ANALYTICAL

201,43-47

BIOCHEMISTRY

(19%)

Deuterium Exchange on Micrograms of Proteins by Attenuated Total Reflection Fourier Transform Infrared Spectroscopy on Silver Halide Fiber Stella

M. Chiacchieral

Biophysical

Received

June

Organic

and Edward

Chemistry

Unit,

M. Kosower2

School of Chemistry,

Tel-Aviv

Press,

Tel-Aviv

69978, Israel

18, 1991

We illustrate the use of polycrystalline silver halide fibers (2-20 Frn transparency range) for attenuated total internal reflection Fourier transform infrared (IR) spectroscopic measurements of microsamples (10 pg of protein). A powerful adjunct technique is a simple method for carrying out deuterium for proton exchange. Spectra of trypsin, soybean trypsin inhibitor, and their complex are easily obtained. Two kinds of difference spectra (DS) are revealing: DSl (changes in protein on combination with ligand), IR of the trypsinsoybean trypsin inhibitor complex (T.SBTI complex) C [IR of trypsin (T) + IR of soybean trypsin inhibitor (SBTI)], the small values at all wavelengths indicating no conformational change of the proteins upon complexation, and DS2 (changes in materials on deuteration), IR of protioprotein - IR of deuterioprotein, which reveals the infrared bands affected by deuteration. The rate and the extent of the exchange are additional valuable parameters readily measured with this technique. In the present instance, the rate and the amount of the exchange for T.SBTI complex after 30 min was substantially less than that expected from the simple sum of the same parameters for the two individual proteins, T and SBTI. The enzymatic activity of trypsin on the fiber survived for more than a day, no autodegradation being detected by SDS-gel electrophoresis. o 1992 Academic

University.

Inc.

Infrared spectroscopy provides specific types of relevant information about protein structure. Furthermore, changes in conformation occurring upon binding differ’ Permanent address: Departomento dad National de Rio Cuarto, Estafete doba 5800, Argentina. ’ To whom correspondence should 0003-2697/92 $3.00 CopyrIght 0 1992 by Academic Press, All rights of reproduction in any form

de Quimica y Fisica, UniversiPostal No. 9, Rio Cuarto, Corbe addressed.

ent substrates are easily and nondestructively detected by means of differential spectroscopy. In our laboratory, a convenient cell for attenuated total internal reflection Fourier transform infrared (ATR-FT-IR)3 spectroscopic measurements (1,2) of microsamples on AgX fibers has been developed (3). It is useful over a broad wavelength range (2-20 pm) and is suitable for the study of films as well as for aqueous and other solutions. We now describe a potentially useful technique for following conformational changes in proteins through monitoring of the hydrogen-deuterium (H/D) exchange process by ATR-FT-IR spectroscopy using AgX fibers. The high sensitivity of the FT-IR, coupled with the use of optical fibers as an internal reflectance element in ATR-FT-IR spectroscopy, allowed us to develop the present microsampling method for conformational studies on proteins. We have found that proteins (and by extension, rare and valuable natural products) can be studied with this technique on a microgram scale. Microtechniques applying to infrared spectroscopy are quite new and usually require expensive instrumentation (4). The cell used in the present work is based on an infrared transmitting optical AgX fiber (5,6). This cell has the advantages of convenience, relatively low cost, and a geometry that allows easy and quantitative sample loading of small amounts of material. ATR-FTIR measurements of aqueous urea solutions flowing over 3.3-cm AgCl fibers have been reported (7), extending the utility of our earlier report about static aqueous glycine and acetone solutions on lo-cm AgCl fibers (8). Moreover, microplanimetric studies on the optical AgX fibers have suggested that the surface has undula-

3 Abbreviations used: ATR-FT-IR, attenuated total internal reflection Fourier transform infrared; T, trypsin; SBTI, soybean trypsin inhibitor; BAEE, Nol-benzoyl-L-arginine ethyl ester; HPr, protioprotein; DPr, deuterioprotein. 43

Inc. reserved.

