Anal. Chem. 1995, 67,3775-3781
Dynamics of Acrylodan-Labeled Bovine and Human Serum Albumin Sequestered within Aerosol-OT Reverse Micelles Jeffrey S. Lundgren, Mark P. Heitz, and Frank V. Bright*
Department of Chemistty, Natural Sciences and Mathematics Complex, State University of New York at Buffalo, Buffalo, New York 14260-3000
We investigate the effects of hydration on acrylodanlabeled bovine and human serum albumin @SA-Ac and HSA-Ac) in aerosol-OT(AOT) reverse micelles solubilized in n-heptane. Ti”-resolved fluorescence intensity decay experimentsreveal a dipolar relaxation process surrounding the acrylodan cybotactic region. This process is best described by a two-term rate law wherein the average r e h t i o n increases with increased hydration. However, the actual rate constants describingthe relaxation process either remain unchanged or actually decrease with increased hydration. The results illustrate that the fractional contribution associated with the individual relaxation pathways causes the observed changes in relaxation dynamics. The recovered rotational reorientation dynamics of the acryfodan residue are also affected by the extent of protein hydration. As hydration is increased, the semiangle through which the acrylodan residue precesses increases by 10”for both protein systems. Interestingly, the recovered semiangles for the native proteins equal those recovered at lower hydration when the proteins are sequestered within the AOT reverse micelle. These results demonstrate the importance of hydration on protein behavior in environments where water is limited (e.g., biosensor interfaces and sol-gel-derived biocomposites). At the most elementary level, proteins are identitied by their amino acid sequence or primary This structure can be altered only by breaking and/or forming new covalent bonds within the peptide chain. To a f i s t approximation, protein conformation, function, and stability depend strongly on primary structure, but secondary, tertiary, and quaternary structures also affect protein behavi~r/function.~-~ Plasma proteins, enzymes, and antibodies are systems in which ultimate function/performance are governed by subtle structural changes within the protein micro domain^.^ A portion of the research in our group has focused on the development of analytical biosens0rs.5-~ During the course of this work, we have become acutely aware of the need to understand (1) Stryer, L.Biochemistry; W. H. Freeman and Co.: New York, 1988. (2) Schultze, H. E.: Heremans, J. F. Molecular Biology of Human Proteins; Elsevier Publishing Co.: New York, 1966,Vol. 1, Chapter 1, pp 7-40. (3)DeVoe, H. In Structure and Stability ofBiologica1 Macromolecules: Theory of the Conformation ofBiologica1 Macromolecules in Solution; Timasheff, N. S., Fasman, G. D., Eds.; New York, 1969; pp 2-63. (4)Creighton, T.E. Protein Function: A Practical Approach; Oxford University Press: New York, 1989. 0003-2700/95/0367-3775$9.00/0 0 1995 American Chemical Society
and thus control the behavior of a biorecognition element (e.g., enzymes and antibodies) at or within a sensor interface. For example, it has become evident that the act of immobilizing/ entrapping the biorecognition element to form the biosensor can alter its function, behavior, performance, and/or stability and ultimately affect a biosensor‘s analytical figures of merit. We have also shown that a complex interdependence exists between the biorecognition element and the behavior of the reporter group (i.e,, fluorophore) used to “sense” target analyte-biorecognition element binding. Many studies have shown that “hydration”plays a key role in protein structure and perf0rmance.8-~~ Thus, the degree of protein hydration and/or local solvent composition can affect a protein’s structure and dynamics and, in turn, its performance. For example, Goryunov and co-worker~’~ have illustrated how human serum albumin (HSN is affected by solvation. These authors followed the spin-spin relaxation times of water protons and found that they were a function of the percent age of DzO in the system, demonstrating that the stability of HSA is influenced by the degree of hydration. Bolton and Schererl*have shown that the structure of bovine serum albumin @SA),cast as a thin film, is affected by the relative humidity (RH). In this work, Raman spectroscopy was used to follow changes in the a-helicity of BSA as a function of RH.Together these results demonstrate that protein hydration can control protein conformation and behavior. Thus, it seems clear that any attempt toward exploiting proteins as chemical (5)(a) Bright, F. V.; Litwiler, IC S.; Vargo, T. G.; Gardella, J. A, Jr. Anal. Chim. Acta 1992,262,323-30.