Anal. Chem. 2007, 79, 416-425
Articles
Method for Stabilizing Protein-Ligand Complexes in Nanoelectrospray Ionization Mass Spectrometry Jiangxiao Sun, Elena N. Kitova, and John S. Klassen*
Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2
The interaction between the bovine pancrease trypsin (Tryp) and its competitive inhibitor benzamidine (1), in solution and the gas phase, is investigated using nanoflow electrospray ionization (nanoES) and Fourier transform ion cyclotron resonance mass spectrometry. In a recent study (Clark, S.M.; Konermann L. Anal. Chem. 2004, 76, 7077-7083), it was reported that the (Tryp + 1) complex could not be detected by ES-MS. Here, it is shown that, with gentle sampling conditions, it is possible to detect gaseous protonated ions of the (Tryp + 1) complex with nanoES-MS. However, the relative abundance of the detected (Tryp + 1)n+ ions is lower than expected, based on solution composition, which suggests that dissociation of (Tryp + 1)n+ ions occurs during MS sampling. The dissociation pathways and corresponding Arrhenius parameters for the protonated (Tryp + 1)n+ ions, at n ) 7-9, are determined from time-resolved thermal dissociation experiments, implemented with the blackbody infrared radiative dissociation technique. The gaseous (Tryp + 1)n+ ions are found to have short lifetimes, e.g., 100 °C. The use of solution additives, including polyols, carbohydrates, amino acids, and small organic molecules, to stabilize the (Tryp + 1)n+ ions during nanoES-MS analysis is investigated. Notably, the addition of imidazole to the nanoES solution is shown to preserve the (Tryp + 1)n+ ions. The Kassoc value, (1.9 ( 0.2) × 104 M-1, determined for the (Tryp + 1) complex by the direct ES-MS method is in agreement with values determined by other analytical methods. The stabilizing effect of imidazole in nanoESMS is further demonstrated for the interaction between carbonic anhydrase II and 5-(dimethylamino)naphthalene-1-sulfonamide. The stabilizing effect of imidazole may be due to enhanced evaporative cooling achieved by the dissociation of molecules of imidazole, bound nonspecifically, from the protein-ligand complex in the ion source. The direct electrospray ionization mass spectrometry (ES-MS) technique is a promising method for establishing the binding * To whom correspondence should be addressed. E-mail: john.klassen@ ualberta.ca.
416 Analytical Chemistry, Vol. 79, No. 2, January 15, 2007
stoichiometry and affinity (Kassoc) of protein-ligand complexes,1-3 as well as other noncovalent biological complexes.4,5 As described in more detail below, the direct ES-MS approach to quantifying Kassoc for protein-ligand complexes requires that the intensity ratio, IPL/IP, for the ligand-bound (PL) and free protein (P) ions, as measured by ES-MS, be equivalent to the ratio of the equilibrium concentrations of PL and P in solution, [PL]eq/[P]eq. Absolute values of Kassoc for a variety of protein-ligand complexes, with affinities ranging from 102 to 106 M-1, have been determined by the direct ES-MS approach and shown to be in agreement with values determined by other analytical methods.1 The key advantages of the direct ES-MS technique are speed of analysis, which can usually be completed within a few seconds, and specificity, which allows for the direct determination of binding stoichiometry and the ability to study binding between multiple proteins and ligands simultaneously. When combined with nanoflow ES (nanoES), the MS-based assay also affords high sensitivity, normally consuming picomoles or less of analyte per analysis. NanoES also readily allows for the transfer of noncovalent complexes from buffered aqueous solution to the gas phase and is amenable to in vitro binding studies under near-physiological conditions. Although the direct ES-MS technique has been successfully applied for the determination of Kassoc values for a variety of protein-ligand complexes, there have been reports of complexes that could not be detected by ES-MS, or, if detected, the distribution of bound and unbound protein in the gas-phase did not match the distribution in solution.6,7 One notable example is the interaction between the serine protease trypsin (Tryp) and its competitive inhibitor benzamidine (1), for which Kassoc values of 2.0 × 104 and 4.5 × 104 M-1 have been reported.6,8 Konermann and co-workers recently investigated the Tryp-1 interaction by ES-MS and reported that ions of the specific (Tryp + 1) complex (1) Daniel, J. M.; Friess, S. D.; Rajagopalan, S.; Wendt, S.; Zenobi, R. Int. J. Mass. Spectrom. 