Enhancement of immunotoxin efficacy by acid-cleavable cross-linking agents utilizing diphtheria toxin and toxin mutants

We have utilized a new class of acid-cleavable protein cross-linking reagents in the construction of antibody-diphtheria toxin conjugates (Srinivaschar, K., and Neville, D. M., Jr. (1989) Biochemistry 28, 2501-2509). The potency of anti-CD5 conjugates assayed by inhibition of protein synthesis on CD5 bearing cells (Jurkat) is correlated with cross-linker hydrolytic rates. The maximum increase in potency of the cleavable conjugates over non-cleavable conventional conjugates is 50-fold and is specific for the CD5 uptake route as judged by competition with excess anti-CD5. The potency of conjugates made from diphtheria toxin and the anti-high molecular weight melanoma-associated antigen (HMW-MAA) is enhanced 3-10-fold by a cleavable cross-linker. However the potency of transferrin or anti-CD3 diphtheria toxin conjugates is only minimally enhanced (2-3-fold). Mutant diphtheria toxins, CRM103 and CRM9, previously shown to express less than 1/100 of the wild type in binding affinity were substituted into these conjugates as probes for possible intracellular toxin receptor interactions. Both mutants were equally as toxic to Jurkat target cells exhibiting 1/700 the wild-type potency. CRM9 non-cleavable conjugates were equally as potent as wild-type conjugates for transferrin and anti-CD3-mediated uptake but not for anti-CD5-mediated uptake where toxicity was reduced 60-fold over the wild-type analog. The cleavable cross-linker enhanced the toxicity of anti-CD5-CRM103 and anti-CD5-CRM9 conjugates, but potency was only 1/10 that of the analogous wild-type cleavable conjugate. These data are consistent with a model in which potentiation of toxicity of the anti-CD5 and anti-high molecular weight melanoma-associated antigen conjugates by the cleavable cross-linker occurs from an enhanced intracellular toxin-toxin receptor interaction that ultimately results in increased toxin translocation to the cytosol compartment. In contrast, these data indicate that the anti-CD3 and transferrin uptake systems do not require this interaction in agreement with previous work (Johnson, V.G., Wilson, D., Greenfield, L., and Youle, R. J. (1988) J. Biol. Chem. 263, 1295-1300).     

Membrane interactions of diphtheria toxin analyzed using in vitro synthesized mutants.

We have developed a system to study the interactions of diphtheria toxin with the cell surface using non-toxic mutant proteins synthesized in vitro. Proteins obtained by N-terminal deletions containing the whole B fragment bound strongly to cells. Deletions extending into the B fragment did not yield an autonomous binding domain. Loss of only the N-terminal 3 kd of the B fragment significantly impaired the ability to recognize the receptor. This, together with previous reports that the C-terminal end of the B fragment is required for binding, suggests that both ends of the B fragment are necessary for receptor recognition. Receptor bound diphtheria toxin undergoes a conformational change at pH less than 5.3 that results in translocation of the A fragment to the cytosol and the appearance of a B fragment-derived 25 kd polypeptide (P25) resistant to externally applied protease. Only the B fragment was required for generation of P25. N-terminal deletions of 130 amino acids or more resulted in proteins that gave rise to P25 at higher pH than full length toxin. Furthermore, a second protease-inaccessible polypeptide of 18 kd (P18) was observed.     

Ion channel and membrane translocation of diphtheria toxin

Abstract Diphtheria toxin is the best studied member of a family of bacterial protein toxins which act inside cells. To reach their cytoplasmic targets, these toxins, which include tetanus and botulinum neurotoxins and anthrax toxin, have to cross the hydrophobic membrane barrier. All of them have been shown to form ion channels across planar lipid bilayer and, in the case of diphtheria toxin, also in the plasma membrane of cells. A relation between the ion channel and the process of membrane translocation has been suggested and two different models have been put forward to account for these phenomena. The two models are discussed on the basis of the available experimental evidence and in terms of the focal points of difference, amenable to further experimental investigations.     

GPI-anchored diphtheria toxin receptor allows membrane translocation of the toxin without detectable ion channel activity.

We have investigated the role of the transmembrane and cytoplasmic domains of the diphtheria toxin (DT) receptor [heparin-binding epidermal growth factor (HB-EGF) precursor] in the intoxication pathway. Two mutants were constructed in which these domains were replaced by either a 37 amino acid sequence signalling membrane attachment via a glycosylphosphatidylinositol (GPI) anchor (DTR-GPI) or by the transmembrane and cytoplasmic domains of the human EGF receptor (DTR-EGFR). Similar amounts of DTA fragment were translocated through the plasma membrane of NIH 3T3 cells transfected with the wild-type receptor (DTR), DTR-GPI and DTR-EGFR, but translocation was about six times less efficient in the case of DTR-GPI and DTR-EGFR when taking into account the number of receptors expressed. Interestingly, DT-induced 22Na+ influx was weak in DTR-EGFR cells and not detectable in DTR-GPI cells. Whole cell patch-clamp analysis showed the DT at low pH induced depolarization and decreased input resistance in DTR cells (and to a lesser extent also in DTR-EGFR cells) but not in DTR-GPI cells. These results suggest that the transmembrane and cytoplasmic part of the receptor might be involved in channel activity and that translocation of the A fragment is independent of toxin-induced cation channel activity.     

