Production of aggregation prone human interferon gamma and its mutant in highly soluble and biologically active form by SUMO fusion technology.

The Escherichia coli expression system is a preferable choice for production of recombinant proteins. A disadvantage of this system is the target protein aggregation in "inclusion bodies" (IBs) that further requires solubilisation and refolding, which is crucial for the properties and the yield of the final product. In order to prevent aggregation, SUMO fusion tag technology has been successfully applied for expression of eukaryotic proteins, including human interferon gamma (hIFNγ) that was reported, however, with no satisfactory biological activity. We modified this methodology for expression and purification of both the wild type hIFNγ and an extremely prone to aggregation mutant hIFNγ-K88Q, whose recovery from IBs showed to be ineffective upon numerous conditions. By expression of the N-terminal His-SUMO fusion proteins in the E. coli strain BL21(DE3)pG-KJE8, co-expressing two chaperone systems, at 24 °C a significant increase in solubility of both target proteins (1.5-fold for hIFNγ and 8-fold for K88Q) was achieved. Two-step chromatography (affinity and ion-exchange) with on-dialysis His-SUMO-tag cleavage was applied for protein purification that yielded 6.0-7.0mg/g wet biomass for both proteins with >95% purity and native N-termini. The optimised protocol led to increased yields from 5.5 times for hIFNγ up to 100 times for K88Q in comparison to their isolation from IBs. Purified hIFNγ showed preserved thermal stability and antiproliferative activity corresponding to that of the native reference sample (3 × 10(7)IU/mg). The developed methodology represents an optimised procedure that can be successfully applied for large scale expression and purification of aggregation-prone proteins in soluble native form.

Characterization of functional domains of human interferon gamma by specific monoclonal antibodies.

For the study of functional domains of hIFNgamma, two MABs were characterized. Both MABs, named 1E11 and 5D6, were sequence-specific for hIFNgamma. According to the estimated additivity index, they recognized epitopes located at distinct non-overlapping areas of the hIFNgamma molecule. When pre-incubated with hIFNgamma, MAB 1E11 was able to neutralize the antiviral as well as the antiproliferative activity of the cytokine. This indicated that MAB 1E11 was specific either for the domain responsible for binding to the cell receptor or for a domain required for both functions. By contrast, the second MAB 5D6 did not interfere with any of the two hIFNgamma biological activities. The epitope of MAB 5D6 was located between the amino acid residues Leu 135 and Glu 143 by using different forms of C-terminally truncated hIFNgamma. These data allow the conclusion that the last nine C-terminal amino acids are not essential for the receptor binding and biological functioning of this cytokine. The possible role of hIFNgamma C-terminus in the intracellular cascade of events is discussed.

Role of the C-terminal chain in human interferongamma stability: an electrostatic study.

Electrostatic interactions in two structures of human interferon gamma (hIFNgamma), corresponding to interferon molecule alone and bound to its receptor, were analyzed on the basis of a continuum dielectric model. It was found that a number of titratable groups, mainly basic, show large pK shifts and remain in their neutral forms at physiologically relevant pH. The fact that these groups are largely common to both structures and that most of them belong to the set of most conserved sites suggests that this is a property inherent to the hIFNgamma molecule rather than an artifact of the crystal packing. His111 was also found deprotonated at neutral pH. It was concluded that receptor recognition involving His111 is driven by aromatic coupling of His111 and Tyr52 from the receptor rather than by electrostatic interactions. The structure corresponding to hIFNgamma in complex with its receptor shows a reduction in number and in degree of desolvation of the buried titratable sites. This finding suggested that on receptor binding, hIFNgamma adopts energetically more favorable, relaxed, conformation. It was experimentally shown that in contrast to the full-size hIFNgamma, the construct having 21 amino acid residues deleted from the C-terminus is soluble. The hydrophobicity profile analysis suggested that factors other than the exposure of hydrophobic parts of the molecule are responsible for the low stability and propensity for aggregation. On the basis of these results, it was assumed that the electrostatic influence of the C-terminal part contributes particularly to the low solvent exposure of the titratable groups, and hence to the low structural stability and propensity for aggregation of the recombinant hIFNgamma. Proteins 2001;43:125-133.

