Revisiting monomeric HIV-1 protease. Characterization and redesign for improved properties

Interactions between the C-terminal interface residues (96-99) of the mature HIV-1 protease were shown to be essential for dimerization, whereas the N-terminal residues () and Arg(87) contribute to dimer stability (Ishima, R., Ghirlando, R., Tozser, J., Gronenborn, A. M., Torchia, D. A., and Louis, J. M. (2001) J. Biol. Chem. 276, 49110-49116). Here we show that the intramonomer interaction between the side chains of Asp(29) and Arg(87) influences dimerization significantly more than the intermonomer interaction between Asp(29) and Arg(8'). Several mutants, including T26A, destablize the dimer, exhibit a monomer fold, and are prone to aggregation. To alleviate this undesirable property, we designed proteins in which the N- and C-terminal regions can be linked intramolecularly by disulfide bonds. In particular, cysteine residues were introduced at positions 2 and 97 or 98. A procedure for the efficient preparation of intrachain-linked polypeptides is presented, and it is demonstrated that the Q2C/L97C variant exhibits a native-like single subunit fold. It is anticipated that monomeric proteases of this kind will aid in the discovery of novel inhibitors aimed at binding to the monomer at the dimerization interface. This extends the target area of current inhibitors, all of which bind across the active site formed by both subunits in the active dimer.

Biosynthetically directed fractional 13C labeling facilitates identification of Phe and Tyr aromatic signals in proteins

Analysis of 2D [(13)C,(1)H]-HSQC spectra of biosynthetic fractionally (13)C labeled proteins is a reliable, straightforward means to obtain stereospecific assignments of Val and Leu methyl sites in proteins. Herein we show that the same fractionally labeled protein sample facilitates observation and identification of Phe and Tyr aromatic signals. This is the case, in part, because the fractional (13)C labeling yields aromatic rings in which some of the (13)C-(13)C J-couplings, present in uniformly labeled samples, are absent. Also, the number of homonuclear J-coupling partners differs for the delta-, epsilon- and zeta-carbons. This enabled us to vary their signal intensities in distinctly different ways by appropriately setting the (13)C constant-time period in 2D [(13)C,(1)H]-HSQC spectra. We illustrate the application of this approach to an 18 kDa protein, c-VIAF, a modulator of apoptosis. In addition, we show that cancellation of the aromatic (13)C CSA and (13)C-(1)H dipolar interactions can be fruitfully utilized in the case of the fractionally labeled sample to obtain high resolution (13)C constant-time spectra with good sensitivity.

An NMR experiment for the accurate measurement of heteronuclear spin-lock relaxation rates

Rotating-frame relaxation rates, R(1)(rho), are often measured in NMR studies of protein dynamics. We show here that large systematic errors can be introduced into measured values of heteronuclear R(1)(rho) rates using schemes which are usually employed to suppress cross-correlation between dipole-dipole and CSA relaxation mechanisms. For example, in a scalar-coupled two-spin X-H spin system the use of (1)H WALTZ16 decoupling or (1)H pulses applied at regularly spaced intervals leads to a significant overestimation of heteronuclear R(1)(rho) values. The problem is studied experimentally and theoretically for (15)N-(1)H and (13)C-(1)H spin pairs, and simple schemes are described which eliminate the artifacts. The approaches suggested are essential replacements of existing methodology if accurate dynamics parameters are to be extracted from spin-lock relaxation data sets.

Rapid structural fluctuations of the free HIV protease flaps in solution: relationship to crystal structures and comparison with predictions of dynamics calculations

Crystal structures have shown that the HIV-1 protease flaps, domains that control access to the active site, are closed when the active site is occupied by a ligand. Although flap structures ranging from closed to semi-open are observed in the free protease, crystal structures reveal that even the semi-open flaps block access to the active site, indicating that the flaps are mobile in solution. The goals of this paper are to characterize the secondary structure and fast (sub-ns) dynamics of the flaps of the free protease in solution, to relate these results to X-ray structures and to compare them with predictions of dynamics calculations. To this end we have obtained nearly complete backbone and many sidechain signal assignments of a fully active free-protease construct that is stabilized against autoproteolysis by three point mutations. The secondary structure of this protein was characterized using the chemical shift index, measurements of (3h)J(NC') couplings across hydrogen bonds, and NOESY connectivities. Analysis of these measurements indicates that the protease secondary structure becomes irregular near the flap tips, residues 49-53. Model-free analysis of (15)N relaxation parameters, T(1), T(2) (T(1rho)) and (15)N-[(1)H] NOE, shows that residues in the flap tips are flexible on the sub-ns time scale, in contrast with previous observations on the inhibitor-bound protease. These results are compared with theoretical predictions of flap dynamics and the possible biological significance of the sub-ns time scale dynamics of the flap tips is discussed.