44

CHIACCHIERA

tions that may correspond to natural wells (9) (cf. enzyme-linked immunosorbent assay techniques). Under experimental conditions (one or two monolayers of protein), the wells might behave as “cells” in which the intimate contact between molecules is minimized. Hence, enzymes such as trypsin (10) that are prone to autodegradation are stable. Hydrogen exchange is extremely sensitive to the changes in protein conformation because the exchange is effectively irreversible and cumulative. Thus, ligand binding effects usually undetected by common spectrometric techniques often produce substantial perturbation of hydrogen exchange (11). The assays of isotope specific activity as a function of time for total hydrogen exchange are usually adaptations of three basic methods: infrared spectroscopy of the amide II band (H/D exchange), isolation and analysis of the solvent after freeze drying (H/D or H/T exchange), or radioassay after rapid dialysis or gel filtration (H/T exchange). Total hydrogen exchange of peptide NH, free from the interference of side-chain exchange, can be followed by the decay in the intensity of the amide II band in the IR spectrum in deuterated solvent. Conventional IR techniques require a high protein concentration (>>2%) and although less precise than measurements of radioactivity, they are less time consuming and less expensive and do not pose hazards due to radioactivity. IR measurements of hydrated protein films in equilibrium with deuterium vapor have been made in order to overcome the problem of the water interference in the absorbance estimation of the N-H and N-D bands. Transmission measurements carried out using protein between calcium fluoride windows (12), as well as ATR studies of samples on germanium plates (l3), allowed a more reliable quantitation of the H/D exchange process. Deutchsmann and Ullrich (13) found no discontinuity in the exchange behavior of proteins upon going from a solution to an amorphous solid phase when the humidity of the film is maintained, and the diffusion rate of water molecules in the film is not the limiting process in the exchange reaction. The use of the AgX fiber as the internal reflection element in ATR-FT-IR spectroscopy has the advantages of having a wider range of IR transparency and requiring a smaller amount of materials. Also, the diffusion of water into and out of the film would be fast and not rate limiting. The utility of differential spectroscopy should be emphasized. The difference spectra show in an elegant and easy way the frequencies affected in the complexation or exchange processes under consideration. Moreover, the use of this technique yields results that enable one to estimate the number of water molecules per protein molecule in the hydrated film.

AND KOSOWER

MATERIALS

AND METHODS

Bovine pancreatic trypsin type III (salt-free) (T), soybean trypsin inhibitor type I-S (SBTI), and Na-benzoyl-L-arginine ethyl ester (BAEE) were purchased from Sigma and used without further purification. Deuterium oxide (99%) (Aldrich) and twice-deionized distilled water were used. The proteins were dissolved in pure water without any pH adjustment. The concentrations of the solutions were ca. 0.2%. All infrared spectra are measured with a Nicolet 5DX FTIR spectrophotometer equipped with a standard pyroelectric triglycine phosphate detector. The cell contains a silver chloride/bromide fiber (AgX), 0.9 mm in diameter, inserted into solvent-resistant gaskets at the end. The interior and cover of the cell are lined with Teflon. Inlets allow the passage of nitrogen. The cell fits a base that controls its position precisely. The form of the cell has been previously illustrated. An initial zinc selenide lens focuses the light to impinge on the fiber; a second lens collimates the output light for the detector. The AgX fibers were supplied by Professor Aharon Katzir and his group in the School of Physics, Tel-Aviv University. Two kinds of difference spectra are collected: DSl (changes in protein on combination with ligand), IR of T.SBTI complex - C (IR of T + IR of SBTI), and DS2 (changes in materials on deuteration), IR of protioprotein (HPr) - IR of deuterioprotein (DPr). The cleaning of the cell between experiments was done by rinsing the cell sequentially with water, dilute hydrochloric acid solution ( low3 M), and water. The cell was dried by placing it in a vacuum desiccator for at least 30 min. The cell was flushed with dry N, in order to reestablish the conditions used for measurement. This procedure allowed multiple determinations with the same fiber. The quality of the fiber is monitored both by the energy output reaching the detector andby an examination of its background absorption. Depending on the care with which the fibers were handled (for example, minimizing exposure to fluorescent and/or uv light), a fiber could be used over a period of months. The N, was dried by passing through columns of Drierite (8 mesh) and activated 4A molecular sieves. RESULTS AND DISCUSSION

Microsampling

Method

An appropriate background spectrum is taken. A small volume (normally 5 ~1) of a protein solution (typically 0.2%) is placed on the fiber using a polypropylenetipped micropipette. Care is taken to spread the solution evenly without scratching the fiber. The solvent is then evaporated by flushing with a slow stream of nitro-

INFRARED

TRYPSIN

ON AgX FIBER

SPECTROSCOPY

ON

;

t

z

0 s: T

00191-

m

L% @z 00143::

i ‘01 z

f CJ

00094-

5

g

s w

0 0045 0 0003 3500

3022

2544

2056

1794

1555

WAVENUMBER

FIG.