2002, 216, 1-27. (2) Heck, A. J. R.; van den Heuvel, R. H. H. Mass Spectrom. Rev. 2004, 23, 368-389. (3) Loo, J. A.; Berhane, B.; Kaddis, C. S.; Wooding, K. M.; Xie, Y.; Kaufman, S. L.; Chernushevich, I. V. J. Am. Soc. Mass Spectrom. 2005, 16, 998-1008. (4) Sannes-Lowery, K. A.; Griffey, R. H.; Hofstadler, S. A. Anal. Biochem. 2000, 280, 264-271. (5) Hagan, N.; Fabris, D. Biochemistry 2003, 42, 10736-10745. (6) Clark, S. M.; Konermann, L. Anal. Chem. 2004, 76, 7077-7083. (7) Robinson, C. V.; Chung, E. W.; Kragelund, B. B.; Knudsen, J.; Aplin, R. T.; Poulsen, F. M.; Dobson, C. M. J. Am. Chem. Soc. 1996, 118, 8646-8653. (8) Talhout, R.; Engberts, J. B. F. N. Eur. J. Biochem. 2001, 268, 1554-1560. 10.1021/ac061109d CCC: $37.00
© 2007 American Chemical Society Published on Web 12/06/2006
could not be detected, in either positive or negative ion mode.6 It was proposed that the absence of complex in the mass spectrum reflected the instability of the gaseous complex owing to the loss of the stabilizing hydrophobic interactions. In the present work, we have applied nanoES and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICRMS) to investigate the (Tryp + 1) complex in solution and the gas phase. Specifically, we have demonstrated that, with gentle sampling conditions, it is possible to detect gaseous protonated ions of the (Tryp + 1) complex using nanoES-MS, albeit with a relative abundance lower than expected in solution. Time-resolved thermal dissociation experiments, implemented with the blackbody infrared radiative dissociation (BIRD) technique,9,10 were used to determine the Arrhenius activation parameters for dissociation of the protonated (Tryp + 1)n+ ions at n ) 7-9. The ability of a number of solution additives, including carbohydrates, polyols, amino acids, and small organic molecules, to stabilize the (Tryp + 1) complex was investigated. Notably, it was found that imidazole helps to preserve the complex during nanoES-MS analysis. The general stabilizing effect of imidazole in nanoESMS of protein-ligand complexes is further demonstrated for the interaction between bovine carbonic anhydrase II and 5-(dimethylamino)-1-naphthalenesulfonamide (2). EXPERIMENTAL SECTION Proteins and Ligands. Bovine pancreas trypsin (23 332 Da), bovine carbonic anhydrase II (29 089 Da), benzamidine hydrochloride, 5-(dimethylamino)-1-naphthalenesulfonamide (DNSA), histidine, galactose, imidazole, sorbitol, and triethylamine bicarbonate were purchased from Sigma-Aldrich Canada (Oakville, ON, Canada). Formamide and ethanol were products of Fisher Scientific Co. (Nepean, ON, Canada). Acetonitrile was purchased from BDH Inc. (Toronto, ON, Canada). The synthetic monosaccharide abequose was donated by D. Bundle (University of Alberta). All of the reagents were used without further purification. A single-chain variable fragment, scFv (26 539 Da), of the carbohydrate-binding IgG antibody of Se155-4 was produced using recombinant technology.11 The nanoES solutions were prepared from aqueous stock solutions of protein and ligand with known concentrations. Unless otherwise indicated, aqueous ammonium acetate was added to the nanoES solution to yield a final buffer concentration of 10 mM (pH 7). The structures of the ligands, benzamidine (1) and DNSA (2), are shown in Figure 1. Mass Spectrometry. All experiments were performed on an Apex II 9.4-T Fourier transform ion cyclotron resonance mass spectrometer (Bruker, Billerica, MA) equipped with an external nanoES ion source. NanoES was performed using an aluminosilicate capillary (1.0-mm o.d., 0.68-mm i.d.), pulled to 4-7-µm o.d. at one end using a P-2000 micropipette puller (Sutter Instruments, Novato, CA). The electric field required to spray the solution was established by applying a voltage of ∼( 800 V to a platinum wire inserted inside the glass tip. The solution flow rate was typically 20-50 nL/min. The droplets and gaseous ions produced by (9) Price, W. D.; Schnier, P. D.; Jockusch, R. A.; Strittmatter, E. F.; Williams, E. R. J. Am. Chem. Soc. 1996, 118, 10640-10644. (10) Dunbar, R. C.; McMahon, T. B. Science 1998, 279, 194-197. (11) Zdanov, A.; Li, Y.; Bundle, D. R.; Deng, S.; MacKenzie, C. R.; Narang, S. A.; Young, N. M.; Cygler, M. Proc. Natl. Acad. Sci. U. S. A. 1994, 91, 64236427.