(b) BettsT. A;Catena, G. C.; Huang, J.; Litwiler, IC S.; Zhang, J.; Zagrobelny, J.; Bright, F. V. Anal. Chim.Acta 1991,246, 55-63. (c) Bright, F. V.; Betts, T. A; Litwiler, IC A Ana!. Chem. 1990,62, 1065-9. (6) (a) Lundgren, J. S.; Bekos, E. J.; Wang, R; Bright, F. V. Anal. Chem. 1994, 66,2433-40. (b) Wang, R;Narang, U.: Prasad. P. N.; Bright F. V. Anal. Chem. 1993,65,2671-5. (c) Narang, U.;Wang, R; Prasad, P. N.; Bright, F.V. 5 Phys. Chem. 1994, 98,17-22. (d) Jordan, J. D.; Dunbar, R A; Bright, F. V. Anal. Chem. 1995,67,2436-43. (7)Wang, R;Sun, S.; Bekos, E. J.; Bright, F. V. Anal. Chem. 1995,67,14959. (8)Steinmann, B.;Jackle, H.; Luisi, P. L. Biopolymers 1986,25, 1133-56. (9)Gekko, IC; Morikawa, T. J. Biochem. 1981,90,39-50. (10)Otting, G.; Liepinish, E.; Wuthrich, IC, Science 1991,254,974-80. (11) Bemdt, K D.;Beunink. J.; Schrtider. W.; Wiilthrich, K Biochemisty 1993, 32,4564-70. (12)Khuurgin, Y.;Maksareva, E. FEES Le#. 1993,315, 149-52. (13)KtiivMiinen, A I.; Sukhanova, G.; Goryunov, A S. Folia [Biol.] 1984.30, 84-92. (14)Gallay, J.; Vincent. M.; Nicot, C.; Waks, M. Biochemistry 1987,26,573847. (15)Holmberg, K.Ado. Colloid Intetfnce Sci. 1994,51, 137-74. (16)Ruckenstein, E.;Karpe, P. Biotechnol. Le#. 1990, 12,241-6. (17)Han, D.; Rhee, J. S. Biotechnol. Bioeng. 1986,28,1250-5. (18)Bolton, B. A;Scherer, J. RJ.Phys. Chem. 1989,93,7635-40.
Analytical Chemistry, Vol. 67,No. 20,October 15, 1995 3775
recognition elements should be carried out with attention to the effect of hydration on the protein-reporter group couple. Aerosol-OT (AOT) reverse micelles allow one to simultaneously solubilize proteins within a defined domain and control the extent of protein h y d r a t i ~ n . AOT ~ ~ ’ ~reverse ~ micelles consist of a polar core surrounded by ionic head groups that are, in turn, attached to aliphatic tails, suspended in an organic liquid phase.21,22 Thus, by adding protein and the desired amount of water to this system, both of which are sequestered within the reverse micelle, one can effectively control the degree of protein hydration. In this paper, we report on the effects of hydration on BSA and HSA These proteins were selected because they have been used previously as models6a,7,13~’*~21.z~ and they both have a single, free cysteine residue at position 34.23,24This particular site provides a convenient target for site-selective labelingz5with the fluorescent probe &acryloyl-2-(dmethylamino)naphthalene (acrylodan, Ac). Thus, these systems serve as simple biorecognition element-reporter group couples. Further, because all fluorescence results from a single, well-defined site within the protein, we can follow how protein hydration influences the dynamics surrounding the cysteine34 residue. Previous work from our group on BSA-Ac has demonstrated that there are nanosecond and subnanosecond dipolar relaxation processes within the cybotactic region surrounding the acrylodan reporter group7 Complementary timeresolved decays of anisotropy revealed that the rotational motion of the acrylodan reporter group in the native protein is limited to a semiangle of 23 & 1”. In the current work, our goals are to determine the effects of hydration on (1) the average local environment surrounding the acrylodan reporter group, (2) the dynamics of the cybotactic region surrounding the acrylodan residue, and (3) the rotational dynamics associated with the acrylodan probe. Steady-state and time-resolved fluorescence are used to address these issues, and AOT reverse micelles formed in n-heptane serve to control the extent of protein hydration. THEORY The theory of frequency-domain fluorescence for recovering intensity and anisotropy decays and determining the time evolution of the fluorescence emission spectrum (e.g., dipolar relaxation of the cybotactic region) has been described e l s e ~ h e r e . ~The .