Codominant translational mutants of Chinese hamster ovary cells selected with diphtheria toxin.

Diphtheria toxin-resistance markers in two translational mutants, CH-RE1.22c, possessing no toxin-sensitive EF-2 (class IIa), and CH-RE1.32, with 50% toxin-sensitive and 50% toxin-resistant EF-2 (class IIb), behaved codominantly in somatic cell hybrids. There was no complementation in hybrids formed between the two resistant mutants. The mutant parents and their hybrids, except those formed by fusion of CH-RE1.32 and wild-type cells, grew in the presence of toxin. To explain these results we suggest that CHO-K1 cells possess two functional copies of the gene for EF-2 and that CH-RE1.22c and CH-RE1.32 represent the homozygous (R/R) and heterozygous (R/S) states of resistance at the EF-2 gene locus. The failure of hybrids formed between CH-RE1.32 and wild-type cells to grow in toxin is a gene dosage effect. Codominant class IIa translational resistance is a selectable marker for the isolation of hybrids. It can be combined with a second, recessive, marker to provide a cell which is a "universal hybridizer" (10).     

Requirement of specific receptors for efficient translocation of diphtheria toxin A fragment across the plasma membrane

The role of specific receptors in the translocation of diphtheria toxin A fragment to the cytosol and for the insertion of the B fragment into the cell membrane was studied. To induce nonspecific binding to cells, toxin was either added at low pH, or biotinylated toxin was added at neutral pH to cells that had been treated with avidin. In both cases large amounts of diphtheria toxin became associated with the cells, but there was no increase in the toxic effect. There was also no increase in the amount of A fragment that was translocated to the cytosol, as estimated from protection against externally added Pronase E. In cells where specific binding was abolished by treatment with 12-O-tetradecanoyl-phorbol 13-acetate, trypsin, or 4,4'-diisothiocyano-2,2'-stilbenedisulfonic acid, unspecific binding did not induce intoxication or protection against protease. This was also the case in untreated L cells, which showed no specific binding of the toxin. When Vero cells with diphtheria toxin bound to specific receptors were exposed to low pH, the cells were permeabilized to K+, whereas this was not the case when the toxin was bound nonspecifically at low pH or via avidin-biotin. The data indicate that the cell-surface receptor for diphtheria toxin facilitates both insertion of the B fragment into the cell membrane and translocation of the A fragment to the cytosol.     

Insertion of diphtheria toxin B-fragment into the plasma membrane at low pH. Characterization and topology of inserted regions

When the enzymatically active A-fragment of diphtheria toxin is translocated to the cytosol, the B-fragment inserts into the membrane in such a way that a 25-kDa polypeptide becomes shielded from proteases added to the external medium. We have attempted to determine the boundaries of this polypeptide within the toxin B-fragment as well as the topology of the B-fragment in the membrane. Chemical cleavage of the 25-kDa polypeptide with hydroxylamine and o-iodosobenzoic acid yielded fragments of sizes indicating that the 25-kDa polypeptide starts at residue approximately 300 and extends to the COOH-terminal end. Experiments where the toxin was labeled with [35S]cysteine at distinct positions of the B-fragment supported this conclusion. Treatment of cells with inserted B-fragment with L-1-tosyl-amido-2-phenylethyl chloromethyl ketone-treated trypsin and with V8 protease from Staphylococcus aureus yielded protected 27- and 30-kDa fragments in addition to 25 kDa, indicating that the region 240-264 is also at the outside. The topology of the inserted B-fragment is discussed.     

Crystallization of diphtheria toxin.

Two new crystal forms (forms III and IV) have been grown of diphtheria toxin (DT), which kills susceptible cells by catalyzing the ADP-ribosylation of elongation factor 2, thereby stopping protein synthesis. Forms III and IV diffract to 2.3 A and 2.7 A resolution, respectively. Both forms belong to space group C2; the unit cell parameters for form III are a = 107.3 A, b = 91.7 A, c = 66.3 A and beta = 94.7 degrees and those for form IV are a = 108.3 A, b = 92.3 A, c = 66.1 A and beta = 90.4 degrees. Both forms have one protein chain per asymmetric unit with the dimeric molecule on a twofold axis of symmetry. Form IV is exceptional among all crystal forms of DT in that it can be grown reproducibly. Thus the form IV crystals should yield a crystallographic structure giving insight into the catalytic, receptor-binding and membrane-insertion properties of DT.     