1.

Infrared

spectrum

of 10 rg trypsin

1316

1077

83%

600

halide

fiber.

(cm-l)

on a silver

gen until the stretching frequency of the water band at 3420 cm-’ disappears and the intensity of the amide I band (C=O stretching of the peptide group) remained constant. Any differences in the water vapor and CO, background content during the acquisition time were corrected by subtracting their spectra from that of the sample. A typical trypsin spectrum is shown in Fig. 1. The characteristic protein absorptions are amide I at 1639 cm-‘, amide II at 1524 cm-‘, and amide III at 1236 cm-‘. Spectra of wet films were obtained after the proteincoated fiber was exposed to a water-saturated stream of nitrogen. Experiments carried out using different amounts of protein enabled one to estimate the extension of the film hydration under the experimental conditions. This was accomplished by subtracting a liquid water spectrum from that corresponding to the hydrated film. The “actual” cell volume that can be estimated from the value of the effective penetration of the evanescent electric field was determined (5) to be 0.283 ~1. The factor of reference subtraction provided by the Nicolet DXFTIR routine can be used to evaluate the amount of water responsible for the IR absorption in the hydrated film. In the case of trypsin, a value of 0.41 g of H,O per gram of trypsin is found, in good agreement with values reported for other globular proteins (14). The HID

Exchange

HALIDE

45

FIBERS

bond). The amide I band underwent slight changes upon

IO/l9

00288 0 0240

SILVER

deuteration. This can be easily appreciated in the difference spectra (DS2 type; see Materials and Methods) for the reversible exchange as shown in Fig. 2. Thus, on going from HPr to DPr the following changes are observed: the N-D stretching band appears and the N-H counterpart disappears; the peptide bond C=O stretching absorption (amide I) is shifted from 1682 to 1631 cm-l due to deuteron bonding, while the amide II band (mixture of the peptide C-N stretch and N-D in plane bend) at 1529 cm-’ disappears; and the amide II’band of the DPr develops at 1445 cm-‘. Minor changes are observed in other regions of the spectra. A mirror image difference spectrum is obtained for the reverse process, that is, the replacement of exchangeable deuterium with protium, carried out by flowing H,O-saturated nitrogen gas over the deuterated film. The apparent exchange rates were obtained by plotting the absorbance ratio of the amide II to amide I bands as a function of time. The number of remaining peptide hydrogens (Nu) at each time was calculated in the following way

Nu = (An~A~),&W4),, where A, and AI1 are the intensities of the amide I and II bands, respectively. The zero time value is taken from the undeuterated protein spectrum. In order to increase the SIN ratio, each spectrum illustrated is the average of five collections, each of which involved 300 scans. The exchange process is followed for 30 min, since much of the facile exchange occurred during this period. In the early experiments, the changes were followed for pe-

DIFFERENCE

SPECTRA FOR H/D EXCHANGE N-D stretchmng

OF TRYPSlN

Experiment

A D,O- or H,O-saturated N, stream was passed for different lengths of time through the cell containing the protein-coated fiber. The rate of gas flow was maintained at 50 ml/min to ensure the reproducibility of the measurements. After the film was dried with N,, the cell was placed on the mount and spectra were collected. The H/D exchange was followed by a measurement of the absorbance decrease in the intensity of the amide II band (mainly the N-H bending vibration of the peptide

-0

00921 3500

I 3022

I 2544

I 2066

I 1794

’ ‘,r’ 1555

1316

I 1077

I 83%

/ 600

WAVENUMBERtcm-1)

FIG. 2. Infrared difference spectra (type DS2; see Materials and Methods) for deuteriotrypsin-protiotrypsin and protiotrypsin-deuteriotrypsin. The important difference bands are labeled. The amide II and amide II’ bands involve some C-N stretching in addition to N-H or N-D bending as indicated in the text.