~~-~ interested reader is referred to the aforementioned references for a more thorough description of the acquisition techniques, data analysis schemes, and basic interpretation. Time-Dependent Spectral Relaxation. Photoexcitation of a fluorophore generally leads to an instantaneous change in the (19) Marzola, P.; Pinzino, C.; Veracini, C. A Langmuir 1991,7, 238-42. (20) Marzola, P.; Gratton, E. J. Phys. Chem. 1991,95, 9488-95. (21) Costantino, L.; Volpe, C. D.; Ortona, 0.;vitagliano, V. J. Chem. Soc., Faraday Trans. 1992,88,61-3. (22) Fletcher, P. D. I. J. Chem. Soc., Faraday Trans. 1987,83, 1493-506. (23) Rosenoer, V. M.; Oratz, M. Albumin Structure, Function, and Uses;Pergamon Press: New York. 1977. (24) Bing, D. H. The Chemistry and Physiologv of Human Plasma Proteins; Pergamon Press: New York, 1979. (25) Gamson, M. D.; Iuliano, D. J.; Saavedra, S. S.;Truskey, G. A; Reichert, W. M. J. Colloid Intelface Sci. 1992,148,415-24. (26) (a) Topics in Fluorescence Spectroscopy; Lakowicz, J. R, Ed.; Plenum Press: New York, 1991; Vols. 1-3. (b) Bright, F. V.; Betts, T. A; Litwiler, K. S. CRC Cnt Reo. Anal. Chem. 1990,21, 389-405. (27) Gratton, E.; Jameson, D. M.; Hall, R D. Annu. Reo. Biophys. Bioeng. 1984, 13,105-24. (28) Lakowicz, J. R Pknciples of Fluorescence Spectroscopy; Plenum Press: New York, 1983; Chapter 6, pp 155-85. (29) Lakowicz, J. R.; Cherek H. Chem. Phys. Left. 1985,122,380-4.
3776 Analytical Chemistry, Vol. 67, No. 20, October 75, 7995
fluorophore’s dipole This disrupts the equilibrium between the fluorophore and its surrounding environment. In order to reestablish equilibrium, the solvent molecules and/or local groups within the cybotactic region “relax” or “reorganize” to minimize the unfavorable interactions produced following the sudden change in the fluorophore dipole moment. If this relaxation process occurs on a time scale similar to the fluorophore excited-state lifetime, it results in a quantitiable, time-dependent shifting of the emission s p e ~ t r u m . ~In~ turn, ~ ~ - this ~ ~ provides information on the dynamics of the local environment surrounding the fluorescent reporter group.7,27-32 Decay of FluorescenceAnisotropy. Time-resolved fluores cence anisotropy measurements on ‘biomolecule-reporter group systems provides information on the global motion of the protein and local motion of the fluorescent reporter group. Our previous work7 on BSA-Ac showed that the decay of anisotropy, r(t), was described by a double exponential decay of the form
where r, is the limiting anisotropy, 41 is the rotational correlation time associated with the local motion of the probe, and & is the rotational correlation time of theglobal motion of the protein. The terms B1 and BZ are the fractional contributions to the total anisotropy decay from the local and global motions, respectively (CPi = 1). If the probe precesses about its axis, and the protein is free to rotationally reorient, one can associate a semiangle (e) with the range of motion that the probe experiences over the course of the excited-state liietime:7,26a
Thus, if the cybotactic region surrounding the reporter group were totally restrictive to the local motion of the fluorescent reporter group, 8 would approach 0”. In contrast, an increase in the semiangle indicates that the probe has more local mobility. EXPERIMENTAL SECTION Reagents. The following chemicals were used: &acryloyl-2(dimethy1amino)naphthalene(acrylodan, Molecular Probes); essentially fatty acid-free BSA, HSA, and 12 OOO MW cutoff cellulose dialysis tubing (Sigma Chemical Co.); sodium bis(2ethylhexyl)sulfosuccinate and 1,4bis(4methyl-5phenyl-2-oxazolyl) benzene (MeZPOPOP) (Aldrich Chemical Co.); n-heptane (Mallinckrodt SpecialtyChemicals Co.); NazHPOc7Hz0, NaN03, and NaClOqHzO (Fisher Chemical); and N~H~PO~.HZO a. T. Baker Inc.). Preparation of BSA-and HSA-Ac. BSA- and HSA-Ac were prepared according to published protocols.7~25 Encapsulation of BSA- and HSA-Ac in Am. A stock solution of 0.2 M AOT was prepared by adding the dry AOT solid to a 250 mL volumetric flask. Two hundred milliliters of n-heptane was then added, the solution was sonicated until all of the AOT had dissolved and diluted to volume with n-heptane. Protein was loaded into the reverse micelles by first placing an aliquot of the (30) Maroncelli, M.; Fleming, G. R. J. Chem. Phys. 1987,86,6221-9. (31) Maroncelli, M.; Castner, E. W.; Bagchi, B.; Fleming, G. R. Faraday Discuss. Chem. SOC.1988,85, 199-210. (32) Maroncelli, M.; MacInnis, J.; Flemming, G. R. Science 1989,3,1674-81.