Folded monomer of HIV-1 protease

The mature human immunodeficiency virus type 1 protease rapidly folds into an enzymatically active stable dimer, exhibiting an intricate interplay between structure formation and dimerization. We now show by NMR and sedimentation equilibrium studies that a mutant protease containing the R87K substitution (PR(R87K)) within the highly conserved Gly(86)-Arg(87)-Asn(88) sequence forms a monomer with a fold similar to a single subunit of the dimer. However, binding of the inhibitor DMP323 to PR(R87K) produces a stable dimer complex. Based on the crystal structure and our NMR results, we postulate that loss of specific interactions involving the side chain of Arg(87) destabilizes PR(R87K) by perturbing the inner C-terminal beta-sheet (residues 96-99 from each monomer), a region that is sandwiched between the two beta-strands formed by the N-terminal residues (residues 1-4) in the mature protease. We systematically examined the folding, dimerization, and catalytic activities of mutant proteases comprising deletions of either one of the terminal regions (residues 1-4 or 96-99) or both. Although both N- and C-terminal beta-strands were found to contribute to dimer stability, our results indicate that the inner C-terminal strands are absolutely essential for dimer formation. Knowledge of the monomer fold and regions critical for dimerization may aid in the rational design of novel inhibitors of the protease to overcome the problem of drug resistance.

Use of the carbonyl chemical shift to relieve degeneracies in triple-resonance assignment experiments

We illustrate an approach that uses the backbone carbonyl chemical shift to relieve resonance overlaps in triple-resonance assignment experiments conducted on protein samples. We apply this approach to two cases of simultaneous overlaps: those of ((1)H(N), (15)N) spin pairs and those of ((1)H(alpha), (13)C(alpha)) spin pairs in residues preceding prolines. For these cases we employed respectively CBCACO(N)H and H(CA)CON experiments, simple variants of the commonly used CBCA(CO)NH and HCA(CO)N experiments obtained by replacing one of the indirect dimensions with a carbonyl dimension. We present data collected on ribosomal protein S4 using these experiments, along with overlap statistics for four other polypeptides ranging in size from 76 to 263 residues. These data indicate that the CBCACO(N)H, in combination with the CBCA(CO)NH, can relieve >83% of the ((1)H(N), (15)N) and ((1)H(N), (13)C') overlaps for these proteins. The data also reveal how the H(CA)CON experiment successfully completed the assignment of triply and quadruply degenerate X-Pro spin systems in a mobile, proline-rich region of S4, even when X was a glycine. Finally, we discuss the relative sensitivities of these experiments compared to those of existing sequences, an analysis that reinforces the usefulness of these experiments in assigning extensively overlapped and/or proline-rich sequences in proteins.

Optimized labeling of 13CHD2 methyl isotopomers in perdeuterated proteins: potential advantages for 13C relaxation studies of methyl dynamics of larger proteins

13CHD2 methyl isotopomers are particularly useful to study methyl dynamics in proteins because, as compared with other methyl isotopomers, the 13C relaxation mechanism for this isotopomer is straightforward. However, in the case of proteins, where (omega tau)2 >> 1, the refocused INEPT pulse sequence does not completely suppress unwanted 13CH3 signals. The presence of weak 13CH3 peaks is usually not a serious problem for smaller proteins because there are relatively few methyl signals and they are sharp; however, signal overlap becomes more common as the size of the protein increases. We overcome this problem by preparing a protein using a 98% D2O cell culture medium containing 3-(13)C pyruvic acid, 50-60% deuterated at the 3-position, and 4-(13)C 2-ketobutyric acid, 98% and 62% deuterated at the 3- and 4-positions, respectively. This approach significantly reduces the population of the CH3 isotopomer while optimizing the production of 13CHD2, the isotopomer desired for 13C relaxation measurements. In larger proteins where the deuterium T2 may be too short to measure accurately, we also suggest the alternative measurement of the proton T2 of the 13CH2D methyl isotopomer, because these protons are well-isolated from other protons in these highly deuterated samples.

Comparison of methyl rotation axis order parameters derived from model-free analyses of (2)H and (13)C longitudinal and transverse relaxation rates measured in the same protein sample

Recombinant HIV-1 protease was obtained from bacteria grown on a 98% D(2)O medium containing 3-(13)C pyruvic acid as the sole source of (13)C and (1)H. The purified protein is highly deuterated at non-methyl carbons, but contains significant populations of (13)CHD(2) and (13)CH(2)D methyl isotopomers. This pattern of isotope labeling permitted measurements of (1)H and (13)C relaxation rates of (13)CHD(2) isotopomers and (2)H (D) relaxation rates of (13)CH(2)D isotopomers using a single sample. The order parameters S(axis)(2), which characterize the motions of the methyl rotation axes, were derived from model-free analyses of R(1) and R(2) data sets measured for (13)C and (2)H spins. Our primary goal was to compare the S(axis)(2) values derived from the two independent types of data sets to test our understanding of the relaxation mechanisms involved. However, S(axis)(2) values derived from the analyses depend strongly on the geometry of the methyl group, the sizes of the quadrupolar and dipolar couplings, and the effects of bond vibrations and librations on these couplings. Therefore uncertainties in these basic physical parameters complicate comparison of the order parameters. This problem was circumvented by using an experimental relationship, between the methyl quadrupolar, (13)C-(13)C and (13)C-(1)H dipolar couplings, derived from independent measurements of residual static couplings of weakly aligned proteins by Ottiger and Bax (J. Am. Chem. Soc. 1999, 121, 4690-4695) and Mittermaier and Kay (J. Am. Chem. Soc. 1999, 121, 10608-10613). This approach placed a tight experimental restraint on the values of the (2)H quadrupolar and (13)C-(1)H dipolar interactions and greatly facilitated the accurate comparison of the relative values of the order parameters. When applied to our data this approach yielded satisfactory agreement between the S(axis)(2) values derived from the (13)C and (2)H data sets.