Conformations of fibroblast and E. coli-derived recombinant human interferon-beta s as studied by nuclear magnetic resonance and circular dichroism

The conformations of fibroblast and E. coli-derived recombinant human interferon-beta s were studied by circular dichroism and nuclear magnetic resonance spectroscopy in the acidic pH region of 4.6 to 1.6. Both interferons have very similar conformations with high alpha-helix contents (approximately 70%). These results suggest that glycosylation does not appreciably change the conformation of human interferon-beta. Moreover, a slow conformational change is observed below pH 2.0, which induces the disruption of beta-sheets.     

Disulfide bond interchange in Escherichia coli-derived recombinant human interferon-beta 1 under denaturing conditions

Disulfide bond interchange has been pointed out as a considerable problem in preparing recombinant proteins from Escherichia coli cells. This has been reported in the system of reducing denaturation followed by a refolding process, where incorrectly folded molecules are sometimes produced. As the possibility of disulfide bond interchange may also arise in the cytoplasm of E. coli cells, the state of sulfhydryl groups of recombinant proteins obtained from a nonreducing and nondenaturing process should be examined. The state of sulfhydryl groups of E. coli-derived recombinant human interferon-beta 1, which had been purified under nonreducing and nondenaturing conditions, was examined by using the N-(7-dimethylamino-4-methylcoumarinyl)maleimide (DACM) labeling technique. Among the three cysteine residues in E. coli-derived human interferon-beta 1, the 17th cysteine was identified as being unpaired, as in the natural molecule. However, it was found that three isomers of the recombinant protein could be formed when the protein was denatured with 6 M guanidine hydrochloride. These three isomers were identified as having unpaired cysteine residues at positions 17, 31, and 141, respectively. These results indicate that disulfide bond interchange occurs in E. coli-derived recombinant human interferon-beta 1 under denaturing conditions in spite of the absence of a reducing agent.     

Characterization of four different mammalian-cell-derived recombinant human interferon-beta 1s. Identical polypeptides and non-identical carbohydrate moieties compared to natural ones

In order to evaluate the structural identification of various recombinant human interferon-beta 1s, the recombinant proteins were produced in four different mammalian cells (human PC12 and PC8 lung adenocarcinoma cells, Chinese hamster ovary cells and mouse C127 cells) and characterized. Each mammalian-cell-derived recombinant human interferon-beta 1 represented a single band of 23 kDa on sodium dodecyl sulphate/polyacrylamide gel electrophoresis, the same molecular mass as fibroblast-derived natural human interferon-beta 1. Specific activities, amino acid compositions, amino-terminal sequences, peptide maps on C18 reversed-phase high-performance liquid chromatography and circular dichroic spectra of recombinant proteins were in good agreement with natural ones. On the other hand, the patterns of isoelectric focusing were different between mammalian-cell-derived recombinant human interferon-beta 1s and natural human interferon-beta 1. Sugar composition analysis revealed that the recombinant protein from Chinese hamster ovary cells has a similar sugar composition to that of natural protein and the other recombinant proteins have increased amounts of galactose and glucosamine in comparison to the natural protein. Furthermore, there is no galactosamine in the natural protein, while small amounts of galactosamine were detected in the oligosaccharides released from PC8- and C127-derived recombinant proteins by N-glycanase. These results indicate that mammalian-cell-derived recombinant human interferon-beta 1s have identical polypeptides to those of natural human interferon-beta 1 but their carbohydrate moieties, including unusual N-linked oligosaccharides, are individually different from natural ones and depend on the host cell.     

Renaturation of Escherichia coli-derived recombinant human macrophage colony-stimulating factor.

Recombinant human macrophage colony-stimulating factor (rhM-CSF), a homodimeric, disulfide bonded protein, was expressed in Escherichia coli in the form of inclusion bodies. Reduced and denatured rhM-CSF monomers were refolded in the presence of a thiol mixture (reduced and oxidized glutathione) and a low concentration of denaturing agent (urea or guanidinium chloride). Refolding was monitored by nonreducing gel electrophoresis and recovery of bioactivity. The effects of denaturant type and concentration, protein concentration, concentration of thiol/disulfide reagents, temperature, and presence of impurities on the kinetics of rhM-CSF renaturation were investigated. Low denaturant concentrations (<0.5 M urea) and high protein concentrations (>0.4 mg/ml) in the refolding mixture resulted in increased formation of aggregates, although aggregation was never significant even when refolding was carried out at room temperature. Higher protein concentration resulted in higher rates but did not lead to increased yields, due to the formation of unwanted aggregates. Experiments conducted at room temperature resulted in slightly higher rates than those conducted at 4 degrees C. Although the initial renaturation rate for solubilized inclusion body protein without purification was higher than that of the reversed-phase purified reduced denatured rhM-CSF, the final renaturation yield was much higher for the purified material. A maximum refolding yield of 95% was obtained for the purified material at the following refolding conditions: 0.5 M urea, 50 mM Tris, 1.25 mM DTT, 2 mM GSH, 2 mM GSSG, 22 degrees C, pH 8, [protein] = 0.13 mg/ml.     

