PROTEINS: Structure, Function, and Genetics Suppl. 2:28–37 (1998) Application of Electrospray Ionization Mass Spectrometry for Studying Human Immunodeficiency Virus Protein Complexes Joseph A. Loo,* Tod P. Holler, Susan K. Foltin, Patrick McConnell, Craig A. Banotai, Nicole M. Horne, W. Tom Mueller, Tracy I. Stevenson, and David P. Mack Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, Ann Arbor, Michigan ABSTRACT Mass spectrometry (MS) with electrospray ionization (ESI) has shown utility for studying noncovalent protein complexes, as it offers advantages in sensitivity, speed, and mass accuracy. The stoichiometry of the binding partners can be easily deduced from the molecular weight measurement. In many examples of protein complexes, the gas phasebased measurement is consistent with the expected solution phase binding characteristics. This quality suggests the utility of ESI-MS for investigating solution phase molecular interactions. Complexes composed of proteins from the human immunodeficiency virus (HIV) have been studied using ESI-MS. Multiply charged protein dimers from HIV integrase catalytic core (F185K) and HIV protease have been observed. Furthermore, the ternary complex between HIV protease dimer and inhibitor pepstatin A was studied as a function of solution pH. Zinc binding to zinc finger-containing nucleocapsid protein (NCp7) and the NCp7-psi RNA 1:1 stoichiometry complex was also studied by ESI-MS. No protein-RNA complex was observed in the absence of zinc, consistent with the role of the zinc finger motifs for RNA binding. Proteins Suppl. 2:28–37, 1998. r 1998 Wiley-Liss, Inc. Key words: electrospray ionization mass spectrometry; noncovalent complexes; protease; integrase; nucleocapsid protein INTRODUCTION The development of new volatilization/ionization methods such as electrospray ionization (ESI) has allowed mass spectrometry (MS) to play an increasingly important role in biomedical research. Sensitive detection and accurate molecular mass determinations for a variety of macromolecules are important advantages of MS applications. ESI has also been used to study the solution phase complexation processes of molecules. Biological processes are typically mediated by the intramolecular and intermolecular interactions of macromolecules. Multiply charged gas phase ions representing intact noncovar 1998 WILEY-LISS, INC. lently bound complexes can be detected with ESIMS, as ESI is a gentle ionization method, yielding little to no molecular fragmentation. The reports of the globin-heme interaction of myoglobin1 and the FKBP-FK506 receptor-ligand complex2 obtained from aqueous solution at near neutral (or physiological) pH were the first to suggest that complexes that associate through specific noncovalent interactions in solution could be detected by ESI-MS. Since these initial reports, a variety of noncovalent binding systems have been studied and reported, including antibody-antigen, protein-cofactor, protein and oligonucleotide coordination with metal ions, enzymesubstrate pairings, protein subunit complexes (quaternary structure), oligonucleotide duplex (and higher order) assemblies, and protein-DNA/RNA complexes.3 We have used ESI-MS to study protein complexes relevant to drug discovery efforts against the human immunodeficiency virus (HIV). Several HIV proteins have been identified as possible targets for drug therapies, such as protease, integrase, and nucleocapsid protein. As part of their function, many of these HIV proteins interact with other molecules. This report will present several examples of protein interactions we have studied by ESI-MS, such as metal ion binding, inhibitor binding, protein quaternary structures, and protein-RNA interactions, as part of our efforts in HIV drug discovery. Because of the accuracy of the mass measurements, stoichiometry of the binding partners is readily determined, even for small molecule binding (e.g., inhibitors and drug molecules) to larger molecular mass targets. MATERIALS AND METHODS Integrase Catalytic Core (F185K) Expression and Purification The synthetic HIV-1 integrase gene was shuttled between pKK223–3 and the T7 expression vector pET21a (Novagen, Madison, WI), using the unique NdeI and HindIII sites, to produce the plasmid pET21a-NY5IN. This plasmid was used to transform Escherichia coli strain BL21(DE3). BL21(DE3) con*Correspondence to: Joseph A. Loo, Parke-Davis Pharmaceutical Research, 2800 Plymouth Road, Ann Arbor, MI 48105. E-mail: Joseph.Loo@wl.com Received 7 April 1998; Accepted 14 July 1998 ESI MASS SPECTROMETRY OF HIV PROTEIN COMPLEXES taining pET21a-NY5IN were grown at 37° C with vigorous stirring and aeration in a 2 L Omniculture fermentor (Virtis Company, Gardiner, NY) containing superbroth (Digene, Silver Spring, MD) and 100 µg/mL ampicillin. When the optical density reached 8 (600 nm), integrase expression was induced by addition of IPTG to a final concentration of 1 mM. Approximately 90 minutes later, the cells were harvested by centrifugation at 5,000 x g for 20 minutes, frozen in liquid nitrogen, and stored at ⫺80° C. Purification, carried out at 4° C, was begun by suspending the cells in 5 mL of lysis buffer (50 mM HEPES, pH 7.5, 5 mM DTT, 1 mM DTPA, and 1 µg/mL leupeptin and pepstatin A) per gram of cell paste. The suspension was passed twice through a French pressure cell at 12,000–14,000 psi and spun at 12,000 x g