Figure 1. Structures of benzamidine (1) and 5-(dimethylamino)-1naphthalenesulfonamide, DNSA (2).
nanoES were introduced into the mass spectrometer through a stainless steel capillary (i.d. 0.43 mm) maintained at an external temperature of 66 °C. The ion/gas jet sampled by the capillary ((48-52 V) was transmitted through a skimmer ((0-2 V) and stored electrodynamically in an rf hexapole. Unless specified otherwise, a hexapole accumulation time of e1 s was used for all experiments. Ions were ejected from the hexapole and accelerated to ∼ ( 2700 V into a 9.4-T superconducting magnet, decelerated, and introduced into the ion cell. The trapping plates of the cell were maintained at a constant potential of ( 1.4-1.8 V throughout the experiment. The typical base pressure for the instrument was ∼5 × 10-10 mbar. The temperature of the ion cell for the BIRD experiments was controlled with two external flexible heating blankets placed around the vacuum tube in the vicinity of the cell.12 Isolation of the reactant ions was achieved using a combination of single rf and broadband rf sweep excitation. The isolated ions were stored inside the heated cell for varying reaction times before excitation and detection. Data acquisition was controlled by an SGI R5000 computer running the Bruker Daltonics XMASS software, version 5.0. Mass spectra were obtained using standard experimental sequences with chirp broadband excitation. The time domain signal, consisting of the sum of 50-100 transients containing 128K data points per transient, were subjected to one zero-fill prior to Fourier transformation. Determination of Kassoc by the Direct ES-MS Technique. The equilibrium expression for the association of a protein (P) and ligand (L) is given by eq 1. The equilibrium concentrations,
Kassoc ) [PL]equil/[P]equil[L]equil
(1)
[PL]equil, [P]equil, and [L]equil, can be calculated from the initial concentration of protein and ligand in solution, [P]o and [L]o, and the relative abundance of the corresponding bound and unbound protein ions measured in the mass spectrum, for example, PLn+ and Pn+ in positive ion mode. It is assumed that the ionization and detection efficiencies for the PLn+ and Pn+ ions are similar such that the ratio (R) of the ion intensity (I) of the bound and unbound protein ions (I(PLn+)/I(Pn+)) determined from the mass spectrum is equivalent to the ratio of the concentrations in solution at equilibrium ([PL]equil/[P]equil). Because the nanoES process typically produces Pn+ and PLn+ ions with a distribution of charge (12) Felitsyn, N.; Kitova, E. N.; Klassen, J. S. Anal. Chem. 2001, 73, 46474661.
