Molecular bases for the recognition of short peptide substrates and cysteine-directed modifications of human insulin-degrading enzyme

Abstract

Insulin degrading enzyme (IDE) utilizes a large catalytic chamber to selectively bind and degrade peptide substrates such as insulin and amyloid beta (Abeta). Tight interactions with substrates occur at an exosite located approximately 30 A away from the catalytic center that anchors the N-terminus of substrates to facilitate binding and subsequent cleavages at the catalytic site. However, IDE also degrades peptide substrates that are too short to occupy both the catalytic site and the exosite simultaneously. Here, we use kinins as a model system to address the kinetics and regulation of human IDE with short peptides. IDE specifically degrades bradykinin and kallidin at the Pro/Phe site. A 1.9 A crystal structure of bradykinin-bound IDE reveals the binding of bradykinin to the exosite and not to the catalytic site. In agreement with observed high K(m) values, this suggests low affinity of bradykinin for IDE. This structure also provides the molecular basis on how the binding of short peptides at the exosite could regulate substrate recognition. We also found that human IDE is potently inhibited by physiologically relevant concentrations of S-nitrosylation and oxidation agents. Cysteine-directed modifications play a key role, since an IDE mutant devoid of all 13 cysteines is insensitive to the inhibition by S-nitrosoglutathione, hydrogen peroxide, or N-ethylmaleimide. Specifically, cysteine 819 of human IDE is located inside the catalytic chamber pointing toward an extended hydrophobic pocket and is critical for the inactivation. Thiol-directed modification of this residue likely causes local structural perturbation to reduce substrate binding and catalysis.

Figure 1

Mass spectrometry analysis of bradykinin and kallidin degradation by IDE-CF. (A) MALDI-TOF spectra of bradykinin degradation at 0 min (top) and 30 min (bottom). The amino acidic sequence of bradykinin is shown for the peaks corresponding to the mass of full-length bradykinin (top) and for the cleavage product (bottom). (B) MS-MS spectra of precursor mass 757.38 for bradykinin as detected by MALDI-TOF. De novo prediction of the peptide sequence is labeled on the corresponding y and b ions on the peaks from the spectra and on the chemical structure of bradykinin. (C) and (D) Same as panels A and B for kallidin.

Figure 2

Kinetic analysis of bradykinin and kallidin degradation by IDE. Specific activities (sec⁻¹) of IDE-CF using the indicated bradykinin (A) and kallidin (B) concentrations and 0.5 μM substrate V are shown. The activity was measured at 37 °C for 5 min. K_m values were determined to be 4.2±0.6 mM, and 7.3±0.7 mM for bradykinin and kallidin, respectively. Data utilized for fitting kinetic curves are mean ± SD representative of two independent assays performed in duplicates.

Figure 3

Structural analysis of bradykinin-bound IDE. (A) Stereo view of the interactions between bradykinin and the exosite of IDE. The N-terminal residues Arg(Ala)¹-Pro²-Pro³ of bradykinin are shown as yellow sticks. Protein residues belonging to the exosite are shown as salmon sticks, while their secondary structures are shown as green cartoon. 2Fo-Fc simulated-annealing omit map is shown around bradykinin as blue mesh contoured at 1.5σ. Hydrogen bonds are shown as red dashed lines. (B) Surface representation of IDE-N catalytic chamber with shown positions of the active site (red surface and magenta sphere for catalytic zinc) and exosite (green surface for residues 336–342 and 359–363 of IDE). N-terminal bradykinin is shown as yellow sticks, and distance from active site to exosite is shown as dashed line.

Figure 4

NEM sensitivity of human IDE. Relative activities of IDE-WT, IDE-CF, IDE-CF-123, and IDE-CF-4 (A), and of IDE mutants M1-M5 (B), are shown at two concentrations of NEM (0.2 and 1 mM). Activities were measured after 1hour incubation of IDE samples partially purified by a small-scale protocol (see Experimental procedures for details). (C) NEM-concentration (0–1 mM) dependent relative activities of IDE-WT, IDE-CF-4 and IDE-WT-C819A. (D) NEM sensitivity of fully purified IDE mutants at three different concentrations of NEM (0.15, 0.6, and 2.5 mM) after 1 hour incubation. Samples shown in panel D were fully purified, in large scale, as previously reported (12) to exclude effects of contaminant from less purified samples. Relative activities are reported assuming as 100% the control reaction were each sample was incubated with buffer only. Incubations were performed mixing 1mg/ml purified IDE samples with the shown concentration of NEM, and activity was followed by measuring fluorescence as resulting from the cleavage of 0.5 μM Substrate V.

Figure 5

GSNO and H₂O₂-mediated inhibition of IDE. Concentration dependent inhibition of IDE-WT, IDE-CF123 and IDE-CF4 by GSNO (A) and H₂O₂ (B) as measured by monitoring the degradation of substrate V. Inhibition of IDE measured by monitoring the degradation of either 0.5 μM fluorogenic substrate V (C) or 1.5 μM fluorogenic FAβB (D) in the presence of a single concentration of NEM, GSNO, and H₂O₂.(1 mM).

Figure 6

Structural analysis of the area around IDE cysteine 819. (A-Left) Cartoon representation of IDE highlighting the position of Cys 819 and Cys 812 (cyan balls and sticks), residues involved in substrate binding (red sticks), and catalytic site (magenta sphere and grey sticks). IDE-N and IDE-C are depicted as blue and green cartoons, respectively. The 28-aa loop joining IDE-N and IDE-C is shown as pink ribbon. (A-Right) Close-up view of the Cys 812-Cys819 and active site regions, after a rotation of 180° around the y-axis. Cys 812 and Cys 819 are shown as ball and sticks, their relative distance and the distance to the active site is shown as black dashes. Interactions between residues F820, R824, Y831 and Aβ_1–40 from PDB ID 2G47 (13) are shown as red dashes. The position of the buried Cys 110 close to the active site is also shown. (B-Left) View of the secondary structures in the vicinity of residue Cys 819 (approximately same orientation as in A-Right). Residues and secondary structures depicted as in panel A. Location of other cysteines (110, 178, 789, and 812) not involved in inactivation of IDE is also shown as balls and sticks. (B-Right) Stereo-diagram showing close-up view of the hydrophobic local environment around the side-chain of Cys 819. Hydrophobic residues are depicted as yellow sticks. See text for details.