for 30 minutes. The supernatant was discarded, and the pellet was resuspended in buffer A (25 mM HEPES, pH 7.5, 1 mM DTPA, 1 mM DTT, 1 µg/mL each leupeptin and pepstatin A), homogenized with a Dounce homogenizer, and re-pelleted. This second pellet was resuspended in buffer B (25 mM HEPES, pH 7.5, 10 mM CHAPS, 1 mM DTT, 1 mM DTPA, 1 µg/mL each leupeptin and pepstatin A) containing 1 M NaCl and stirred for 1 hour to extract the integrase. The suspension was clarified by centrifugation at 80,000 x g for 1 hour. The supernatant was slowly diluted ten-fold with buffer B containing 150 mM NaCl, mixed with DEAE Sephacel (Pharmacia, Uppsala, Sweden), stirred on ice for 30 minutes, and filtered to remove the resin. Heparin sepharose (Pharmacia) was added to the filtrate, stirred on ice for 10 minutes, and then packed into a column. Integrase was eluted with a gradient of 250–1,000 mM NaCl in buffer B. Integrase-containing fractions were pooled, dialyzed overnight (25 mM HEPES, pH 7.5; 10 mM CHAPS, 10 mM DTT; 500 mM NaCl and 20% glycerol), aliquoted, and stored at ⫺80° C. Integrase concentration was determined by quantitative amino acid analysis. Solutions for ESI-MS measurements were desalted and concentrated by centrifugal ultrafiltration. Intregrase solutions were washed with two 400 µL volumes of 5% (v/v) acetic acid, followed by two 400 µL volumes of 10 mM ammonium acetate (pH 6.5) and passing the solutions through 5,000 Da cut-off centrifugal filtration cartridges (Ultrafree-MC with low-binding regenerated cellulose membrane; Millipore, Bedford, MA). The protein solutions were concentrated to 0.2–0.5 µg/µL. Protease Expression and Purification Cell strain The E. coli strain expressing the HIV protease NY5 was a gift from Dr. Sergei Gulnik (SAIC Frederick, Frederick, MD). The host cell line was BL21(DE3) and the vector was pET24. 29 Expression The culture was started by inoculating a shake flask containing 200 mL of Luria Broth (Becton Dickinson, Cockeysville, MD) plus 100 µg/mL ampicillin with 1 mL of a 20% glycerol stock. The shake flask was grown at 30° C. After the shake flask became cloudy (8 hours), 1 mL of the shake flask contents was used to inoculate a 10 L fermentor. The media for the fermentation consisted of 2% (w/v) of the following: yeast extract (Difco, Detroit, MI); acidicase peptone (Becton Dickinson); casitone (Difco); gelatone (Difco); 0.2% (w/v) of the following: KH2PO4 anhydrous, K2HPO4 anhydrous, Na2HPO4●7H2O; and 100 µg/mL ampicillin. The culture was maintained at pH 6.8 with lactic acid. Air was introduced into the fermentor at 8 L/min, and the temperature was maintained at 30° C. The culture was grown in the fermentor to an optical density of 10 at 600 nm, the temperature was raised to 37° C, and recombinant expression was induced by addition of isopropyl ␤-D-thiogalactopyranoside (IPTG) to a concentration of 3.5 mM. The fermentor was harvested 3 hours after induction by using hollow-fiber microfiltration and centrifugation. The cell paste was stored at ⫺70° C. Cell disruption Cell lysis was carried out by using a Dyno-Mill KDL (Glen Mills, Clifton, NJ) with a 600-mL chamber containing 500 mL of 0.25–0.5 mm lead-free glass beads. The lysis solution was 50 mM sodium phosphate buffer, pH 7.2, containing 1.0% Triton X-100, 10 mM MgCl2, and 0.045 mL of Benzonase (American International Chemicals, Natick, MA) per liter. The cell paste (350–400 g) was resuspended to 1.2 L with lysis solution and run through the Dyno-Mill twice at 100 mL/min. The jacket temperature was ⫺13° C and the agitator speed was 4,200 rev/min. The suspension was then centrifuged for 90 minutes at 4° C in a DuPont Sorvall (Wilmington, DE) centrifuge using a GS-3 rotor at 9,000 rev/min. Inclusion body isolation The inclusion body isolation was based on the procedures of Estape and Rinas4 and Marston et al.5 The pellet from the centrifugation was thoroughly resuspended using a Polytron PT 3000 homogenizer equipped with a PT DA 3012/ S generator (Brinkmann Instruments, Westbury, NY) at 10,000 rev/min with 750 mL of 50 mM potassium phosphate buffer, pH 8.0, containing 10 mM EDTA, 0.06 mL/L Benzonase, and 2,500 U/mL lysozyme from chicken egg white (Sigma Chemical, St. Louis, MO). The suspension was adjusted to pH 7.5 with NaOH or HCl and stirred for 2 hours at 37° C. Glycine was then added to a concentration of 50 mM and the pH was adjusted to 10 with NaOH. Triton X-100 was added to a concentration of 1%, DTT was added to a concentra- 30 J.A. LOO ET AL. tion of 10 mM, and the suspension was stirred for 18 hours at 4° C. Water (750 mL) was added, and the suspension was centrifuged as described above. The pellet was resuspended as described above with 1.5 L of water; the suspension was stirred for 30 minutes and centrifuged as described above. The water wash step was repeated once more. The suspension was resuspended with 1.5 L of 95% ethanol, stirred for 1 hour, and centrifuged as described above. The pellet was lyophilized in a Dura-Dry µP Freeze Dryer (FTS Systems, Stone Ridge, NY) at 10 mT for 18 hours and stored at ⫺70° C. the enzyme solution was diafiltered versus 2 L of 50 mM sodium acetate buffer, pH 5.5, concentrated to 7–10 mg/mL with an Amicon-Millipore YM 10 membrane at 5° C and stored at ⫺70° C. Protease solutions for ESI-MS analysis were desalted by using 5 kDa cutoff centrifugal ultrafiltration cartridges (Millipore). The 7–10 mg/mL protease samples were washed with several volumes of a solution containing 0.1 mM pepstatin A inhibitor (Sigma), 0.3 mM DTT, and 10 mM ammonium acetate (pH 5). The addition of pepstatin A prevented degradation of the enzyme. S-Sepharose chromatography Activity assay All steps were carried out at room temperature. The dried pellet (6–9 g) was extracted by resuspending in 330 mL of 50 mM sodium borate buffer, pH 8.6, containing 8 M urea, 2 mM EDTA, and 10 mM DTT (solvent A) using the Polytron homogenizer and centrifuged as described above. The extraction of the pellet was repeated twice. The supernatants from the three extractions were combined and Triton X-100 was added to a concentration of 1%. The solution was adjusted to pH 8.6 with NaOH and applied to a S-Sepharose Fast Flow column (9 ⫻ 10 cm; Pharmacia) equilibrated with solvent A. The column was washed with solvent A at 200 mL/min until A280nm of the column effluent was zero. The enzyme was eluted with a linear salt gradient (solvent B: 0.5 M NaCl in solvent A, total gradient volume of 4 L) at 200 mL/min. Fractions containing the purified, denatured enzyme as determined by SDS-PAGE were combined and stored at ⫺70° C. Matrix-assisted laser desorption/ionization (MALDI)-MS (Voyager Elite-DE time-of-flight mass spectrometer, PerSeptive Vestec, Framingham, MA) was used to verify the molecular mass of the protein. The activity of the enzyme was measured by the hydrolysis of the substrate His-Lys-Ala-Arg-Val-Leup-nitro-Phe-Glu-Ala-Nle-Ser-NH2 (BACHEM, King of Prussia, PA)7,8 in 72 mM sodium acetate buffer, pH 4.7, containing 0.9 mM EDTA, 0.14 M NaCl, 0.09% (w/v) PEG 8000, and 0.9 mM DTT. Samples of enzyme (40 nM) were incubated with substrate (25–400 µm) at 37° C for 10 minutes, and the reaction was quenched by the addition of 1.67% (v/v) trifluoroacetic acid (TFA). Fifty microliters of each quenched reaction mixture was separated on a DYNAMAX C18, 5 µm column (4.6 ⫻ 50 mm) equipped with a DYNAMAX C18, 5 µm guard column (4.6 ⫻ 15 mm) (RAININ Instruments, Woburn, MA). Unconverted substrate and hydrolysis products were resolved in a linear acetonitrile gradient (10% solvent B to 40% solvent B in 4 minutes; solvent A: 0.1% TFA in water, solvent B: 0.07% TFA in acetonitrile) and detected by monitoring at 220 nm. Chromatography was performed at 1.5 mL/min at room temperature. Kinetic analysis of the data generated a KM ⫽ 23 µM and kcat of 0.64 sec-1. Enzyme refolding Nucleocapsid Protein-RNA Sample Preparation Purified, denatured enzyme (225 mg) was diluted to 0.16 mg/mL with 8 M urea, and DTT was added to a concentration of 10 mM. The enzyme concentration was determined at 280 nm, using an A280nm of 1.17 for a 1 mg/mL solution, calculated from the amino acid composition (modification of the procedure of Gill and Von Hippel6) and an average molecular mass of 10,779 Da. The solution was diafiltered versus 2.4 L of 25 mM sodium phosphate buffer, pH 7.0, containing 25 mM NaCl, 10 mM DTT, and 10% glycerol with an Amicon-Millipore (Bedford, MA) S1Y10 spiral cartridge. The temperature of the enzyme solution was maintained at 0–5° C by the use of a heat exchanger coupled to a refrigerated recirculator. The enzyme solution was concentrated to 200 mL in the spiral cartridge and diafiltered as described above versus an additional 400 mL of 25 mM sodium phosphate buffer, pH 7.0, containing 25 mM NaCl, 10 mM DTT, and 10% glycerol. Finally, The expression and purification of nucleocapsid protein (NCp7) was described previously.9 Peptides corresponding to the zinc finger regions of NCp7 were synthesized by the University of Michigan Protein and Carbohydrate Structure Facility (Ann Arbor, MI). RNA samples were chemically synthesized using phosphoramidite chemistry, purified by polyacrylamide gel electrophoresis, and characterized by enzymatic sequencing and ESI-MS. Gel-purified RNA samples for ESI-MS binding studies were further desalted by cold ethanol precipitation as ammonium acetate salts and lyophilized.10,11 To assure that the RNA had the proper secondary and tertiary structure, the 10 µM solutions in 10 mM ammonium acetate, pH 6.9, were annealed by heating for 4 minutes at 95° C and cooled slowly (3 hours). Methanol (10% v/v) was added to the protein-RNA solution before ESI-MS experiments to enhance the stability ESI MASS SPECTROMETRY OF HIV PROTEIN COMPLEXES 31 Fig. 1. ESI mass spectra (plotted on the same m/z axis) of HIV integrase catalytic core (F185K) showing only the (a) 18 kDa monomer species detected from a pH 2.5 solution (2:1 v/v acetonitrile:water with 3% acetic acid) and (b) the 36 kDa dimer protein from a pH 6.5 solution (10 mM ammonium acetate). The glass capillary ESI interface was used to acquire the spectra. of the ESI-MS signal without altering the resulting mass spectra. Mass Spectrometry ESI-MS was performed with a double focusing hybrid mass spectrometer (EBqQ geometry, Finnigan MAT 900Q, Bremen, Germany) with a mass-tocharge (m/z) range of 10,000 at 5 kV full acceleration potential.12 A position and time-resolved ion counting (PATRIC) scanning array detector with an 8% m/z range of the m/z centered on the detector was used.13,14 For some of these experiments, an ESI interface based on a heated glass capillary inlet was used.15 This interface uses a countercurrent flow of warm nitrogen gas (⬃60° C) to aid the droplet and ion desolvation