Production, purification, and crystallization of human interleukin-1 beta converting enzyme derived from an Escherichia coli expression system.

Interleukin-1 beta converting enzyme (ICE) is a cysteine protease that catalyzes the conversion of the inactive precursor form of IL-1 beta to an active mature form. The mature form of IL-1 beta is involved in mediating inflammatory responses and in the progression of autoimmune diseases. We recently reported on the production of active human ICE in insect cells using the baculovirus expression system (Wang XM et al., 1994, Gene 145:273-277). Because the levels of expression achieved with this system were limiting for the purpose of performing detailed biochemical and biophysical studies, we examined the production of ICE in Escherichia coli. By using a tac promoter-based expression system and fusion to thioredoxin we were able to recover high levels of active ICE protein. The expressed protein, which was distributed between the soluble and insoluble fractions, was purified to homogeneity from both fractions using a combination of classical and affinity chromatography. Comparisons of ICE derived from both fractions indicated that they were comparable in their specific activities, subunit composition, and sensitivities to specific ICE inhibitors. The combined yields of ICE obtained from the soluble and insoluble fractions was close to 1 mg/L of induced culture. Recombinant human ICE was crystallized in the presence of a specific ICE inhibitor in a form suitable for X-ray crystallographic analysis. This readily available source of ICE will facilitate the further characterization of this novel and important protease.

High-resolution structure of Ascaris trypsin inhibitor in solution: direct evidence for a pH-induced conformational transition in the reactive site.

BACKGROUND: The Ascaris trypsin inhibitor (ATI) is a member of a new family of serine protease inhibitors isolated from the helminthic worm Ascaris lumbricoides var suum. This family comprises five chymotrypsin/elastase inhibitors and one trypsin inhibitor. Members are characterized by the presence of five disulfide bonds (two of which are located on either side of the reactive site) in a single small protein domain of 61-62 residues. RESULTS: The solution structure of ATI has been determined at pH 2.4 and pH 4.75 by NMR spectroscopy. Iterative refinement permitted the unambiguous assignment of the pairing of the five disulfide bridges (Cys5-Cys38, Cys15-Cys33, Cys18-Cys29, Cys22-Cys60, and Cys40-Cys54) which were previously unknown. The structure includes four short beta-strands arranged in two approximately perpendicular beta-sheets. The reactive site loop is bounded by two disulfide bridges (Cys15-Cys33 and Cys18-Cys29) and is part of the long loop (residues 15-25) connecting strands beta 1 and beta 2. Comparison of the nuclear Overhauser enhancement data at the two pH values revealed significant differences centered around the reactive site. While the reactive site at pH 2.4 closely resembles that of other protease inhibitors, at pH 4.75 the reactive site loop undergoes a major conformational rearrangement involving the psi backbone torsion angles of the P2, P1 and P1' residues (residues 30-32). This is associated with a change in the positions of the side chains of Arg31 and Glu32. CONCLUSIONS: The overall three-dimensional structure of ATI posesses an unusual fold and, with the exception of the reactive site, shows no similarity to other serine protease inhibitors. The observation that the reactive site of the low pH form of ATI is similar to that of other serine proteases suggests that this is the active form of the protein.

Analysis of the backbone dynamics of interleukin-8 by 15N relaxation measurements.

The backbone dynamics of the cytokine interleukin-8, a symmetric homodimer of overall molecular mass 16 kDa, has been investigated at pH 5.2 by means of 15N relaxation measurements using heteronuclear two-dimensional inverse detected 1H-15N spectroscopy. 15N T1, T2 and NOE data were obtained for 66 out of a total of 67 backbone amide groups. The overall correlation time is 9.10(+/- 0.05) ns at 26.6 degrees C. All residues exhibit very rapid motions on a time-scale of < or = 20 ps. These very rapid motions alone can account for the 15N relaxation behaviour of 30 residues. The 15N relaxation data for another 21 residues can only be accounted for by the inclusion of an additional internal motion on a time-scale ranging from 0.5 to 3.5 ns. These residues are clustered at the N and C termini, and in the loop regions connecting elements of secondary structures. Finally, the 15N relaxation data for another 15 residues could only be accounted for by the presence of chemical exchange on a time-scale ranging from approximately 170 ns to 2.25 ms. In addition, the inclusion of chemical exchange improved the fit to the experimental data for 10 of the 30 residues whose 15N relaxation behaviour could be accounted for by very fast motions alone. The residues exhibiting chemical exchange line broadening cluster at the interface of the long C-terminal alpha-helix and the underlying beta-sheet. It is suggested that this clustering is indicative of concerted rather than independent motions in regions of secondary structure, with motion at any one residue being propagated to neighbouring residues in van der Waals contact.