quantitative analysis of nucleic acidr,, proteins and viruses by Raman band deconvolution. Biophys J 46, 763-768 Kitagawa T, Azuma T, Hamaguchi K (1979) The Raman spectra of Bence-Jones proteins. Disulfide stretching frequencies and dependence of Raman intensity of tryptophan residues on their environments. Biopolymers 18, 451--465 Fasman GD, Itoh K, Lui CS, Lord RC (1978) Laser Raman spectroscopy of biomolecules. XI Conformational study of poly (L-valine) and copolymers of L-valine and Lalanine. Biopolymers 17, 125-143

1>~6 37 38

39 40

41 42 43 44

45

46

47 48

49 50 51

D Aslanian et al

Yu NT. Jo BH. O'Shea DC ~1973) Laser Raman scattering of cobramine B. a basic protein from cobra venom. Atz'h Biochem Biophys 156, 71-76 Siamwitza MN, Lord RC, Chen MC, Takamatsu T, Harada I, Matsuura H, Shimanuouchi T (1975) Interpretation of the doublet at 850 and 830 cm -I in the Raman spectra of tyrosyl residues in proteins and certain model compounds. Biochemistry 14, 4870-4876 Yu NT (1975) Laser P,~aman scattering of neurotoxins isolated from the venoms of sea snakes Lapemis hardwickii and Enhvdrina schistosa. J Biol Chem 250, 1783-1785 Miuva 1", Takeuchi H, Harada J (1988) Characterization of individual tryptophan side chains m proteins using Raman spectroscopy and hydrogen-deuterium exchange kinetics. Biochemisnw 27, 88-94 Lord RC, "~'u NT (1970) Laser-excited Raman spectroscopy of biomolecules. II. Native ribonuclease and 0tchymotrypsin. J Mol Biol 51,203-213 Chen MC. Lord RC (1976) Laser Raman spectroscopy studies of the thermal unfolding of ribonuclease A. Biochemistry 15, 1889-1897 Pezolet M, Pigeon-Gosselin M, Caill6 JL (1978) Laser Raman investigation of intact single muscle fibers. Protein conformations. Biochim Biophys Acta 533, 263-269 Pezolet M, Pigeon-Gosselin M, Nadeau J, Caill6 JP (1980) Laser Raman scattering. A molecular probe of the contractile state of intact single muscle fibers. Biophys J 31, 1-8 Thomas G Jr, Prescott B, Middaugh CR, Litman GW (1979) Raman spectra and conformational structures of Fabla and (Fc)5~t fra~mnents of cryoglobulin JgM-kMcE. Biochim Biophys Acta 577, 285-290 Thomas GI Jr, Prescott B, Day LA (1983) Structure similarity, difference and variability in the filamentous viruses fd, lfl, IKe, Pfl and Xf. Investigation by laser Raman spectroscopy. J Mol Bioi 165, 321-365 Pezolet M, Gosselin MP, Coulombe L (1976) Laser Raman investigation of the conformation of human immunoglobulin G. Biochim Biophys Acta 453, 501-512 Snyder RG, Schachtschneider JP (1963) Vibrational ,~.,,..y~.~. L,~ u,~ n-pas--~fins. I. Assignments of infrared bands in the spectra of Calls through n-CtgH40. Spectrochim Acta 19, 85-116 Mizushima S (1954) Structure of molecules and infi'ared rotations, New York Academic Press, Vol VI, 117 Snyder RG (1970) Vibrational study of the chain conformation of the liquid n-paraffins and molten polyethylene. J Chem Phys 47, 1316-1360 Bellamy LJ (1975) The infrared spectra of complex molecules, Wiley NY, Vol 1

52

Gaber BE Yager E Peticolas WL (1978) Interpretation of biomembrane structure by Raman difference spectroscopy. Biophys J 21,161-176 53 Stojan J, Pavlic MR (1988) The influence of medium on the stability of cholinesterases. Yugoslav Physiol Pharmacol Acta 24, 305-311 54 Edwards JA, Brimijoin S (1983) Thermal inactivation of the molecular forms of acetylcholinesterase and butyrylcholinesterase. Biochim Biophys Acta 742, 509-516 55 Wu C-SC, Gan L, Yang JT (1987) Conformational similarities of the globular and tailed forms of acetylcholinesterase from Torpedo californica. Biochim Biophys Acta 911, 25-36 56 Masson E Laurentie M (1978) Stability of butyryicholinesterase: thermal inactivation in water and deuterium oxide. Biochim Biophys Acta 957, 111-121 57 Payne CS, Saed M, Wolfe AD (1989) Ligand stabilization of cholinesterases. Biochim Biophys Acta 999, 4651 58 Masson E Balny C (1990) Conformational plasticity of butyrylcholinesterase as revealed by high pressure experiments. Biochim Biophys Acta 104 l, 223-23 l 59 Krejci E, Duval N, Chatonnet A, Vincens P, Massouli6 J (1991 ) Cholinesterase-like domains in enzymes and structural proteins: functional and evolutionary relationships and identification of a catalytically essential aspartic acid. Proc Natl Acad Sci USA 88, 6647-6651 60 Manavalan E Taylor P, Johnson WC Jr (1985) Circular dichroism studies of acetylcholinesterase conformation. Comparison of the 11S and 5.6S species and the differences induced by inhibitory ligands. Biochim Biophys Acta 829, 365-370 61 Myers DK (1952) Effect of salt on the hydrolysis of acetylcholine by cholinesterases. Arch Biochem 37, 469--487 62 Walsh MT, Watzlawick H, Putnam FW, Schmid K, Brossmer R (1990) Effect of the carbohydrate moiety on the secondary structure of 132-glycoprotein. I. Implications for the biosynthesis and folding of glycoproteins. Biochemiso~y 29, 6250-6257 63 MacPhee-Quigiey K, Drotar AM, Vedvick i-, Taylor P, Taylor SS (1987) Characterization of the glycosylation sites of acetylcholinesterase from Torpedo californica. Fed Proc 46, 2150 64 Bon S, M6flah K, Musset F, Grassi J, Massouli6 J (1987) An immunoglobulin M monoclonal antibody, recognizing a subset of acetylcholinesterase molecules from electric organs of Electrophorus and Torpedo, belongs to the HNK-I anticarbohydrate family. J Neurochem 49, 1720-1731