Artificial hybrid protein containing a toxic protein fragment and a cell membrane receptor-binding moiety in a disulfide conjugate. II. Biochemical and biologic properties of diphtheria toxin fragment A-S-S-human placental lactogen.

The biochemical and biologic properties of a purified disulfide conjugate of diphtheria toxin fragment A and human placental lactogen (toxin A-hPL) have been studied by (a) assaying the ADP-ribosyltransferase activity of the intact conjugate, (b) assaying the binding of the intact conjugate to mammary gland plasma membrane lactogenic receptors, and (c) assaying the effect of the conjugate on the rate of protein synthesis in rabbit mammary gland explants maintained in organ culture. The toxin A-hPL conjugate retains one-third of the NAD+:EF-2 ADP-ribosyltransferase activity of toxin A, and 26% of the hPL-binding activity to lactogenic receptors. Binding activity was demonstrated by radioreceptor assay and by assaying toxin A activity bound to membranes which was competitively displaced by excess hPL. Since the toxin A-hPL conjugate retained activities of its separate subunits, it could be regarded as a structural analogue of nicked diphtheria toxin with replacement of the original membrane-binding chain by another binding chain that is specific for lactogenic receptor. However, the conjugate failed to inhibit protein synthesis in organ-cultured mammary gland explants, although these were sensitive to native diphtheria toxin and could bind hPL. It is concluded from these results that the toxin A-hPL conjugate does not act as a functional analogue of diphtheria toxin with altered receptor specificity, and that the hPL receptor cannot mediate the entry of toxin A or toxin A-hPL from membrane-bound conjugate into the cytosol site of action of toxin A.     

Synthesis and in vitro activity of a hormone-diphtheria toxin fragment A hybrid

The synthesis and in vitro biological activity of a hybrid protein composed of intact human chorionic gonadotropin and fragment A of diphtheria toxin in a disulfide conjugate is reported. This hybrid retained greater than 90% of the binding ability of uncoupled hCG and was shown to be specifically toxic to a mouse Leydig cell tumor which binds hCG while being non-toxic towards cells which lack receptors for hCG.     

Synthesis of a cytotoxic insulin cross-linked to diphtheria toxin fragment A capable of recognizing insulin receptors

Insulin has been cross-linked via a disulfide bond to the diphtheria toxin fragment A which is catalytically active in ADP-ribosylating elongation factor-2 but does not retain binding sites for toxin receptors. The purified conjugate proved to be cytotoxic to mouse Swiss/3T3 cells which are toxin resistant but express insulin receptors. This cytotoxicity coincided with a decrease in protein synthesis and with drastic morphology changes. In contrast, IN-2 cells, which are insulin-nonresponsive variants derived from mouse $\text{BALBc}\text{3T3}$ cells, were resistant to the conjugate. Thus, the conjugate (a chimeric insulin) appears to mediate entry of the toxic fragment A into 3T3 cells through insulin receptors.     

Studies of the mechanism of cell intoxication by diphtheria toxin fragment A-asialoorosomucoid hybrid toxins. Evidence for utilization of an alternative receptor-mediated transport pathway