The intrinsic fluorescence of the recombinant human leukocyte interferon-alpha A and fibroblast interferon-beta 1

The environment of Trp residues of the recombinant human interferons has been studied by the analysis of the emission spectra of native and denatured proteins at pH 1.5-8.5 and temperature 10-65 degrees C in the presence and absence of the anionic, cationic and neutral to charge contact quenchers--KI, CsCl and acrylamide, respectively. The obtained data allow to suppose that in IFN-alpha A and IFN-beta 1 Trp141 interacts with Arg145 and one or several from the following residues--Phe124, Ile127, Tyr130, Leu131, whereas Trp77--with Arg33 and Phe36, Phe78, Leu81 or Leu82 (Ile81 or Val82 for IFN-beta 1).     

Evaluation of different primary recovery methods for E. coli-derived recombinant human growth hormone and compatibility with further down-stream purification

Due to advances in fermentation technology, it is now possible to obtain fermentation broth with over 30% solids. The high solid content makes the clarification step difficult, especially at large scale. The primary protein recovery step is challenging due to the heterogeneous solution of soluble and insoluble material. In this study, we compare different primary recovery routes and the compatibility with the initial capture chromatography step. The primary recovery routes studied are standard clarification by centrifugation and extraction in aqueous two-phase systems. The compatibility of the feed streams from the different primary recovery steps with the first chromatography step is addressed. An anion-exchange column was used as the first capture column in the purification process. The aqueous two-phase system was composed of a random copolymer of ethylene oxide and propylene oxide (EOPO) in combination with a waxy starch. The target protein in this study was human growth hormone (hGH) produced in recombinant Escherichia coli. The purity of hGH in the top phase after aqueous two-phase extraction was found to be significantly higher than in clarified homogenate supernatant and increased as the EOPO polymer concentration in the aqueous two-phase system increased. Stability of the supernatant and EOPO top phases and hGH were determined by turbidity measurements and LC-MS assay. All of the feed-streams from the primary recovery steps were compatible with the anion-exchange chromatography step; however, the capacity of the resin was strongly dependent on the purity of the load. Different process aspects, e.g., resin capacity, viscosity, purification, and yield of hGH and scalability are compared.     

重组人干扰素—β在大肠杆菌中的表达与活性分析

利用定点突变技术将人干扰素-β第17位半胱氨酸编码序列突变为丝氨酸(IFN-βser17),DNA序列分析证明了其核苷酸序列的正确性,并在大肠杆菌TAP106中获得高效表达,IFN-βser17表达量达20%,用Westemblot得到单一条带,细胞病变抑制法测定其比活性达(2～3)×107U/mg.INF-βser17的比活性和稳定性均超过天然人IFN-β.     

Molecular characterization of recombinant human acidic fibroblast growth factor produced in E. coli: comparative studies with human basic fibroblast growth factor.

Synthetic cDNA coding for human acidic fibroblast growth factor (haFGF) was expressed in E. coli under the control of the T7 promoter. The haFGF produced was purified extensively using heparin-Sepharose and phenyl-Sepharose columns. The mitogenic activity of haFGF on 3T3 and endothelial cells was significantly potentiated in the presence of heparin (10-50 micrograms/ml), while angiogenic activity was observed on chick embryo chorioallantoic membrane without exogenously added heparin. This significant potentiation of mitogenic activity was observed specifically with haFGF, not human basic fibroblast growth factor (hbFGF). Circular dichroism spectra of haFGF was not affected by the presence of heparin. The affinity of haFGF for heparin was examined using heparin affinity HPLC and was precisely confirmed to be relatively lower than that of hbFGF. These results implied that haFGF was potentiated by heparin and that this potentiation did not involve a significant change in the conformation of the haFGF molecule. The affinity of haFGF for copper was also confirmed to be higher than that of hbFGF using a copper affinity HPLC column. In addition, under acidic conditions, haFGF appeared more stable than hbFGF and was further stabilized in the presence of heparin.     