Analytical Chemistry, Vol. 79, No. 2, January 15, 2007
417
states, the R value is obtained from the sum of the intensities for the complex ions divided by the sum of the intensities for the protein ions over all the observed charge states. In FT-ICR-MS, the ion signal is proportional to the abundance and charge state (n) of the ions. Therefore, the measured ion intensities must be normalized for charge state. In this case, R should be calculated from eq 2.13
R)
[PL]equil [P]equil
∑(I
(PL)n+/n)
n
)
∑
(2) (I(PL)n+/n)
n
The equilibrium concentration, [PL]equil, can be determined from the value of R and [P]o using the following expression:
[PL]equil )
R[P]o 1+R
(3)
The equilibrium concentration [L]equil can be found from eq 4, and Kassoc can then be determined with eq 5:
[L]equil ) [L]o - [PL]equil Kassoc )
[PL]equil [P]equil([L]o - [PL]equil)
Kassoc )
R R[P]o [L]o 1+R
(4) (5a)
(5b)
RESULTS AND DISCUSSION Detection of the (Tryp + 1) Complex by NanoES-MS. Shown in Figure 2a is a nanoES mass spectrum obtained in positive ion mode for an aqueous solution of trypsin (12 µM) and 1 (53 µM) with ammonium acetate (10 mM, pH 7). The dominant ions revealed by the mass spectrum correspond to protonated trypsin, (Tryp + nH)n+ ≡ Trypn+, where n ) 7-9. Also present are ions corresponding to the protonated 1:1 complex, (Tryp + 1)n+ at n ) 8 and 9. The discrepancy between the charge-state distribution observed for the Trypn+ and (Tryp + 1)n+ ions is attributed to gas-phase dissociation of the (Tryp + 1)n+ ions. Specifically, the Tryp7+ ion is believed to form predominantly in the source via the loss of 1+ from (Tryp + 1)8+, vide infra. To rule out the possibility that the (Tryp + 1)n+ ions originate from nonspecific binding during the nanoES process, mass spectra were also acquired for a solution of trypsin, 1 and a second protein that does not interact with 1. It was recently shown in our laboratory that the occurrence of nonspecific ligand binding to proteins during nanoES is independent of the structure of the protein14 and that the contribution of nonspecific ligand binding can be quantified by adding a reference protein, which does not interact in solution with the ligand of interest, to the nanoES solution.15 Shown in Figure 2b is a mass spectrum acquired for a (13) Wang, W.; Kitova, E. N.; Klassen, J. S. Anal. Chem. 2003, 75, 4945-4955. (14) Wang, W.; Kitova, E. N.; Klassen, J. S. Anal. Chem. 2005, 77, 3060-3071. (15) Sun, J.; Kitova, E. N.; Wang, W.; Klassen, J. S. Anal. Chem. 2006, 78, 30103018.
418 Analytical Chemistry, Vol. 79, No. 2, January 15, 2007
Figure 2. NanoES mass spectra obtained from aqueous ammonium acetate (10 mM) solutions of (a) trypsin (12 µM) and 1 (53 µM), (b) trypsin (12 µM), 1 (53 µM), and scFv (10 µM), in positive ion mode, and (c) Tryp (11 µM) and 1 (106 µM) in negative ion mode.
solution of trypsin (12 µM), 1 (53 µM), scFv (10 µM), and ammonium acetate (10 mM, pH 7). Peaks corresponding to protonated scFvn+ ions, at n ) 9-11, are observed in the spectrum. Importantly, no (scFv + 1)n+ ions are detected. The absence of (scFv + 1)n+ ions in the spectrum strongly suggests that the (Tryp + 1)n+ ions originate from a specific interaction in solution. Shown in Figure 2c is an illustrative mass spectrum obtained in negative ion mode for a solution of trypsin (11 µM) and 1 (106 µM) with ammonium acetate (10 mM, pH 7). According to the mass spectrum, the deprotonated Trypn- ions, at n ) 8 and 9, are produced by nanoES with extensive adducts, corresponding to acetate (Ac-) and chloride ions (Cl-). Due to the presence of the adducts, ions corresponding to the specific complex, (Tryp + 1)n-, could not be positively identified. The influence of solution composition and instrumental parameters on the magnitude of the Kassoc value for the (Tryp + 1) complex, as derived from the mass spectrum, was investigated. Notably, it was found that the magnitude of Kassoc was sensitive to the concentration of ammonium acetate used. Shown in Figure 3 are values of Kassoc, measured at constant concentrations of trypsin (12 µM) and 1 (53 µM), plotted versus the concentration of ammonium acetate, which was varied from 1 to 30 mM. It can be seen that the value of Kassoc increases with increasing concentration of ammonium acetate, reaching a maximal value at ∼10 mM. The dependence of Kassoc on buffer concentration may be due to greater electrostatic shielding between trypsin, which
Figure 3. Influence of ammonium acetate concentration on the Kassoc value determined by nanoES-FT-ICR-MS for the (Tryp + 1) complex in an aqueous solution of trypsin (12 µM) and 1 (53 µM).