In order to study the variables important for determining the cytotoxicity of protein-protein hybrid toxins, the disulfide and thioiether conjugates between diphtheria toxin fragment A (DTA) and asialoorosomucoid (ASOR) were synthesized, purified, and tested for cytotoxicity toward isolated rat hepatocytes as monitored by the inhibition of protein synthesis. After 4-h incubation at 37 degrees C, both DTA-ASOR conjugates were nontoxic, but were highly toxic in the presence of an inhibitor of 125I-ASOR degradation. At concentrations inhibiting 125I-ASOR degradation to the same extent, the efficiency of inhibitors to enhance the toxicity of both conjugates varied with the following rank order: colchicine greater than chloroquine greater than NH4Cl greater leupeptin greater than cytochalasin B. In the absence of an inhibitor, both conjugates, after 10-h incubation, also exhibited measurable toxicity at concentrations as low as 10(-9) M, and DTA-S-S-ASOR was 60 times more toxic than DTA-S-ASOR. In a physiological salt solution containing energy source and bovine serum albumin, both conjugates also exhibited partial inhibition of protein synthesis after 4-h incubation and were enhanced by 2 microM colchicine toward complete inhibition with similar C0.5 of 2 X 10(-10) M. The toxicity of the conjugates were prevented in the presence of excess ASOR. These results not only indicate that intracellular degradation is an important factor restricting the toxicity of the conjugates, but also indicate that the toxic entry of DTA activity from these hybrids is independent of lysosomal degradation. Consequently, it appears that conjugate toxicity is dependent on the release of DTA activity from an extralysosomal compartment whose pool size is increased by the inhibitors and the change of medium. The effects of the inhibitors to inhibit 125I-ASOR uptake and degradation were also studied to facilitate interpretation of their enhancement of conjugate toxicity. Specifically, colchicine inhibited both internalization and degradation of the ligand in a saturable fashion with a C0.5 of 0.5 microM and a maximum inhibition of 65%. The data of the present study are interpreted so as to support a hypothesis that two receptor-mediated pathways are involved in the internalization and transport to lysosomes of ASOR and DTA-ASOR conjugates, only one of which is colchicine-sensitive.     

Receptor-mediated transport of the hybrid protein ricin-diphtheria toxin fragment A with subsequent ADP-ribosylation of intracellular elongation factor II.

A hybrid protein of ricin and the enzymatically active fragment A of diphtheria toxin (toxin A) has been synthesized and purified. The diphtheria toxin A fragment of the hybrid protein is shown to enter the cytosol compartment of HeLa cells, its presence assayed by the fall of intracellular elongation factor II (EF-2) and the rise of ADP-ribosylated EF-2. Hybrid entrance to HeLa cells is blocked by lactose which blocks receptor-mediated entry of ricin but not by NH4Cl which blocks the transport of diphtheria toxin. It is concluded that the diphtheria toxin fragment A moiety of the hybrid enters the cell cytosol via the ricin receptor-mediated transport system. The kinetics of intracellular ADP-ribosylation of EF-2 by diphtheria toxin have also been studied. Ribosylation is preceded by a toxin dose-dependent lag period. The data suggest that the time constant responsible for the lag period is in the transport step. Models consistent with these data are discussed.     

A fragment comprising the last third of bovine serum albumin which accounts for almost all the antigenic reactivity of the native protein.

The fragmentation of native bovine serum albumin by trypsin has been studied in aqueous solution under various conditions with regard to the yield and size of the fragments obtained. From a partial tryptic hydrolysate at pH 8.2 (40 degrees, 1 hour), a homogeneous fragment was isolated in high yield by gel filtration on Sephadex G-100, followed by chromatography on DEAE-cellulose. The molecular weight of the fragment by gel filtration on calibrated Sephadex G-100 columns and by sodium dodecyl sulfate electrophoresis was 22,500. After reduction of the disulfide bonds followed by alkylation of the resultant thiol groups with iodoacetamide, the fragment retained homogeneity by disc electrophoresis and its molecular weight remained unchanged, indicating that it was composed of a single polypeptide chain. From its amino acid composition, sequence of the first 20 residues, and actions of carboxypeptidases A or B, it was unequivocally assigned to positions 377-571 in albumin. The inhibitory activity of the fragment was 90 to 93% towards the immune reaction of the protein with the IgG fraction of the antisera. The IgGfraction accounted for 96% of the total antibody activity in the antisera. An immunoabsorbent of fragment 377-571 removed 89 to 95% of the antibody to albumin. A fluorescent derivative of the fragment, which retained full immunochemical activity, was found to bind 2 mol of antibody/mol of peptide. The disulfides in peptide 377-571 were essential for its immunochemical reaction because the latter was entirely abolished upon reduction and S-alkylation of the disulfides. Since this fragment comprised only a third of the albumin molecule, but accounted for 90 to 95% of its antigenic reactivity, the results indicated that native albumin carries identical repeating antigenic reactive sites.     