Characterization of a cysteine-free analog of recombinant human basic fibroblast growth factor

Using oligo site-directed mutagenesis, we have modified our synthetic gene for human basic fibroblast growth factor (bFGF) to replace all four cysteine codons with serine codons. The corresponding protein was expressed in Escherichia coli and purified from inclusion bodies by solubilization in urea followed by a series of column chromatographies and a folding step. The resulting protein, having no cysteine residues, is unable to form either intramolecular or intermolecular disulfide bonds. The secondary and tertiary structures of the purified analog, as determined by circular dichroism and fluorescence spectroscopy, were identical within experimental error to recombinant bovine and human bFGF with unaltered amino acid sequences. Reflecting the similar conformation, the analog protein exhibited mitogenic activity on NIH 3T3 cells which was indistinguishable from the natural sequence molecule.     

Simple purification ofEscherichia coli-derived recombinant human interleukin-2 expressed with N-terminus fusion of glucagon

Simple procedures have been devised for purifying recombinant human interleukin-2 (hIL-2), which was expressed inEscherichia coli using sequences of glucagon molecules and enterokinase cleavage site as an N-terminus fusion partner. The insoluble aggregates of recombinant fusion protein produced inE. coli cytoplasm were easily dissolved by simple alkaline pH shift (8→12→8). Following enterokinase cleavage, the recombinant hIL-2 was finally purified by one-step reversed-phase HPLC with high purity. The ease and high efficiency of this simple purification process seem to mainly result from the role of used glucagon fusion partner, which could be applied to the production of other therapeutically important proteins.     

Three-dimensional crystal structure of recombinant murine interferon-beta.

The crystal structure of recombinant murine interferon-beta (IFN-beta) has been solved by the multiple isomorphous replacement method and refined to an R-factor of 20.5% against 2.6 A X-ray diffraction data. The structure shows a variant of the alpha-helix bundle with a new chain-folding topology, which seems to represent a basic structural framework of all the IFN-alpha and IFN-beta molecules belonging to the type I family. Functionally important segments of the polypeptide chain, as implied through numerous gene manipulation studies carried out so far, are spatially clustered indicating the binding site(s) to the receptor(s). Comparison of the present structure with those of other alpha-helical cytokine proteins, including porcine growth hormone, interleukin 2 and interferon gamma, indicated either a topological similarity in chain folding or a similar spatial arrangement of the alpha-helices.     

Synthesis of human fibroblast interferon by E. coli

A cDNA library was constructed using mRNA from human fibroblasts induced with poly(I):poly(C). A bacterial clone containing fibroblast interferon cDNA sequences was identified by hybridization to a cDNA probe synthesized using deoxyoligonucleotide primers which hybridize to fibroblast interferon mRNA specifically. Expression plasmids were constructed which permitted the synthesis in E. coli of 8 x 10(7) units of human fibroblast interferon per liter of culture. The bacterially produced fibroblast interferon is indistinguishable from authentic human fibroblast interferon by several criteria.     

Comparative study of the asparagine-linked sugar chains of natural human interferon-beta 1 and recombinant human interferon-beta 1 produced by three different mammalian cells

The asparagine-linked sugar chains of natural interferon-beta 1 secreted from human foreskin fibroblasts by poly I:poly C induction and of three recombinant human interferon-beta 1 produced by Chinese hamster ovary cells, mouse epithelial cells (C127), and human lung adenocarcinoma cells (PC8) were released quantitatively as oligosaccharides by hydrazinolysis followed by N-acetylation. After being reduced with either NaB3H4 or NaB2H4, their structures were comparatively analyzed. More than 80% of the sugar chains of natural interferon-beta 1 occur as biantennary complex-type sugar chains, approximately 10% of which contain N-acetyllactosamine repeating structure in their outer chain moieties. The remainders are 2,4- and 2,6-branched triantennary complex-type sugar chains. The sugar chains of the recombinant interferon-beta 1 derived from Chinese hamster ovary cells were very similar to those of its natural counterpart. In contrast, two other recombinant proteins contain quite different sugar chains. The protein derived from C127 cells contains complex-type sugar chains with the Gal alpha 1----3Gal beta 1----4GlcNAc group in their outer chain moieties. Their sialic acid residues occur solely as the Sia alpha 2----6Gal group, where Sia is sialic acid. In contrast, the sialic acid residues of other interferon-beta 1 occur as the Sia alpha 2----3Gal group only. A part of the sugar chains of the protein derived from PC8 cells contains bisecting N-acetylglucosamine residue in addition to the Gal alpha 1----3Gal beta 1----4GlcNAc group.     