process. For the detection of noncovalently bound gas phase complexes, it is important to have fine control of parameters affecting the desolvation process. The energy of gas phase collisions in the ESI interface region is controlled by the voltage difference between the tube lens at the metalized exit of the glass capillary and the first skimmer element (⌬ VTS) and is used to augment the desolvation of the ESI-produced droplets and ions. Additionally, spraying the aqueous solutions ‘‘off axis’’ relative to the capillary inlet produces mass spectra with minimal solvation of the ions and with an increase in sensitivity. Spraying ‘‘on axis’’ relative to the capillary inlet produces broad peaks, as larger droplets are sampled by the MS interface. A stream of SF6 coaxial to the spray suppressed corona discharges, especially important for ESI of aqueous solutions. Solution flow rates delivered to the ESI source were in the 0.5–1.0 µL min-1 range. Other experiments used a heated metal capillary inlet16 with a low flow micro-ESI source.11 Control of the analyte flow rate from 50–200 nL min-1 was accomplished by pneumatic pressure. Capillary temperatures were maintained around 150–200° C for noncovalent complex studies and 210–225°C for other work. A countercurrent gas flow was not necessary for desolvation. However, as for the glass capillary inlet, the voltage difference between the tube lens and the skimmer lens (⌬ VTS) was critical to control the energy of collisions in the interface region. RESULTS AND DISCUSSION HIV Integrase: Protein Dimerization In addition to other viral and host cell proteins, HIV infection requires the activity of the three enzymes encoded by the viral pol gene: protease (PR), reverse transcriptase (RT), and integrase (IN). Integration of a double-stranded DNA copy of the retroviral genome into the host chromosome is essential for viral replication. To facilitate this process, IN catalyzes a coordinated series of DNA cutting and joining reactions. RT and PR are the targets of several HIV inhibitors currently used in clinical practice. However, HIV integrase protein may potentially be a new target for anti-AIDS drugs.17 The crystal structure of the catalytic core of HIV-1 integrase was solved by Dyda et al.18 The limited solubility of the protein hindered its detailed characterization. However, crystallization was made possible by the discovery of a mutation, F185K, which enhances solubility but does not affect the biochemical properties of integrase. Integrase functions minimally as a dimer, consistent with the crystal structure.18–20 The ESI mass spectrum of IN F185K catalytic core (residues 50–212, Mr 18171) from denaturing solution conditions shows a pattern of multiply charged ions of relatively high charge consistent for the 18 kDa monomeric species (Fig. 1; see Table I for molecular weight measurements). From an aqueous solution at pH 6.5, only three multiply charged ions of relatively low charge (11⫹–13⫹) and at relatively high mass-to-charge (m/z ⬎ 3,000) from m/z 3,000– 32 J.A. LOO ET AL. TABLE I. ESI-MS Molecular Weight Measurements HIV integrase Monomer† Dimer HIV protease Monomer† Dimer Dimer ⫹ pepstatin A HIV NCp7 Apo-protein† Holo-protein (⫹2 Zn2⫹) SL3 -RNA‡ Holo-NCp7 ⫹ SL3 -RNA N-terminal Zn-finger peptide Apo-peptide† Holo-peptide (⫹1 Zn2⫹) C-terminal Zn-finger peptide Apo-peptide† Holo-peptide (⫹1 Zn2⫹) Apo-peptide (mutant)† Holo-peptide (mutant, ⫹1 Zn2⫹) Mr (sequence) Mr (measured) 18170.8 36341.6 18172.0 ⫾ 0.4 36345.9 ⫾ 4.7 10778.8 21557.6 22243.5 10780.0 ⫾ 0.6 21555.5 ⫾ 1.8 22242.0 ⫾ 2.8 6369.5 6496.2 4524.8 11021.0 6369.7 ⫾ 0.3 6496.6 ⫾ 0.4 4525.0 ⫾ 0.5 11023.0 ⫾ 0.5 1978.4 2041.7 1977.7 ⫾ 0.2 2041.3 ⫾ 0.3 2068.4 2131.8 2068.4 2068.0 ⫾ 0.1 2131.4 ⫾ 0.3 2067.9 ⫾ 0.2 2131.8 2131.1 ⫾ 0.2 †ESI-MS measurements from acetonitrile/water with 2–5% (v/v) acetic acid solutions. ‡Negative ion ESI-MS measurement from acetonitrile/water with 0.8% (v/v) ammonium hydroxide solution. 3,700 representing the 36 kDa dimeric species are observed. The production of a limited number of charge state ions is believed to be related to the protein’s native, compact structure.21 Higher protein concentrations do not result in observation of higher order aggregates in the mass spectrum. Increasing ⌬VTS of the atmosphere/vacuum interface from ⫹50 V to ⫹140 V promotes dissociation of the gas phase dimer to the monomer. Further increase of ⌬VTS results in dissociation of covalent polypeptide backbone bonds that provide confirmation of the primary structure (data not shown). Fragment ions of yn-type localized near Pro amino acids are preferentially formed. The preference for bond dissociation near proline residues was discussed by Loo et al.22 By carefully controlling instrumental parameters necessary for ion desolvation and increasing solution pH to near physiological conditions, ions for the specific noncovalently bound IN catalytic core dimer can be detected by ESI-MS. HIV Protease: Dimerization and Inhibitor Binding HIV-1 protease is a member of the aspartyl proteinase family of enzymes, including pepsin, renin, chymosin, and cathepsin D.23 PR catalyzes specific cleavage of the viral gag and gag-pol translation products to yield the viral enzymes and structural proteins required for assembly of the virions. It is a homodimeric protein made up of two 99-residue polypeptide chains. Enzymatic