The role of binding ligand in toxic hybrid proteins: a comparison of EGF-ricin, EGF-ricin A-chain, and ricin

To analyze the influence of ricin B-chain on the toxicity of hybrid-protein conjugates, the rate of cellular uptake of conjugates, and the rate at which ricin A-chain (RTA) is delivered to the cytoplasm, we have constructed toxic hybrid proteins consisting of epidermal growth factor (EGF) coupled in disulfide linkage either to ricin or to RTA. EGF-ricin is no more toxic on A431 cells than EGF-RTA. The two conjugates demonstrate similar kinetics of cellular uptake (defined as antibody irreversible toxicity). EGF-RTA and EGF-ricin, like ricin, required a 2-2 1/2 hour period at 37 degrees before the onset of protein synthesis inhibition occurred. Our results suggest that RTA determines the processes which carry it, either in conjugate or toxin, from the plasma membrane binding site to the cytoplasm following endocytosis, and the ricin B chain is not required for these processes.     

Species variability in plasma S-sulfonate levels during and following sulfite administration.

It has been shown that S-sulfonate compounds (R-S-SO-3) are produced by the action of sulfite on reactive disulfide bonds [4,5]. Plasma S-sulfonate production was determined as a function of sulfite ingestion and intraperitoneal injection in rats, mice and rhesus monkeys. The tendency of these species and of the rabbit [8] to produce S-sulfonates in plasma was related to the availability of sulfite and of reactive disulfide bonds and to the stability of plasma protein S-sulfonates. The rhesus monkey and the rabbit accumulated plasma S-sulfonates much more readily than did the rat, while the mouse produced little, if any, under the same test conditions. Plasma protein S-sulfonate fractions in the rat and rhesus monkey were characterized by half-lives of approximately 4 and 8 days respectively. The sensitivity and precision of the analytical method for plasma protein S-sulfonate were improved by incorporation of 35S into the outer sulfur atom of the S-sulfonate moiety (R-S-35SO-3).     

Radioiodinated iodobenzoyl conjugates of a monoclonal antibody Fab fragment. In vivo comparisons with chloramine-T-labeled Fab

A comparative investigation of the biodistributions of radioiodinated p- and m-iodobenzoyl conjugates of a monoclonal antibody Fab fragment, NR-LU-10 Fab, and the same antibody Fab fragment radioiodinated by the chloramine-T (ChT) method has been carried out in mice. Coinjected, dual-isotope studies in athymic mice with tumor xenografts have demonstrated that there are only minor differences in the in vivo distributions of the iodobenzoyl-labeled Fabs, except in the excretory organs, kidneys, and intestines, where major differences were observed. Similarly, coinjection of either the p-iodobenzoyl or m-iodobenzoyl conjugate of NR-LU-10 Fab with the Fab radioiodinated with ChT/radioiodide into BALB/c mice provided additional data that indicated that the two iodobenzoyl conjugates distributed similar in a number of selected tissues. The tissue-distribution differences of the regioisomeric iodobenzoyl conjugates in relation to the ChT-radioiodinated Fab were large for the stomach and neck, consistent with previous studies. The most notable difference between the two iodobenzoyl conjugates was the kidney activity, where the m-iodobenzoyl conjugate was similar to the directly labeled Fab, but the p-iodobenzoyl-conjugated Fab was higher by nearly a factor of 2.     

Formation of the intrachain disulfide bond in the constant fragment of the immunoglobulin light chain

Regeneration by glutathione of the constant fragment of the immunoglobulin light chain was studied in the absence and presence of 8 m-urea. The species that appeared during the reaction of the reduced constant fragment with oxidized glutathione were trapped by alkylation with iodoacetamide and identified by electrophoresis in 15% polyacrylamide gel at pH 9.5. The kinetics of the reactions starting from various species formed during the reaction of the reduced constant fragment were also studied, and the overall reaction kinetics of the formation of the intrachain disulfide bond in the constant fragment were established in the absence and presence of urea.The reaction of the reduced constant fragment with oxidized glutathione was much slower but the yield of the constant fragment with the disulfide bond was much higher in the absence than in the presence of 8 m-urea. The slowness of the reaction in the absence of urea is due to the two cysteinyl residues of the reduced constant fragment being buried in the interior of the molecule and because oxidized glutathione is capable of reacting with the thiols only in the opened form of the protein molecule. The high yield is due to the cysteinyl thiol and the mixed disulfide in the intermediate forming an intrachain disulfide bond through thiol-disulfide interchange, the reaction sites being exposed to solvent and located at the appropriate proximity. These findings indicate first, that the appropriate proximity of a pair of cysteinyl residues is essential to form a disulfide bond and second, that they are not easily oxidized to disulfide if they are buried in the interior of the protein molecule.     