46

CHIACCHIERA PROTELN

% SBTI Estimated

1

H/D

EXCHANGE

KINETICS

Complex for the T-SBTI

mixture

-“~6040000 TIME(s)

FIG. 3. Plots of the log of the fraction of remaining hydrogen in trypsin (T), soybean trypsin inhibitor (SBTI), and trypsin-soybean trypsin inhibitor complex (TSBTI) as a function of exchange time after exposure to a D,O-saturated nitrogen stream. A plot is also given for the same quantity estimated for a sum of T and SBTI.

riods up to several hours without yielding additional information. The baseline of the amide II band was taken as the straight line connecting the sloping background. The definition this baseline constitutes one of the main error sources in the calculation. The denatured protein spectrum has been extensively used to establish the baseline; the advantages of the straight linear interpolation have also been discussed (15). DIH Exchange with the Trypsin-Soybean Complex

AND

KOSOWER

which exchange within 3 min, (ii) the moderately fastexchanging protons, exchanging in less than an hour, (iii) the slowly exchanging protons, and (iv) the nonexchanging protons, those that remain after 24 h (18). The sampling times used in this work allowed the study of classes 1 and 2, essentially only the fast- and moderately fast-exchanging protons, those that are relatively accessible to solvent. As can be seen in Fig. 3, first-order kinetics were obtained in each case and no change in the rate constant occurred during the course of the experiment. The expected exchange curve for the trypsin-soybean trypsin inhibitor mixture, the solid line in Fig. 3, was estimated by adding the contributions of the two polypeptides, on the basis of their individual rates of exchange, the molar fraction of each of them in the mixture, and their respective number of hydrogens. Woodward et al. (19) found that one-third of the exchangeable hydrogens in T are unexchanged in 24 h within the T.SBTI complex. This number exceeds the sum of the maximum number of amino acids in contact with the interface plus the number of new hydrogen bonds, indicating that the region of reduced solvent accessibility extends beyond the contact region. In the present results, the proportion after 30 min seemsto be larger, due to the fact that of the exchangeable protons, those more affected by the complexation are probably those that are the faster exchanging, that is, those that we had been following in the time scale of the experiment.

Trypsin

SBTI binds to the active site of T in a 1:l ratio to form a highly stable complex (T.SBTI) with loss of the catalytic activity. Even though no conformational changes have been detected by the standard techniques (16), hydrogen exchange experiments have shown a noticeable change in the mobility of both peptides upon complexation (17). After the spectrum of trypsin was recorded, the inhibitor solution was added, the film dried, and the new spectrum recorded. By spectral subtraction, the spectra of trypsin upon complexation was obtained and compared with that of the free protein through a difference spectrum, type DSl. No changes were observed. The D/H exchange experiment was performed as described above. Plots of proton absorbance versus time are shown in Fig. 3 for T, SBTI, and T.SBTI complex along with a plot represented the changes expected for T + SBTI. From hydrogen exchange experiments, the hydrogens in native proteins can be placed into four classes, each of which exchanges at a rate that represents the average of the rate distribution of the individual protons in the group. The classes are the (i) fast-exchanging protons,

Assay of Trypsin Biochemical Activity The evaluation of trypsin catalytic activity was made using an adaptation of the classical method (20). The changes produced during the trypsin-catalyzed hydrolysis of BAEE (145 mol BAEE to 1 mol T) were followed by FT-IR. A spectrum of the BAEE was recorded immediately after addition of a sample to a trypsin film and drying in a stream of nitrogen. Water (5 ~1) was then spread on the film and water-saturated nitrogen was passed over the film for 1 h to maintain constant humidity. Another spectrum was recorded after drying. Finally, a difference spectrum (first spectrum minus last spectrum) was obtained. The difference spectrum (Fig. 4) clearly shows the affected frequencies due to the substrate hydrolysis. One may observe the loss of the C=O, C(=O)-0, and -O-&H, stretching bands belonging to the ester group, at 1737, 1215, and 1022 cm-‘, respectively. Simultaneously, the development of the symmetric and asymmetric C=O stretching bands of the carboxylate group at 1570 and 1419 cm-’ can be observed. The C,H,OH

INFRARED DIFFERENCE

SPECTRUM~~~-O~~I~.TRYPS~N

*BAEE

SPECTROSCOPY

ON AgX FIBER

ON SILVER

HALIDE

47

FIBERS

ACKNOWLEDGMENTS The authors are grateful to the Israel-United States Binational Research Foundation and the Basic Research Foundation of the Israel Academy of Sciences for support. Partial fellowship support was provided by Consejo National de Investigaciones Cientificas y Tecnicas (CONICET), Argentina. Special thanks are due to Professor Aharon Katzir and his group in the School of Physics, Tel-Aviv University, for much help and cooperation as well as for the supply of the silver halide fibers.

