Expression in Escherichia coli, refolding, and purification of human procathepsin K, an osteoclast-specific protease.

We have constructed and optimized a high yielding Escherichia coli expression system to produce glycosylation-free human procathepsin K and have developed conditions for refolding this enzyme. Recombinant human procathepsin K (EC 3.4.22.38) was expressed in E. coli, refolded from inclusion bodies, and further purified by Superdex 75 size-exclusion chromatography. Purified procathepsin K had a [MH]+ of 35,063 Da which is in agreement with the predicted mass of the construct. Amino-terminal sequence analysis matched the predicted sequence with no secondary sequence detected. Purified procathepsin K activated under autocatalytic conditions to a final specific activity of 23 micromol 7-amido-4-methylcoumarin liberated/min/mg of enzyme using the fluorescent peptide substrate benzyloxycarbonyl-phenylalanine-arginine-7-amido-4-methylcoumarin. This expression and refolding procedure yielded 50 mg of purified, glycosylation-free human procathepsin K from 1 liter of E. coli cell culture and enabled the determination of the structure of human procathepsin K at 2.6 A resolution.

The crystal structure of human procathepsin K.

Cathepsin K is a cysteine protease present in human osteoclasts that plays an important role in bone resorption. Cathepsin K is synthesized as an inactive proenzyme and activated under conditions of low pH. Autoproteolytic processing of the N-terminal 99 amino acid propeptide produces the active, mature form of cathepsin K. It is presumed that the activation of procathepsin K in vivo occurs in the bone resorption pit, which has a low-pH environment. We have determined the structure of human procathepsin K at 2.8 A resolution. The structure of the mature enzyme domain within procathepsin K is virtually identical to that of mature cathepsin K. The fold of the propeptide of procathepsin K is similar to that observed in procathepsins B and L despite differences in length and sequence. A portion of the propeptide occupies the active site cleft of cathepsin K. Hydrophobic interactions, salt bridges, and hydrogen-bonding interactions are observed in the structure of the propeptide and between the propeptide and the mature enzyme of procathepsin K. These interactions suggest an explanation for the stability of the proenzyme. The structure of procathepsin K contributes to an understanding of the molecular basis of inhibition by the propeptide portion of the molecule and activation of this important member of the cysteine protease family.

Inhibition of human immunodeficiency virus type 1 and type 2 Tat function by transdominant Tat protein localized to both the nucleus and cytoplasm.

We introduced various mutations into the activation and RNA binding domains of human immunodeficiency virus type 1 (HIV-1) Tat in order to develop a novel and potent transdominant Tat protein and to characterize its mechanism of action. The different mutant Tat proteins were characterized for their abilities to activate the HIV LTR and inhibit the function of wild-type Tat in trans. A Tat protein containing a deletion of the basic domain (Tat(delta)49-57) localized exclusively to the cytoplasm of transfected human cells was nonfunctional and inhibited both HIV-1 and HIV-2 Tat function in a transdominant manner. Tat proteins containing mutations in the cysteine-rich and core domains were nonfunctional but failed to inhibit Tat function in trans. When Tat nuclear or nucleolar localization signals were fused to the carboxy terminus of Tat(delta)49-57, the chimeric proteins localized to the nucleus or nucleolus, respectively, and remained capable of acting in a transdominant manner. Introduction of secondary mutations in the cysteine-rich and core domains of the various transdominant Tat proteins completely eliminated their abilities to act in a transdominant fashion. Our data best support a mechanism in which these transdominant Tat proteins squelch a cellular factor or factors that interact with the Tat activation domain and are required for Tat to function.

Purification and functional characterization of wild-type and mutant HIV-1 and HIV-2 Tat proteins expressed in Escherichia coli.

We describe a single method to purify milligram amounts of authentic wild-type and mutant HIV-1 and HIV-2 Tat proteins overexpressed in Escherichia coli. This method takes advantage of the highly basic, positively charged RNA binding domain present in both HIV-1 and HIV-2 Tat, which also facilitated purification of HIV-1 and HIV-2 mutant Tat proteins. In contrast to previously described methods, our method does not require use of denaturing or reducing agents, since Tat is present in the soluble fraction after bacterial lysis. The activity of purified wild-type and mutant HIV-1 and HIV-2 Tat proteins was determined using cell-based uptake, in vitro transcription, and TAR binding assays. As expected, both HIV-1 and HIV-2 Tat efficiently transactivated the HIV-2 LTR, whereas HIV-2 Tat transactivated the HIV-1 LTR less efficiently than HIV-1 Tat. Purified HIV-2 Tat proteins in which the glutamic acid residue at position 77 was changed to either glycine or glutamine transactivated the HIV-1 LTR more efficiently than wild-type Tat-2, providing additional evidence that the net charge of this region may be responsible for nonreciprocal transactivation between Tat-1 and Tat-2. Our results demonstrate that this method can be used to rapidly purify authentic wild-type and mutant Tat proteins which are suitable for cell-based and in vitro functional studies.

Extracellular human immunodeficiency virus type 1 Tat protein promotes aggregation and adhesion of cerebellar neurons.

Recombinant human immunodeficiency virus (HIV-1) Tat protein added to the culture medium of rat cerebellar neurons promoted aggregation and formation of spoke-like neurites in a dose-dependent manner. Tat proteins containing mutations in the Arg-Gly-Asp (RGD) cell adhesion motif or a deletion of the cysteine-rich domain had no effect on neuronal morphology. In contrast, a Tat protein that contained a deletion of the proline-rich domain promoted neuronal aggregation. Aggregation of neurons was inhibited by the addition of monoclonal antibodies directed against the RGD and basic domains of Tat, but not against the proline-rich domain. The same domains of Tat required to induce aggregation also mediated adhesion of neurons to Tat-coated substrates. The HIV-2 Tat protein, which lacks an RGD sequence but contains cysteine-rich and basic domains similar to HIV-1 Tat, induced aggregation and acted as a substrate for adhesion when added at higher concentrations than HIV-1 Tat. Vitronectin, fibronectin, and RGD-containing peptides did not induce morphological changes in neurons or act as substrates for adhesion. The ability of Tat to induce morphological changes and promote adhesion was independent of the ability of Tat to transactivate HIV gene expression. Our results suggest that extracellular Tat protein most likely alters neuronal morphology and mediates adhesion by acting in a manner similar to an extracellular matrix protein.