Binding properties of diphtheria toxin to cells are altered by mutation in the fragment A domain

CRM197, CRM176, and CRM228 are products of single or multiple missense mutations in the diphtheria toxin gene. CRM197 differs from wild-type toxin in 1 amino acid residue of the fragment A region, and also CRM176 and CRM228 have amino acid substitution(s) in fragment A. We compared the binding properties of CRM197 to toxin-sensitive Vero cells with those of diphtheria toxin and other CRMs. Nicked CRM197 is about 50 times more effective than intact CRM197 in inhibiting the action of diphtheria toxin on sensitive cells, as shown by inhibition of diphtheria toxin cytotoxicity or inhibition of binding of 125I-diphtheria toxin. The binding of native toxin or other CRMs was not significantly affected by nicking. Moreover, the binding of CRM197 to cells was unaffected by ATP, although ATP clearly inhibits binding of diphtheria toxin, CRM176, and CRM228. Two kinds of hybrid protein were formed using fragment B of CRM197: one with fragment A of diphtheria toxin and one with fragment A of CRM228. ATP inhibited the binding of these hybrid proteins. Furthermore, the affinities of these hybrid proteins for diphtheria toxin-sensitive cells were the same as that of native toxin. Thus, it was concluded that the altered binding properties of CRM197 were due to alteration of fragment A and what the interaction of diphtheria toxin with ATP involves both fragments. The results also suggest that fragment A plays a role in diphtheria toxin-receptor interaction.     

Monomer-dimer equilibria of a Bence Jones protein and its variable fragment.

The circular dichroic (CD) spectra of a type lambda Bence Jones protein (Tod), its variable (VL) fragment, and the constant (CL) fragment of a type lambda protein (Nag) were measured under various conditions. In the pH region from 5.5 to 7.5, the CD spectra of Tod protein with intact interchain disulfide bond (L(SS)) and and CL did not change with pH, while the spectra of Tod protein in which the interchain disulfide bond had been reduced and alkylated (L(RA)) and VL did not change with pH. The dimerization reactions of L(RA) and VL were studied by following the CD change with protein concentration. The CD spectrum of CL did not change with the protein concentration. The dimerization constant for L(RA) was 4 X 10(4) M-1 at at pH 7.5 and 25 degrees C, which was smaller than that for VL (1 X 10(5) M-1). The ellipticity at 278 nm for the L(RA) dimer was different from that for the L(SS) dimer and changed with pH. These findings indicate that the L(RA) dimer and L(SS) dimer have different conformations. The differences in the conformation and L-L interaction between the L(RA) dimer and L(SS) dimer are discussed on the basis of the conformations of VL and CL and the interactions between the paired domains.     

Expressed protein ligation using an N-terminal cysteine containing fragment generated in vivo from a pelB fusion protein

Advances in expressed protein ligation (EPL) methods that permit specific introduction of unique modifications into proteins have facilitated protein engineering, structure-function and protein interaction studies. An EPL-generated hybrid exchangeable apolipoprotein has been constructed from recombinant fragments of apolipoprotein E (apoE) and apolipophorin III (apoLp-III). A recombinant fusion protein comprised of human apoE N-terminal residues 1-111, a modified Saccharomyces cerevisiae intein and a chitin binding domain was subjected to 2-mercaptoethanesulfonic acid (MESNA) induced cleavage to generate apoE(1-111)-MESNA. A second fusion protein was comprised of a bacterial pelB leader peptide fused to a variant form of Galleria mellonella apoLp-III residues 1-91. The N-terminal pelB leader sequence directed the newly synthesized fusion protein to the Escherichia coli perisplamic space where endogenous leader peptidase cleavage generated the desired N-terminal cysteine-containing protein fragment. The resulting apoLp-III fragment, which contained no sequence tags or tails, escaped the bacteria and accumulated in the culture medium. When cultured in M9 minimal medium, Asp1Cys apoLp-III(1-91) was produced in high yield and was the sole major protein in the culture supernatant. Ligation reactions with apoE(1-111)-MESNA yielded an engineered hybrid apolipoprotein. The results document the utility of the pelB fusion protein system for generating active N-terminal cysteine containing proteins for EPL applications.     