----
stock AOT solution into a quartz cuvette. In a second step, we added a known volume of distilled, deionized water to the micelle _ -RR 2.78 5.58 .-a solution. In the final step, we added a known volume of the ...... RR a.33 a ...... R 11.1 protein stock solution. The total water added to the system, 13.9 c _ . - R 22.2 defined as the water loading (R = [waterl/[AOTI), is the sum of Buffer a u the water added in step 2 and the water from the protein stock a solution. The final concentration of the acrylodan-labeled serum a albumin was always 0.5 pM. All samples were equilibrated at 20 !! 0 "C for at least 12 h prior to fluorescence measurements. Under i; these particular conditions,there is little probability of two or more proteins being sequestered within a single micelle. Fluorescence Measurements. All steady-state measure ments were performed with a SLM 48000 MHF spectrofluorom23.0 eter using a Xe arc lamp as the excitation source. All measure B 0 0%-AC b USA-AC ments were made at 20 "C, and spectra were background * 22.5 I subtracted. ' Time-resolved intensity and anisotropy decay data were ac'E ? quired in the frequency domain using a SLM 48000 MHF c 22.0 o n multifrequency phase-modulation fluorometer. An argon ion @ I L O w r laser (Coherent,Model Innova 4W10) operating at 351.1 nm was c PI, P # I c . 5 21.5 used as the excitation source. A 340 f 20 nm bandpass filter 0 was placed in the excitation path to eliminate extraneous plasma discharge. Magic angle polarization was used for all excited-state 21 .o intensity decay experiment^.^^ MezPOPOP in ethanol served as the reference lifetime standard; its lifetime was assigned a value of 1.45 mZ8For all experiments, the Pockels cell was operated Figure I.Steady-state emission spectra of HSA-Ac (panel A), and the emission center of gravity for BSA-Ac (0)and HSA-Ac (0)in AOT at a repetition rate of 5 MHz. Typically data were collected for reverse micelles solubilized in n-heptane as a function of R (panel 60-90 s over a frequency range of 5-125 MHz (25 frequencies). 8). All multifrequency phase and modulation data were analyzed according to the global analysis method described e l ~ e w h e r e ? ~ ~ - ~ ~
.-
wt
Y
_ . I _
-E
.-
Q)
- 0 -
I
RESULTS AND DISCUSSION
Steady-StateFluorescence. The siteselective labeling of the serum albumin at the cysteine34 (loop 1, domain I) defines the cybotactic region of the acrylodan report groups7Thus, by using the reverse micelle system and controlling R, we are able to follow the way protein hydration affects the cybotactic region surrounding cysteine-34. Figure lA presents the steady-state emission spectra of HSA-Ac in AOT at six different water loadings and compares them to the native protein in 0.1 M phosphate buffer @H 7.0). Clearly, as R increases, there are several changes in the average local environment surrounding the acrylodan residue. First, the emission spectra shift from 440 nm at R = 2.78 to 465 nm at R = 22.2. Second, the emission spectra broaden and the total fluorescence decreases with increasing R. The same trend is observed for BSA-Ac (data not shown). Third, as the protein is hydrated (Ris increased), the average environment surrounding the Ac group apparently becomes more similar to the environment seen for the native protein. However, even above R = 22.2, the emission spectra are unchanged and do not reach those seen for native BSA- or HSA-Ac. In order to better compare the effects of hydration on the BSAand HSA-Ac emission, Figure 1B presents the emission center of (33)Spencer, R D.;Weber, G. J. Chem Phys. 1970,52,1654-67. (34) (a) Wang, R;Bright, F. V.J. Phys. Chem. 1993,97,4231-8.