Superinduction of the human interferon-beta promoter.

Superinduction, that is induction with simultaneous blocking of mRNA translation, enhances the induction of interferon-beta in response to virus or double-stranded RNA in human fibroblasts. Expression of the cloned IFN-beta gene upon transfer into heterologous cells reflects the endogeneous interferon expression with respect to induction or superinduction indicating the involvement of cellular mediators in this mode of gene regulation. Expression from gene hybrids in mouse Mo57/2 cells reveals that 5' flanking DNA sequences from the human IFN-beta gene are responsible for induction and superinduction. Superinduction of the human IFN-beta promoter is demonstrated in several rodent and primate cell lines. In addition, expression from promoter mutants in mouse cells indicates that DNA sequences responsible for induction and superinduction are identical.     

New crystal form of recombinant murine interferon-beta

Although we have reported (Matsuda, S., Kawano, G., Itoh, S., Mitsui, Y., and Iitaka, Y. (1986) J. Biol. Chem. 261, 16207-16209) that recombinant murine interferon-beta produced in Escherichia coli was crystallized in an orthorhombic space group C222(1) using polyethyleneglycol 8000 as precipitant, the crystals had an insufficient resolution and a marked tendency for orientational disorder around the c axis. We now report that another form of murine interferon-beta crystals with little disorder was obtained in the presence of dioxane using ammonium sulfate as precipitant. The new crystals belong to a hexagonal space group P6(1) or P6(5) with a = b = 71.4 A and c = 79.6 A having only one murine interferon-beta molecule in an asymmetric unit. The crystals are reasonably stable to x-rays and significantly diffract up to 2.2 A resolution when a synchrotron beam is used.     

Reactivity of E. coli-derived trans-activating protein of human T lymphotropic virus type III with sera from patients with acquired immune deficiency syndrome

Randomly sheared DNA fragments from HTLV-III proviral DNA were cloned into an E. coli open reading frame (ORF) expression vector. The inserted ORF DNA was expressed in E. coli transformants as a polypeptide fused to the lambda CI protein at the amino terminus and to beta-galactosidase at the carboxyl terminus. The reactivity of the recombinant peptides with antibodies from sera of AIDS patients was determined by the Western blot technique. The coordinates of the DNA inserts of the immunoreactive clones were then determined by DNA sequencing. A clone, ORF 628, was found to contain a short DNA segment located between the sor and env genes (nucleotide positions 5367 to 5597), a region previously thought to be noncoding. Inspection of the DNA sequences of this clone and of other HTLV-III isolates revealed the presence of a small ORF located between nucleotide position 5411 and 5625, capable of encoding a polypeptide of 72 amino acids. The biosynthesis of the polypeptide of ORF 628 initiates from an ATG codon within the HTLV-III insert. The fusion protein of ORF 628 was partially purified by affinity chromatography on CH Sepharose 4B coupled to a beta-galactosidase ligand, and tested against a panel of sera from AIDS patients by Western blot analysis. Approximately 35% of the sera from patients with AIDS or ARC contained antibodies reactive with the peptide. The DNA region spanned by ORF 628 is now thought to be the major functional element of the trans-activator gene, tat.     

Characterization of a phosphate transport system in human fibroblast lysosomes.