Expression in Escherichia coli of a chemically synthesized gene for a "mini-C" analog of human proinsulin

A gene has been constructed which codes for an analog of human proinsulin in which the normal 35-amino acid connecting peptide is replaced by a "mini-C" peptide of six amino acids (Arg-Arg-Gly-Ser-Lys-Arg). The gene, composed of oligonucleotide fragments synthesized by the triester method, was cloned and expressed as a beta-galactosidase hybrid protein. The proinsulin analog was separated from beta-galactosidase by cyanogen bromide cleavage and purified. Controlled disulfide exchange in the S-sulfonate of the analog generated a molecule having high-pressure liquid chromatography (HPLC) and radioimmunoassay (RIA) behavior consistent with a proinsulin-like structure.     

Structural disulfide bonds in the Bacillus thuringiensis subsp. israelensis protein crystal.

We examined disulfide bonds in mosquito larvicidal crystals produced by Bacillus thuringiensis subsp. israelensis. Intact crystals contained 2.01 X 10(-8) mol of free sulfhydryls and 3.24 X 10(-8) mol of disulfides per mg of protein. Reduced samples of alkali-solubilized crystals resolved into several proteins, the most prominent having apparent molecular sizes of 28, 70, 135, and 140 kilodaltons (kDa). Nonreduced samples contained two new proteins of 52 and 26 kDa. When reduced, both the 52- and 26-kDa proteins were converted to 28-kDa proteins. Furthermore, both bands reacted with antiserum prepared against reduced 28-kDa protein. Approximately 50% of the crystal proteins could be solubilized without disulfide cleavage. These proteins were 70 kDa or smaller. Solubilization of the 135- and 140-kDa proteins required disulfide cleavage. Incubation of crystals at pH 12.0 for 2 h cleaved 40% of the disulfide bonds and solubilized 83% of the crystal protein. Alkali-stable disulfides were present in both the soluble and insoluble portions. The insoluble pellet contained 12 to 14 disulfides per 100 kDa of protein and was devoid of sulfhydryl groups. Alkali-solubilized proteins contained both intrachain and interchain disulfide bonds. Despite their structural significance, it is unlikely that disulfide bonds are involved in the formation or release of the larvicidal toxin.     

Processing of Pseudomonas exotoxin by a cellular protease results in the generation of a 37,000-Da toxin fragment that is translocated to the cytosol

Pseudomonas exotoxin (PE) was incubated with cells and extracts analyzed for processed fragments. PE was proteolytically cleaved to produce a N-terminal 28-kDa and a C-terminal 37-kDa fragment, the latter being composed of a portion of domain II and all of domain III (the ADP-ribosylating domain). Cleavage was evident at 10 min after toxin addition and endosome preparations contained the processed fragments. Initially, the two fragments were linked by a disulfide bond. Subsequently, the 37-kDa fragment was reduced and translocated to the cytosol where it inactivated protein synthesis. Cytosol from toxin-treated cells was greatly enriched in the 37-kDa fragment. The 37-kDa fragment appears to be essential for toxicity since mutant PE molecules that do not produce this fragment, or cannot deliver it to the cytosol, fail to kill cells.     

Interaction of S-sulfonated human IgG with human complement and its components.

S-sulfonated human IgG (S-sIgG) was prepared by treating IgG with sodium sulfite and sodium tetrathionate. The treatment resulted in the selective cleavage of interchain disulfide bonds of the IgG to give S-sulfonate groups. Complement fixing activities of aggregated S-sIgG and the immune complex formed with the S-sIgG antibody were very weak. S-sIgG at a high dose reduced the activity of the first complement component (C1) in normal human serum without any reduction of other complement components activites, but S-alkylated IgG at the same dose did not. Loss of C1 activity was not caused by either S-sulfonated myeloma proteins (IgA and IgE) or urea-treated S-sIgG, in which both inter- and intra-chain disulfide bonds were cleaved. These results suggest that the selective reduction of C1 by S-sIgG is due to a conformational change of the immunoglobulin.     