activity of HIV-1 PR maximizes near pH 6.24 Dissociation of the active dimeric form of the enzyme results in complete loss of activity. The stability constant, KD, of the HIV-1 PR dimer is strongly pH dependent and is 0.75 nM at pH 5.5 and 48 nM at pH 7.5 (30° C).25 Pepstatin A (isovaleryl-Val-Val-Sta-Ala-Sta, Mr 686) is a specific inhibitor of the aspartyl proteinases.26,27 Binding of the inhibitor is associated with significant conformational changes in the enzyme.28,29 The active site of HIV PR is composed of equal contributions from each of the subunits. Interactions between the inhibitor and PR involve both hydrogen bonding and hydrophobic forces with residues in the body of the protein and with residues in the flap regions. Binding of pepstatin A to HIV-1 protease as a function of pH was investigated by ESI-MS. Pepstatin A was added to an aqueous solution containing PR; the excess inhibitor was removed by centrifugal ultrafiltration. Using the heated metal capillary interface and the low-flow micro-ESI source, MS of PR (monomer Mr 10779) in acetonitrile/water and pH 2.5 (2.5% v/v acetic acid) yields only ions for the monomeric protein (Fig. 2; Table I). Aliquots of the stock PR-inhibitor solution were diluted into ammonium acetate solutions at various pH values (adjusted by addition of acetic acid or ammonium hydroxide) and analyzed by MS. ESI mass spectra acquired from pH 4.7–6.9 solutions show ions for the ternary PR dimer-inhibitor complex as well as uncomplexed PR dimer and monomer (Fig. 3). However, the relative abundance of PR dimer-inhibitor complex was higher at pH less than 6 (Fig. 3a,b) compared with the pH 6.9 spectrum (Fig. 3c). Moreover, ESI-MS of the PR-inhibitor solution at pH 7.4 shows only ions for the PR monomer (Fig. 3d), i.e., evidence for PR dimerization or complexation with pepstatin A was not observed. These results are generally consistent with the solution stability and enzymatic activity of PR dimer. Maximum stability of the dimer and ternary complexes are approximately at pH 5.5–6.0. The HIV PR-pepstatin A ESI-MS results are similar to a previous report published by Baca and Kent.30 The stability of the PR-pepstatin A gas phase complex is highly sensitive to the ESI atmosphere/ vacuum interface conditions. The relative abundance of the ternary complex maximized with low ⌬VTS (⬍10 V) and low metal capillary temperatures (Tcap ⬍ 175° C). Raising ⌬ VTS and/or Tcap results in dissociation of the PR dimer and PR dimer-pepstatin A complexes to the PR monomer. For example, with ⌬VTS ⫽ 0 V and Tcap set to 200° C, a significant amount of the PR dimer complexes dissociated to the monomeric species. An important internal water molecule (water-301) observed in nearly every crystal structure of the PR-inhibitor complex27,28,31,32 was not observed in our study nor in Baca and Kent’s report.30 Although this water molecule is highly coordinated to the PR structure, bridging the flaps of the enzyme to the ESI MASS SPECTROMETRY OF HIV PROTEIN COMPLEXES 33 Fig. 2. ESI mass spectrum of HIV-1 protease/pepstatin A solution (10 µM concentration) under denaturing conditions (2:1 v/v acetonitrile:water with 2.5% acetic acid, pH 2.5). The inset shows isotopic resolution of the (M⫹12H)12⫹ ion. The heated metal capillary interface (Tcap 200° C) and low-flow micro-ESI source (150 nL/min) were used. Fig. 3. Mass spectra of the HIV-1 protease/ pepstatin A mixture acquired from 10 mM ammonium acetate solutions (pH adjusted by addition of acetic acid or ammonium hydroxide). The heated metal capillary temperature was 150° C. inhibitor, the residence time is estimated to be much shorter relative to PR dimerization and inhibitor binding. Water molecules bound to proteins have high mobility and are in rapid motion and exchange in solution, even those that appear to be fixed in a crystal.33 The half-life of dimer dissociation is approximately 30 minutes and inhibitor binding around 1 second,25 whereas the interaction of water with PR has been estimated to be 10 nsec.32 Such kinetic parameters may conspire to preclude the observation of water molecule binding to the gas phase complex by MS. Water clusters or adducts to polypeptides have been observed by ESI-MS.34–36 However, whether these water molecules are structurally relevant to the solution structure is a question that remains to be answered. In general, the loss or 34 J.A. LOO ET AL. Fig. 4. Amino acid sequences of nucleocapsid protein (NCp7) and the synthetic peptides corresponding to the N-terminal and C-terminal zinc fingers used for this study. Fig. 5. Deconvoluted (to the mass domain) ESI mass spectra of nucleocapsid protein (NCp7, 20 µM concentration, 10 mM ammonium acetate, pH 6.9, analyte flow rate 150 nL/min) in the (a) absence and (b) presence of 10 times excess zinc ion (Tcap 150° C). A maximum zinc stoichiometry of two is measured in the presence of excess zinc. The insets show the isotopically resolved deconvoluted mass spectra. dissociation of water from the gas phase complex is observed by MS. Recent results from Chung et al.37 suggest that for selected protein complexes, water molecules important for the solution phase structure may be retained in the gas phase. HIV Nucleocapsid Protein: Zinc and RNA Binding Zinc finger proteins contain Cys and His residues that coordinate zinc and form structural motifs for interacting with nucleic acids, as