The uptake of [32P]KH2PO4 by Percoll-purified human fibroblast lysosomes at pH 7.0 was investigated to determine if lysosomes contain a transport system recognizing phosphate. Lysosomal phosphate uptake was linear for the first 2 min, attained a steady state by 8-10 min at 37 degrees C, and was not Na+ or K+ dependent. Upon entering lysosomes, [32P]phosphate was rapidly metabolized to trichloroacetic acid-soluble and trichloroacetic acid-insoluble products. After 1-min incubations, 50% of the radioactivity recovered from lysosomes was in the form of inorganic phosphate; and after a 2.5-min incubation, 27% of the radioactivity was recovered as inorganic phosphate. When lysosomes are loaded with radioactivity by incubation with 0.03 mM [32P]KH2PO4 for 25 min and then washed at 4 degrees C, lysosomes fail to release the accumulated radioactivity during a subsequent incubation at 37 degrees C. Lysosomal phosphate uptake gave linear Arrhenius plots (Q10 = 1.8) and was inversely proportional to medium osmolarity. Phosphate uptake was maximal at pH 5-6, half-maximal at pH 7.1, with little transport activity at pH greater than 8, suggesting that the transport system recognizes the monobasic form of phosphate. Lysosomal phosphate uptake is saturable, displaying a Km of 5 microM at pH 7.0 and 37 degrees C. High specificity for phosphate is demonstrated since large concentrations of Na2SO4, NaHCO3, KCl, NaCl, 5'-AMP, or the anion transport inhibitor, 4,4'-diisothiocyanatostilbene-2,2'-disulfonate, have no effect on lysosomal phosphate transport. In contrast, the phosphate analog, arsenate, strongly inhibits lysosomal phosphate uptake in a competitive manner with a Ki of 7 microM. Pyridoxal phosphate, CTP, adenosine 5'-(beta,gamma-imino)triphosphate (AMP-PNP), and glucose 6-phosphate were found to be noncompetitive inhibitors of lysosomal phosphate uptake displaying Ki values of 80-250 microM. When lysosomes are incubated with [gamma-32P]ATP, the lysosomal membrane ATPase hydrolyzes the ATP to form inorganic phosphate which then enters lysosomes by this lysosomal phosphate transport route.     

Expression and refolding of recombinant human fibroblast procathepsin D

Procathepsin D is a precursor of the human lysosomal protease cathepsin D. Due to its short half-life, procathepsin D is difficult to obtain in quantities sufficient to allow structural and enzymatic studies. To obtain large quantities of this precursor, procathepsin D was expressed using the T7 promoter vector pET3a in bacteria that carry a chromosomal copy of the T7 RNA polymerase gene under the control of the lac promoter. At high cell density in rich medium, basal levels of T7 RNA polymerase were sufficient to express recombinant procathepsin D without addition of an exogenous inducer of the lac promoter. The recombinant protein, constituting almost half of the total cell protein, accumulated in intracytoplasmic inclusion bodies and was isolated from the insoluble fraction of lysed cells. Antibodies prepared against the purified recombinant protein were shown to crossreact with native human placental and porcine spleen cathepsin D. Recombinant procathepsin D was solubilized in denaturants and was refolded. After extended preincubation of the denatured protein at acid pH to allow folding and activation of the zymogen, pepstatin inhibitable catalytic proteolysis was detected. These data demonstrated that the glycosylated aspartic protease, procathepsin D can be refolded and activated in an unglycosylated form and thus provides a system for the study of procathepsin D structure and function.     

Characterization of 58-kilodalton human neutrophil collagenase: comparison with human fibroblast collagenase

A series of experiments has been carried out to characterize 58-kDa human neutrophil collagenase (HNC) and compare it with human fibroblast collagenase (HFC). N-Terminal sequencing of latent and spontaneously activated HNC shows that it is a distinct collagenase that is homologous to HFC and other members of the matrix metalloproteinase gene family. Activation occurs autolytically by hydrolysis of an M-L bond at a locus homologous to the Q80-F81-V82-L83 autolytic activation site of HFC. This releases a 16-residue propeptide believed to contain the "cysteine switch" residue required for latency. Polyclonal antibody raised against HNC cross-reacts with HFC but with none of the other major human matrix metalloproteinases examined. Treatment of HNC with endoglycosidase F or N-glycosidase F indicates that it is glycosylated at multiple sites. The deglycosylated latent and spontaneously activated enzymes have molecular weights of approximately 44K and 42K, respectively. Differences in the carbohydrate processing of HFC and HNC may determine why HFC is a secreted protein while HNC is stored in intracellular granules. The kinetic parameters kcat and KM for the hydrolysis of the interstitial collagen types I, II, and III in solution by both collagenases have been determined. The strong preferences of HNC for type I collagen and of HFC for type III collagen found in earlier studies have been confirmed. The preference of HNC for type I over type III collagen is almost abolished when fibrillar collagens are used as substrates, but the preference for HFC for type III over type I collagen is only partially decreased.(ABSTRACT TRUNCATED AT 250 WORDS)