Preparation of a verifiable peptide-protein immunogen: direction-controlled conjugation of a synthetic fragment of the monitor peptide with myoglobin and application for sequence analysis

A useful method for preparing a synthetic peptide-carrying protein for specific antibody production was established. The monitor peptide is a trypsin-sensitive cholecystokinin-releasing peptide purified from rat pancreatic juice on the basis of its stimulatory activity toward pancreatic enzyme secretion. The NH2-terminus fragment of the monitor peptide (residues 1-14) was synthesized by a solid phase method. Cysteine at the COOH terminus of the fragment was conjugated with amino groups of myoglobin using a hetero-bifunctional reagent. Sequence analysis of the fragment-myoglobin conjugate indicated that the peptide/myoglobin conjugation ratio was about 1/1 (mol/mol). Antiserum against the conjugate from a rabbit effectively abolished the stimulatory activity of the monitor peptide in the rat small intestine.     

Preparation of protein conjugates via intermolecular disulfide bond formation

Conjugates of two unlike proteins can be prepared via the intermolecular disulfide interchange reaction, namely, protein A containing thiol groups reacts with protein B containing 4-dithiopyridyl groups to yield a conjugate with the release of 4-thiopyridone. Thiol groups can be introduced into proteins upon amidination with methyl 3-mercaptopropionimidate ester or 2-iminothiolane, and 4-dithiopyridyl groups can be introduced into proteins with these same reagents in the presence of 4,4'-dithiodipyridine. 2-Iminothiolane is stable on storage in contrast to the known lability of imidate esters; therefore 2-iminothiolane is a more convenient reagent for the modification of protein than are the imidate esters. All the reactions can be carried out easily under mild conditions in good yields. Conjugates of bovine plasma albumin with itself, ribonuclease, or a copolymer of D-glutamic acid and D-lysine and of sheep antibody and horseradish peroxidase were prepared with modified proteins containing an average of 1 to 5 thiol or dithiopyridyl groups per mol. These conjugates formed mainly dimers, trimers, and tetramers. The peroxidase labeled antibody retained more than 80% of its enzymatic and antigenic binding activities.     

Expression and purification of a trivalent pertussis toxin-diphtheria toxin-tetanus toxin fusion protein in Escherichia coli

Pertussis toxoid, diphtheria toxoid, and tetanus toxoid are key components of diphtheria-tetanus-acellular pertussis vaccines. The efficacy of the vaccines is well documented, however, the vaccines are expensive partly because the antigens are derived from three different bacteria. In this study, a fusion protein (PDT) composed of the immunoprotective S1 fragment of pertussis toxin, the full-length non-toxic diphtheria toxin, and fragment C of tetanus toxin was constructed via genetic means. The correct fusion was verified by restriction endonuclease analysis and Western immunoblotting. Escherichia coli carrying the recombinant plasmid (pCoPDT) produced a 161kDa protein that was recognized by antibodies specific to the three toxins. The expression of the PDT protein was inducible by isopropyl-beta-d-thio-galactoside but the total amount of protein produced was relatively low. Attempts to improve the protein yield by expression in an E. coli strain (Rosetta-gami 2) that could alleviate rare-codon usage bias and by supplementation of the growth media with amino acids deemed to be a limiting factor in translation were not successful. The PDT protein remained in the insoluble fraction when the recombinant E. coli was grown at 37 degrees C but the protein became soluble when the bacteria were grown at 22 degrees C. The PDT protein was isolated via affinity chromatography on a NiCAM column. The protein was associated with five other proteins via disulfide bonds and non-covalent interactions. Following treatment with beta-mercaptoethanol, the PDT fusion was purified to homogeneity by preparative polyacrylamide gel electrophoresis with a yield of 45 microg/L of culture. Antisera generated against the purified PDT protein recognized the native toxins indicating that some, if not all, of the native epitopes were conserved.