displayed by several transcription factors that include zinc finger structures for DNA recognition. HIV nucleocapsid protein NCp7 contains two zinc fingers (or knuckles) of type Cys-X2-Cys-X4-His-X4-Cys (where X is a variable amino acid) that are involved in encapsulation of genomic RNA and viral assembly (Fig. 4).38,39 We40 and others41–43 have used ESI-MS to determine the zinc-binding stoichiometry for NCp7 (2 mol of zinc per mol of protein). An example is shown in Figure 5. Two protons are lost for each zinc ion complexed for CCHC-type zinc finger structures.44 The time course of ejection of zinc from NCp7 caused by the covalent binding of various inhibitors can be monitored by MS.40,42,43 The relative zinc-binding affinity of peptides corresponding to each knuckle can be investigated as well.45 For example, an ESI mass spectrum from a solution containing an equimolar concentration of peptides corresponding to the N-terminal and C-terminal knuckles in the presence of zinc is shown in Figure 6. Each peptide binds a maximum of one zinc ion. However, a competitive binding experiment in which the C-terminal knuckle peptide (Lys-Gly-Cys- ESI MASS SPECTROMETRY OF HIV PROTEIN COMPLEXES 35 Fig. 6. Deconvoluted (to the mass domain) ESI mass spectra from an equimolar mixture (50 µM each) of the N-terminal zinc finger peptide and the (top) C-terminal zinc finger peptide and the (bottom) mutant C-terminal peptide (pH 6.9, 10 mM ammonium acetate) in the presence of zinc. The competitive binding experiment shows greatly reduced zinc affinity for the mutant peptide, as the major species observed for the mutant peptide is the 2,068 Da apoform. Fig. 7. ESI mass spectra from a solution containing an equimolar concentration of NCp7 and SL3 -RNA (10 µM each, 10 mM ammonium acetate, pH 6.9) in the (a) absence and (b) presence of excess zinc ion. Specific formation of the 1:1 stoichiometry protein/RNA complex (where the protein is bound to 2 mol of zinc; peaks labeled with filled circles) is observed in the presence of zinc. Instrumental conditions used were the same as described for Figure 5. Trp-Lys-Cys-Gly-Lys-Glu-Gly-His-Gln-Met-Lys-AspCys-Thr-Glu) is replaced with a mutant peptide where the Cys6-Gly7 stretch is reversed to Gly6-Cys7 shows a greatly reduced zinc-binding affinity of the mutant C-terminal peptide relative to that of the N-terminal peptide (Fig. 6, bottom). Most of the 36 J.A. LOO ET AL. mutant peptide appears as the apo-form under these competitive binding conditions. The sequence and residue spacings are very important for zinc coordination. Experiments performed with only the mutant peptide in the presence of excess zinc show a maximum zinc stoichiometry of one (data not shown; Table I). Located between the 58 long terminal repeat (LTR) and the gag initiation codon is the RNA sequence referred to as the Psi () site, important for encapsidation. The N-terminal zinc finger element is critical for nucleocapsid binding to -RNA. The structure of -RNA contains four double-stranded stem/singlestranded loop (or stem-loop [SL]) elements.46 Three of the four stem-loops show relatively high affinity to NCp7; however, SL3 has been proposed as a particularly important region of -RNA. De Guzman et al.47 recently reported the NMR structure of the NCp7SL3 complex. A 1:1 protein-RNA complex with a dissociation constant of ⬃100 nM was observed. The positive ion ESI mass spectrum of an equimolar solution of NCp7 and an RNA consisting of the sequence for SL3 -RNA (CUAGCGGAGGCUAG, Mr 4525) is shown in Figure 7. Without any zinc added to the solution, the most abundant ions observed represent the apo-form of NCp7 and SL3 RNA. We had previously reported the observation of multiply charged positive ions from RNA and DNA.11,48 Adding zinc to the solution results in the formation of ions for the 1:1 stoichiometry NCp7-Zn2/ SL3 RNA complex (Fig. 7b). Zinc induces structural changes and stabilization in the NCp7 molecule, i.e., the zinc finger structures must be fully functional for RNA binding. Such conformational ordering is a prerequisite for RNA binding. A similar ESI-MS experiment was demonstrated by Nemirovskiy et al.,49 in which calcium-binding was required for the EF-hand calcium-binding protein, calmodulin, to bind specifically to the peptide melittin. Unlike the integrase and protease dimers and the PR dimer-pepstatin A ternary complex, the gas phase nucleocapsid protein-RNA complex is extremely stable. The gas phase protein-RNA complex remains intact at elevated capillary temperatures (Tcap to 250° C) and ⌬ VTS values. The relative stabilities of noncovalent complexes held together largely by electrostatic forces, such as positively charged proteins and anionic oligonucleotides, was recently discussed.3,11 Electrostatic forces may be strengthened in the absence of solvent (i.e., in vacuo), whereas complexes held together by hydrophobic interactions are extremely labile in the gas phase. CONCLUSIONS ESI-MS can be used to study protein complexes important for HIV replication. It is important to understand the structure and biological function of a number of key macromolecules to choose targets for drug discovery and to design potential inhibitors for AIDS therapies. In the examples described in this article and many other published accounts,3 the data from ESI-MS experiments of noncovalent complexes are consistent with known