(b) Wang, R; Bright, F. V.J. Phys. Chem. 1993,97,10872-8. (c) Wang, R;Bright, F. V. Appl. Spectrosc. 1993,47,800-6.(d) Wang, R;Bright, F. V. Appl. Spectrosc. 1993,47,792-9. (35)Beecham, J. M.;Gratton, E. Proc. SPIE-Int. SOC.Opt. Eng. 1988, 909, 70-81. (36)Beecham, J. M.;Graton, E.; Ameloot, M.; Knutson, J. R; Brand, L In Topics in Fluorescence Spectroscopy;Lakowicz,J. R, Ed.; Plenum Press: New York, 1991;Vol. 2, Chapter 5, pp 241-305.
I
gravity (ECOG) as a function of R The ECOG37is simply the weighted average emission. Interestingly, when these serum albumins are sequestered withiin the AOT micelle, there are no statistical differences in the steady-state ECOGs between BSAand HSA-Ac over the R values investigated (Figure 1B). However, the ECOGs of BSA- and HSA-Ac in 0.1 M phosphate buffer (PH 7.0) are 20 240 & 100 and 20 920 f 100 cm-', respectively. Thus, these steady-state experiments reveal that BSA- and HSA-Ac are more similar when sequestered within the AOT micelle than they are in their native forms. This suggests that the native protein structure is not achieved within even the hydrated reverse micelle systems. Tie-Resolved Fluorescence. Previous work from our group7on native, chemically denatured, and silica-adsorbed BSAAc has shown that the acrylodan ground state is homogeneous and the excited state is influenced by dipolar relaxation processes. When BSA- and HSA-Ac are sequestered within AOT micelles, this same pattern persists at all R values. Thus, dipolar relaxation within the environment surroundingthe acrylodan reporter group continues when these proteins are placed in a restricted, dehydrated environment. In order to determine the relaxation rate law and recover the kinetics of this relaxation process? we carried out a series of emission wavelengthdependentmultifrequency phase and modulation (MPM) experiments on BSA- and HSA-Ac in the AOT micelles as a function of R. Figure 2A presents a typical series of wavelengthdependent MPM traces for HW-Ac sequesteredwithin AOT at R = 11.1. The fact that the traces are not superimpossible (37)Lakowicz, J. R;Hogen, D. Biochemisty 1981,20,1366-73.
Analytical Chemistry, Vol. 67, No. 20,October 15, 7995
3777
22.5
n 22.0
c
I
E
0 v
21 .I
c3
0
0
w 10
21 .o
100
20.5
Frequency (MHz)
0
6
4
12
16
20
1 .o
0.8
0.6
n W .w
n 0
z
0.2
-
0.0 400
0.4
,'
425
460
0.2
B
I
475
500
525
0.0
Emission Wavelength (nm) Figure 2. Wavelength-dependent multifrequency phase and modulation traces (panel A), and the recovered time-resolved emission in AOT reverse micelles solubilized spectra for HSA-Ac (R = 11.l) in n-heptane (panel B).
demonstrates that the acrylodan excited-state decay kinetics are heterogeneous. Detailed analysis of the data following our earlier protocol7yield the time-dependent emission spectra of the acrylodan reporter group (Figure 2B). Dipolar relaxation is clearly evident as a red shift (Le., relaxation) in the emission spectra with time. Figure 3A presents the actual time-resolved ECOG for HSAAc in AOT as a function of R. Several aspects of these data merit special mention. First, following optical excitation, there is a redshifting of the spectrum at all R values. Second, at our first time point (-20 ps), the ECOG traces do not all begin at the same ECOG. This suggests that there is at least one faster (