N-terminal methionine in recombinant proteins expressed in two different Escherichia coli strains.

Two genes coding for chloramphenicol acetyltransferase and human interferon gamma, respectively, were overexpressed constitutively in two different strains of Escherichia coli (E. coli LE392 and E. coli XL1). The N-terminal amino acid analysis of the purified proteins showed that: (a) the N-terminal methionine is processed more efficiently in E. coli LE392 rather than in E. coli XL1 cells; (b) the N-terminal methionine is removed better from the heterologous human interferon gamma in comparison with the homologous chloramphenicol acetyltransferase protein: and (c) there is no strong correlation between the efficiency of N-terminal procession and the yield of recombinant protein.

Identification of two RNA-binding proteins in Balbiani ring premessenger ribonucleoprotein granules and presence of these proteins in specific subsets of heterogeneous nuclear ribonucleoprotein particles.

Balbiani ring (BR) granules are premessenger ribonucleoprotein particles (RNPs) generated in giant chromosomal puffs, the BRs, in the larval salivary glands of the dipteran chironomus tentans. Monoclonal antibodies were raised against nuclear proteins collected on a single-stranded-DNA-agarose affinity column, and two of them were used to identify RNA-binding proteins in BR granules. First, in Western blots (immunoblots), one of the antibodies recognized a 36-kDa protein and the other recognized a 45-KDa protein. Second, both antibodies bound to the BRs in immunocytological experiments. It was shown in cross-linking experiments that the two proteins are associated with heterogeneous nuclear RNP (hnRNP) complexes extracted from C. tentans nuclei. By immunoelectron microscopy of isolated and partly unfolded BR RNPs, it was specifically demonstrated that the BR granules contain the two proteins and, in addition, that both proteins are distributed frequently along the RNP fiber of the particles. Thus, the 36- and 45-KDa proteins are likely to be abundant, RNA-binding proteins in the BR particles. To elucidate to what extent the two proteins are also present in other hnRNPs, we studied the binding of the antibodies to chromosomal puffs in general. It was observed that many puffs in addition to the BRs harbor the two proteins, but there are also puffs containing only one of the components, either the 36- or the 45-kDa protein. We conclude that the two proteins are not randomly bound to all hnRNPs but that each of them seems to be linked to a specific subset of the particles.

Change in the pattern of histone binding to DNA upon transcriptional activation.

Patterns of histone binding to DNA of transcriptionally active D. melanogaster hsp70 genes within the nuclei have been analyzed by two methods of histone-DNA chemical cross-linking. When cross-linking is restricted to the central, "globular" regions of histones, it drops most for H1, to an intermediate extent for H2A and H2B, and least for H3 and H4 in transcriptionally active versus transcriptionally silent chromatin. When it occurs via histone terminal regions as well, cross-linking is quantitatively similar for active and inactive chromatin. Neither cross-linking method detects histones on the hsp70 promoter region. It appears that chromatin activation decreases histone binding to DNA via the "globular" regions, known to be essential for the folding of nucleosomes and the 30 nm chromatin fibril, but does not significantly affect the interaction of flexible and loosely bound histone "tails" with DNA. The role of these histone-DNA interaction changes in the unfolding of active chromatin and RNA polymerase reading through histone-bound DNA is discussed.

[Structural changes in chromatin of heat shock protein 70 gene of Drosophila during transcription]. / Strukturnye izmeneniia khromatina gena belka teplovogo shoka 70 Drosofily pri transkriptsii.

Using "protein-image" hybridization technique combined with various crosslinking methods, for formaldehyde-prefixed nuclei we have analysed changes induced by activation in the chromatin structure of HSP-70 genes. From the crosslinking data it follows that chromatin of actively transcribed genes undergoes some structural rearrangements resulting in certain weakening of the contacts between DNA and the globular parts of histones so that the histones remain bound to DNA through their N-terminal regions. In addition, there have been found two specific regions with a reduced content of histones: the 5'-promoter of HSP-70 gene and a region distanced by approximately 1 k.b. from the 3'-end of the HSP-70 gene.