solution phase characteristics. It is this feature that provides the driving force for applying ESI-MS to the study of noncovalent complexes. Whether the gas phase complex has a similar or related structure to its solution phase counterpart is a matter of debate.50 In some isolated examples, the gas phase and solution phase structures share similar characteristics.51 Even if these cases were the exception, however, MS is extremely useful for biochemical research as a bioanalytical tool for measuring binding stoichiometry and complex dynamics in a rapid, accurate, and sensitive manner. ACKNOWLEDGMENTS The authors thank Margaret Whitton for N-terminal sequencing, Patricia J. Harvey for assistance in the expression and isolation of NCp7 protein, and John Domagala and Michael Reily for many helpful discussions. REFERENCES 1. Katta, V., Chait, B.T. Observation of the heme-globin complex in native myoglobin by electrospray-ionization mass spectrometry. J. Am. Chem. Soc. 113:8534–8535, 1991. 2. Ganem, B., Li, Y.-T., Henion, J.D. Detection of noncovalent receptor-ligand complexes by mass spectrometry. J. Am. Chem. Soc. 113:6294–6296, 1991. 3. Loo, J.A. Studying noncovalent protein complexes by electrospray ionization mass spectrometry. Mass Spectrom. Rev. 16:1–23, 1997. 4. Estape, D., Rinas, U. Optimized procedures for purification and solubilization of basic fibroblast growth factor inclusion bodies. Biotechnol. Tech. 10:481–484, 1996. 5. Marston, F.A.O., Lowe, P.A., Doel, M.T., Schoemaker, J.M., White, S., Angal, S. Purification of calf prochymosin (prorennin) synthesized in Escherichia coli. BioTechnology 2:800– 804, 1984. 6. Gill, S.C., Von Hippel, P.H. Calculation of protein extinction coefficients from amino acid sequence data. Anal. Biochem. 182:319–326, 1989. 7. Pennington, M.W., Festin, S.M., Maccecchini, M.L., Dick, F., Scarborough, P.E. HIV protease, chromogenic substrate and inhibitor. In: ‘‘Peptides 1990, Proceedings of the 21st European Peptide Symposium.’’ Giralt, E., Andreu, D., (eds.). Leiden, The Netherlands: ESCOM, 1991:787–789. 8. Pennington, M.W., Zaydenberg, I., Byrnes, M.E., et al. A strategy for characterizing new viral proteases: Applied to HIV PR and HAV 3C PR. In: ‘‘Peptides 1992, Proceedings of the 22nd European Peptide Symposium.’’ Schneider, C.H., Eberle, A.N., (eds.). Leiden, The Netherlands: ESCOM, 1993:936–937. 9. Tummino, P.J., Scholten, J.D., Harvey, P.J., et al. The in vitro ejection of zinc from human immunodeficiency virus (HIV) type 1 nucleocapsid protein by disulfide benzamides with cellular anti-HIV activity. Proc. Natl. Acad. Sci. USA 93:969–973, 1996. 10. Limbach, P.A., Crain, P.F., McCloskey, J.A. Molecular mass measurement of intact ribonucleic acids via electrospray ionization quadrupole mass spectrometry. J. Am. Soc. Mass Spectrom. 6:27–39, 1995. 11. Sannes-Lowery, K.A., Hu, P., Mack, D.P., Mei, H.-Y., Loo, J.A. HIV-1 tat peptide binding to TAR RNA by electrospray ionization mass spectrometry. Anal. Chem. 69:5130–5135, 1997. ESI MASS SPECTROMETRY OF HIV PROTEIN COMPLEXES 12. Loo, J.A., Ogorzalek Loo, R.R., Andrews, P.C. Primary to quaternary protein structure determination with electrospray ionization and magnetic sector mass spectrometry. Org. Mass Spectrom. 28:1640–1649, 1993. 13. Loo, J.A., Pesch, R. Sensitive and selective determination of proteins with electrospray ionization magnetic sector mass spectrometry and array detection. Anal. Chem. 66: 3659–3663, 1994. 14. Loo, J.A., Ogorzalek Loo, R.R. Applying charge discrimination with electrospray ionization-mass spectrometry to protein analysis. J. Am. Soc. Mass Spectrom. 6:1098–1104, 1995. 15. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F., Whitehouse, C.M. Electrospray ionization for mass spectrometry of large biomolecules. Science 246:64–71, 1989. 16. Chowdhury, S.K., Katta, V., Chait, B.T. An electrosprayionization mass spectrometer with new features. Rapid Commun. Mass Spectrom. 4:81–87, 1990. 17. Thomas, M., Brady, L. HIV integrase: a target for AIDS therapeutics. Trends Biotechnol. 15:167–172, 1997. 18. Dyda, F., Hickman, A.B., Jenkins, T.M., Engelman, A., Craigie, R., Davies, D.R. Crystal structure of the catalytic domain of HIV-1 integrase: similarity to other polynucleotidyl transferases. Science 266:1981–1986, 1994. 19. Hickman, A.B., Palmer, I., Engelman, A., Craigie, R., Wingfield, P. Biophysical and enzymatic properties of the catalytic domain of HIV-1 integrase. J. Biol. Chem. 269: 29279–29287, 1994. 20. Jenkins, T.M., Engelman, A., Ghirlando, R., Craigie, R. A soluble active mutant of HIV-1 integrase. Involvement of both the core and carboxyl-terminal domains in multimerization. J. Biol. Chem. 271:7712–7718, 1996. 21. Light-Wahl, K.J., Schwartz, B.L., Smith, R.D. Observation of the noncovalent quaternary associations of proteins by electrospray ionization mass spectrometry. J. Am. Chem. Soc. 116:5271–5278, 1994. 22. Loo, J.A., Edmonds, C.G., Smith, R.D. Tandem mass spectrometry of very large molecules. 2. Dissociation of multiply charged proline-containing proteins from electrospray ionization. Anal. Chem. 65:425–438, 1993. 23. Wlodawer, A., Miller, M., Jaskolski, M., et al. Conserved folding in retroviral proteases: Crystal structure of a synthetic HIV-1 protease. Science 245:616–621, 1989. 24. Smith, R., Brereton, I.M., Chai, R.Y., Kent, S.B.H. Ionization states of the catalytic residues in HIV-1 protease. Nat. Struct. Biol. 3:946–950., 1996. 