is protonated at pH 7 (pI 10.0),16 and 1+ with increasing ionic strength of the solution. The relative abundance of the protonated (Tryp + 1)n+ ions and, consequently, the magnitude of the Kassoc value were also found to be sensitive to the time ions were accumulated in the rf hexapole of the ion source. As shown previously, collisional heating of the gaseous ions can occur during accumulation in the hexapole, leading ultimately to dissociation.13,17 Kassoc values, determined from mass spectra acquired at hexapole accumulation times ranging from 0.6 to 5 s, are shown in Figure 4a. It can be seen that the magnitude of Kassoc decreases significantly with increasing accumulation time, and at accumulation times of g6 s, (Tryp + 1)n+ ions were no longer detected. The effective temperature (Teff) of the (Tryp + 1)n+ ions in the hexapole was estimated from a comparison of the first-order rate constant for dissociation of the complex in the hexapole, established from the change in natural log of the normalized abundance of the (Tryp + 1)n+ ions with accumulation time (Figure 4b), and the Arrhenius plots determined for the dissociation of the (Tryp + 1)n+ ions under thermal conditions, vide infra. Using this procedure, described in more detail elsewhere,13 Teff was estimated to be ∼63 °C. This value is substantially lower than the value of ∼140 °C, which our laboratory previously determined for a 27-kDa proteintrisaccharide complex at charge states +9 to +11.13 The difference in the magnitude of Teff found in these two studies may be due to differences in the extent of solvation of the protein ions entering the hexapole arising from differences in the geometry of, and voltage applied to, the nanoES tips. Based on these measurements, it is concluded that the extent of in-source dissociation of the (Tryp + 1)n+ ions can be minimized by using short hexapole accumulation times. However, there are practical limitations on how short the accumulation event can be. It is our experience that an accumulation time of g0.5 s is necessary to achieve adequate signal intensity in the mass spectrum. Having established optimal experimental conditions (i.e., ammonium acetate concentration of g10 mM, hexapole accumula(16) Gu ¨ nther, A. R.; Santoro, M. M.; Rogana, E. Braz. J. Med. Biol. Res. 1997, 30, 1281-1286. (17) Håkansson, K.; Axelsson, J.; Palmblad, M.; Håkansson, P. J. Am. Soc. Mass Spectrom. 2000, 11, 210-217.
Figure 4. (a) Influence of hexapole accumulation time on the Kassoc value determined by nanoES-FT-ICR-MS for the (Tryp + 1) complex in an aqueous solution of trypsin (12 µM) and 1 (53 µM). (b) Influence of hexapole accumulation time on the natural log of the normalized abundance (Inorm) of the (Tryp+1)n+ ions. Inorm was calculated by considering all of the observed charge states of the Trypn+ and (Tryp+1)n+ ions: Inorm ) [∑(I(Tryp+1)n+)/n)]/[∑(I(Trypn+)/n) + ∑(I(Tryp+1)n+/n)]. Table 1. Association Constant (Kassoc) for Trypsin Binding with Benzamidine (1), As Determined by NanoES-FT-ICR-MS, at 25 °C and Different Concentrations of 1 [trypsin] (µM)
[1] (µM)
Kassoc × 10-4 a-c (M-1)
12 12 12 12
33 53 80 114
0.89 ( 0.12 0.79 ( 0.07 0.75 ( 0.04 0.78 ( 0.06 (0.80 ( 0.06)d
a Ammonium acetate concentration was 10 mM. b Values of K assoc represent average values from five measurements. c Errors correspond d to one standard deviation. The value in parentheses corresponds to the average value of Kassoc determined from individual values measured at the four different concentrations.
tion time