RCOOEt---RCOO-+EtOH

BAEE

REFERENCES

3954

3436

2922

2406

1945

1687

1429

1171

912

654

WAVENUMBERccm-'1

FIG. 4. Methods) nine ethyl fiber. The

Infrared difference spectra (type DSl; see Materials and showing the changes in the substrate, Nol-benzoyl-L-argiester (BAEE), induced by the enzyme trypsin on the AgX infrared spectrum of BAEE is shown as the lower curve.

1. Harrick, N. J. (1967) Internal Reflection Spectroscopy, Wiley, New York. 2. Mirabella, F. M., and Harrick, N. J. (1985) Internal Reflection Spectroscopy: Review and Supplement, Dekker, New York. 3. Simhony S., Katzir, A., and Kosower, E. M. (1988) Anal. Chem. 60, 1908-1910. 4. Schrader, B. (1990) Fresenius J. Anal. Chem. 337,824-829. 5. Simhony, S., Katzir, A., and Kosower, E. M. (1988) J. Appl. Phys. 64(7),3732-3734. 6.

Margalit, Surf.

absorption bands having been lost

were not observed, during the drying

the volatile

alcohol

procedure. A fresh trypsin film as well as films that have been kept wet for periods of 1,5, and 24 h behaved in the same way. The sensitivity of this test allowed its use to investigate the possibility of trypsin absorption on the fiber. No BAEE hydrolysis was detected when the experiment was carried out on a fiber used for trypsin after cleaning. To perform a further check of the stability of the enzyme on the fiber, trypsin was rinsed off from a freshly prepared film, lyophilized, and analyzed by sodium dodecyl sulfate-gel electrophoresis. Samples from the fresh film and also from films kept wet for 1,5, and 24 h showed no degradation. We expect that the technique of ATR-FT-IR fiberoptic spectroscopy in conjunction with hydrogen-deuterium exchange will be useful for the study of protein binding ability and conformational changes occurring upon complexation. The accessibility of the fiber and the possibility of working on a microgram scale offer a powerful new way of studying materials that can be obtained only in limited quantities.

E., Dodiuk, H., Kosower, E. M., and Katzir, A. (1990)

Interface

Anal.

15, 473-478.

I. Swairjo, M., Rothschild, K. J., Nappi, B., Lane, A., and Gold, H. (1991) Proc. SPZE 1437,60-65. 8. Simhony, S., Kosower, E. M., and Katzir, A. (1987) Biochem. Biophys.

Res. Commun.

142,1059-1063.

9. Schnitzer, I. (1990) PhD thesis, Physics Department, Tel-Aviv University. 10. Epstein, C. J., and Anfinsen, C. B. (1962) J. Biol. Chem. 237, 2175.

11. Woodward, Biochem.

C. K., Simon, I., and Tuchsen, E. (1982) Mol.

Cell.

48,135-160.

12. Pilet, J., Szabb, A. G., and Maurinot,

J. C. (1980) Biophys. Chem.

12,279-284.

13. Deutchsmann, G., and Ullrich, V. (1979) Anal. Biochem. 94, 6-14. 14. Bryan, W. P. (1980) J. Theor. Biol. 87,639. 15. Stryker, M. H., and Parker, F. S. (1970) Arch. &o&em. Biophys. 141,313-321.

16. Laskowski, M., and Sealock, R. W. (1971) in Enzymes (Boyer, P. D., Ed.), 3rd ed., Vol. 3, Academic Press, San Diego. 17. Ellis, L. M., Bloomfield, V. A., and Woodward, C. K. (1975) Biochemistry

14,3413-3419.

18. Englander, S. W., and Staley, R. (1969) J. Mol. Biol. 45, 277. 19. Woodward, C. K. (1977) J. Mol. Biol. 111,509. 20. Schwert, G. W., and Takenaka, Y. (1955) Biochim. Biophys. Acta 16,570.

Deuterium exchange on micrograms of proteins by attenuated total reflection Fourier transform infrared spectroscopy on silver halide fiber.

We illustrate the use of polycrystalline silver halide fibers (2-20 microns transparency range) for attenuated total internal reflection Fourier trans...
578KB Sizes 0 Downloads 0 Views