25. Darke, P.L., Jordan, S.P., Hall, D.L., Zugay, J.A., Shafer, J.A., Kuo, L.C. Dissociation and association of the HIV-1 protease dimer subunits: equilibrium and rates. Biochemistry 33:98–105, 1994. 26. Huff, J.R. HIV protease: a novel chemotherapeutic target for AIDS. J. Med. Chem. 34:2305–2327, 1991. 27. Wlodawer, A., Erickson, J.W. Structure-based inhibitors of HIV-1 protease. Annu. Rev. Biochem. 62:543–585, 1993. 28. Fitzgerald, P.M.D., McKeever, B.M., VanMiddlesworth, J.F., et al. Crystallographic analysis of a complex between immunodeficiency virus type 1 protease and acetylpepstatin at 2.0-Å; resolution. J. Biol. Chem. 265:14209– 14219, 1990. 29. Boulanger, Y., Senécal, L., Sauvé, G. Solution structure of the HIV protease inhibitor acetyl-pepstatin as determined by NMR and molecular modeling. J. Biomol. Struct. Dyn. 14:421–428, 1997. 30. Baca, M., Kent, S.B.H. Direct observation of a ternary complex between the dimeric enzyme HIV-1 protease and a substrate-based inhibitor. J. Am. Chem. Soc. 114:3992– 3993, 1992. 31. Miller, M., Schneider, J., Sathyanarayana, B.K., et al. Structure of complex of synthetic HIV-1 protease with a substrate-based inhibitor at 2.3 Å; resolution. Science 246:1149–1152, 1989. 32. Wang, Y.-X., Freedberg, D.I., Grzesiek, S., et al. Mapping hydration water molecules in the HIV-1 protease/DMP323 complex in solution by NMR spectroscopy. Biochemistry 35:12694–12704, 1996. 37 33. Levitt, M., Park, B.H. Water: Now you see it, now you don’t. Structure 1:223–226, 1993. 34. Smith, R.D., Light-Wahl, K.J. Perspectives—The observation of non-covalent interactions in solution by electrospray ionization mass spectrometry - promise, pitfalls and prognosis. Biol. Mass Spectrom. 22:493–501, 1993. 35. Rodriguez-Cruz, S.E., Klassen, J.S., Williams, E.R. Hydration of gas-phase gramicidin S (M⫹2H)2⫹ ions formed by electrospray: The transition from solution to gas-phase structure. J. Am. Soc. Mass Spectrom. 8:565–568, 1997. 36. Loo, J.A., Sannes-Lowery, K.A., Hu, P., Mack, D.P., Mei, H.-Y. Studying noncovalent protein-RNA interactions and drug binding by electrospray ionization mass spectrometry. In: ‘‘New Methods for the Study of Biomolecular Complexes.’’ Ens, W., Standing, K.G., Chernushevich, I.V. (eds.). Dordrecht, The Netherlands: Kluwer, 1998:83–99. 37. Chung, E.W., Rostom, A., Robinson, C.V. Investigating non-covalent protein-ligand interactions using nanoflow electrospray. In: ‘‘Proceedings of the 45th ASMS Conference on Mass Spectrometry and Allied Topics, Palm Springs, CA, June 1997.’’ Santa Fe, NM: American Society for Mass Spectrometry, 1997:359. 38. Darlix, J.-L., Lapadat-Tapolsky, M., de Rocquigny, H., Roques, B.P. First glimpses at structure-function relationships of the nucleocapsid protein of retroviruses. J. Mol. Biol. 254:523–537, 1995. 39. Gorelick, R.J., Chabot, D.J., Rein, A., Henderson, L.E., Arthur, L.O. The two zinc fingers in the human immunodeficiency virus type 1 nucleocapsid protein are not functionally equivalent. J. Virol. 67:4027–4036, 1993. 40. Loo, J.A., Holler, T.P., Sanchez, J., Gogliotti, R., Maloney, L., Reily, M.D. Biophysical characterization of zinc ejection from HIV nucleocapsid protein by anti-HIV 2,2’dithiobis[benzamides] and benzisothiazolones. J. Med. Chem. 39: 4313–4320, 1996. 41. Surovoy, A., Waidelich, D., Jung, G. Nucleocapsid protein of HIV-1 and its Zn2⫹ complex formation analysis with electrospray mass spectrometry. FEBS Lett. 311:259–262, 1992. 42. Yu, X.L., Hathout, Y., Fenselau, C., et al. Specific disulfide formation in the oxidation of HIV-1 zinc finger protein nucleocapsid p7. Chem. Res. Toxicol. 8:586–590, 1995. 43. Hathout, Y., Fabris, D., Han, M.S., Sowder, R.C., II, Henderson, L.E., Fenselau, C. Characterization of intermediates in the oxidation of zinc fingers in human immunodeficiency virus type 1 nucleocapsid protein P7. Drug Metab. Dispos. 24:1395–1400, 1996. 44. Fabris, D., Zaia, J., Hathout, Y., Fenselau, C. Retention of thiol protons in two classes of protein zinc ion coordination centers. J. Am. Chem. Soc. 118:12242–12243, 1996. 45. Mély, Y., de Rocquigny, H., Morellet, N., Roques, B.P., Gérard, D. Zinc binding to the HIV-1 nucleocapsid protein: a thermodynamic investigation by fluorescence spectroscopy. Biochemistry 35:5175–5182, 1996. 46. Sakaguchi, K., Zambrano, N., Baldwin, E.T., et al. Identification of a binding site for the human immunodeficiency virus type 1 nucleocapsid protein. Proc. Natl. Acad. Sci. USA 90:5219–5223, 1993. 47. De Guzman, R.N., Wu, Z.R., Stalling, C.C., Pappalardo, L., Borer, P.N., Summers, M.F. Structure of the HIV-1 nucleocapsid protein bound to the SL3 -RNA recognition element. Science 279:384–388, 1998. 48. Sannes-Lowery, K.A., Mack, D.P., Hu, P., Mei, H.-Y., Loo, J.A. Positive ion electrospray ionization mass spectrometry of oligonucleotides. J. Am. Soc. Mass Spectrom. 8:90–95, 1997. 49. Nemirovskiy, O.V., Ramanathan, R., Gross, M.L. Investigation of calcium-induced, noncovalent association of calmodulin with melittin by electrospray ionization mass spectrometry. J. Am. Soc. Mass Spectrom. 8:809–812, 1997. 50. Wolynes, P.G. Biomolecular folding in vacuo!!!(?). Proc. Natl. Acad. Sci. USA 92:2426–2427, 1995. 51. Loo, J.A., He, J.-X., Cody, W.L. Higher order structure in the gas phase reflects solution structure. J. Am. Chem. Soc. 120:4542–4543, 1998.