Analysis of conformational changes during activation of protein kinase Pak2 by amide hydrogen/deuterium exchange

Abstract

During apoptotic stress, protein kinase Pak2 is cleaved by caspase 3 to form a heterotetramer that is constitutively activated following autophosphorylation. The active protein kinase migrates slightly slower than the inactive holoenzyme when analyzed by gel filtration, suggesting an expanded conformation. Activation of Pak2 comprises a series of structural changes resulting from caspase cleavage, ATP binding, and autophosphorylation of Pak2. Changes at each step were individually analyzed by amide hydrogen/deuterium exchange coupled with mass spectrometry and compared with inactive Pak2. The auto-inhibited form was shown to bind ATP in the active site, with minor changes in the glycine loop and the autoinhibitory domain (AID). Caspase cleavage produced significant changes in solvent accessibility in the AID and upper lobe of the catalytic domain. Cleavage of ATP-bound Pak2 relaxes the allosteric inhibition, as shown by increased solvent accessibility in the upper and lower lobes, including the G-helix, facilitating the autophosphorylation of two sites required for activation, Ser-141 in the regulatory domain and Thr-402 in the catalytic domain. Autophosphorylation increased the amide hydrogen/deuterium exchange solvent accessibility of the contact region between the AID and the G-helix, the E-F loop, and the N terminus. Thus, activation of Pak2 via caspase cleavage is associated with structural relaxation of Pak2 that allows for complete auto-phosphorylation, resulting in a more comprehensive solvent-exposed and conformationally dynamic enzyme.

FIGURE 1.

The autoinhibitory regions of Pak1 and PKA interact at different locations in the structurally homologous catalytic domains. The upper left image shows PKA (1CTP.PDB) complexed to a portion of the protein kinase inhibitor (PKI residues 5–24), and the upper right image of Pak1 shows the catalytic domain (residues 249–542) complexed to the autoinhibitory region (residues 78–147) of Pak1 (1F3M.PDB). The lower figure shows the two images overlaid and illustrates that the pseudosubstrate inhibitor PKI residues 5–24 and the AID bind at different locations on the surface of the homologous structures of the catalytic domains.

FIGURE 2.

The effect of caspase cleavage and autophosphorylation on the solution state of Pak2. a, the linear schematic of Pak2. b, caspasecleaved Pak2, intact inactive Pak2 and caspase-cleaved autophosphorylated Pak2 (25 μg) were analyzed by SDS-PAGE and Coomassie Blue staining. c, the samples were individually chromatographed on a Superdex 200 column as described under “Experimental Procedures.” The calculated molecular masses are shown. A second smaller peak, which ran with the dead volume, was not associated with any detectable DNA or protein, and consequently was judged artificial.

FIGURE 3.

Mass spectrum of the peptic digest and time course of deuterium incorporation for inactive Pak2. a, MALDI-TOF mass spectrum of pepsin-digested inactive Pak2. b, inactive Pak2 was subjected to amide H/D exchange for 0–10 min and analyzed by MALDI-TOF. An example of the expanded isotopic distribution of MALDI-TOF of peak m/z 1697.85 during the time course of amide H/D exchange is shown. c, plot of deuteron incorporation into the exchangeable amides of peptide m/z 1697.85. The peptic fragment was imported into the Prism 4.0 program. The two-phase exponential binding function in the program was used for nonlinear data fitting. The y-axis is the number of exchangeable amides on the peptide.

FIGURE 4.

Solvent accessibility of inactive Pak2. a, the Pak2 sequence and the peptic peptides identified after 10 min of H/D exchange of Pak2 are color coded according to level of deuterium incorporation into exchangeable amides. The numbers under the color bars indicate the mass (m/z) of the peptic fragment. b, the results of amide H/D exchange depicted in a are superimposed on the crystal structure of Pak1 (1F3M.PDB).

FIGURE 5.

Comparison of the changes in the level of deuterium incorporation at different states of activation of Pak2. The calculated changes in deuterium incorporation at 10 min for the different activation states of Pak2 are based on the deuteron incorporation number for inactive Pak2. The color index represents the changes in deuterium incorporation. a, AMP-PCP bound to Pak2; b, caspase-cleaved Pak2; c, AMP-PCP bound to caspase-cleaved Pak2; and d, caspase-cleaved autophosphorylated Pak2.