Roles of Glu 349 and Asp 352 in membrane insertion and translocation by diphtheria toxin.

Acidic conditions within the endosomal lumen induce the T domain of receptor-bound diphtheria toxin (DT) to insert into the endosomal membrane and mediate translocation of the toxin's catalytic domain to the cytosol. A conformational rearrangement in the toxin occurring near pH5 allows a buried apolar helical hairpin of the native T domain (helices TH8 and TH9) to undergo membrane insertion. If the inserted hairpin spans the bilayer, as hypothesized, then the two acidic residues within the TL5 interhelical loop, Glu 349 and Asp 352, should become exposed at the neutral cytosolic face of the membrane and reionize. To investigate the roles of these residues in toxin action, we characterized mutant toxins in which one or both acidic residues had been replaced with nonionizable ones. Each of two double mutants examined showed a several-fold reduction in cytotoxicity in 24-h Vero cell assays (sixfold for E349A + D352A and fourfold for E349Q + D352N), whereas the individual E349Q and D352N mutations caused smaller reductions in toxicity. The single and double mutations also attenuated the toxin's ability to permeabilize Vero cells to Rb+ at low pH and decreased channel formation by the toxin in artificial planar bilayers. Neither of the double mutations affected the pH-dependence profile of the toxin's conformational rearrangement in solution, as measured by binding of the hydrophobic fluorophore, 2-p-toluidinyl-naphthalene 6-sulfonate. The results demonstrate that, although there is no absolute requirement for an acidic residue within the TL5 loop for toxicity, Glu 349 and Asp 352 do significantly enhance the biological activity of the protein. The data are consistent with a model in which ionization of these residues at the cytosolic face of the endosomal membrane stabilizes the TH8/TH9 hairpin in a transmembrane configuration, thereby facilitating channel formation and translocation of the toxin's catalytic chain.     

On the membrane translocation of diphtheria toxin: at low pH the toxin induces ion channels on cells.

Diphtheria toxin (DT) in acidic media forms ion-conducting channels across the plasma membrane and inhibits protein synthesis of both highly and poorly DT-sensitive cell lines. This results in loss of cell potassium and in entry of both sodium and protons with a concomitant rapid lowering of membrane potential. The pH dependency of the permeability changes is similar to that of the inhibition of cell protein synthesis. DT-induced ion channels close when the pH of the external medium is returned to neutrality and cells recover their normal monovalent cation content. Similar permeability changes were induced by two DT mutants defective either in enzymatic activity or in cell binding, but not with a mutant defective in membrane translocation. The implication of these findings for the mechanism of DT membrane translocation is discussed.     

Conformation and model membrane interactions of diphtheria toxin fragment A

Low pH is believed to play a critical role in the penetration of membranes by diphtheria toxin in vivo. In this report, the pH dependence of the conformation of fragment A of diphtheria toxin has been studied using fluorescence techniques. As pH is decreased, fragment A in solution undergoes a reversible conformational change beginning below pH 5. The conformational change occurs rapidly upon exposure to low pH. It involves both an increase in the exposure of tryptophanyl residues to solution and a switch from a hydrophilic state to a hydrophobic state as judged by fragment A binding to micelles of a mild detergent (Brij 96). At low pH fragment A also rapidly and tightly binds to and penetrates model membranes. Binding is reversed when pH is neutralized. The transition pH, the apparent midpoint of the change between the hydrophilic state and the membrane-penetrating hydrophobic state, occurs at about pH 3.5 in the presence of Brij 96 micelles, pH 4 in the presence of small unilamellar vesicles (SUV) composed of zwitterionic phosphatidylcholine, and pH 5 in the presence of SUV composed of 25 mol % anionic phosphatidylglycerol and 75% phosphatidylcholine. The effects of high temperature provide an important clue as to the nature of the changes at low pH. At neutral pH and high temperature, i.e. in the thermally denatured state, a conformational change similar to that observed at low pH occurs, although fragment A does not become hydrophobic. In addition, the effects of low pH and high temperature on the stability of the native state are cumulative. This indicates that the changes in fragment A both at high temperature and at low pH involve denaturation, although there appears to be only partial unfolding under these conditions. Based on the results of this study, the role of fragment A in diphtheria toxin membrane penetration and translocation is evaluated.     

pH-Dependent membrane interactions of diphtheria toxin: A genetic approach

A genetic approach is described for exploring the mechanism by which diphtheria toxin undergoes pH-dependent membrane insertion and transfer of its enzymic A fragment into the cytoplasm of mammalian cells. The cloned toxin expressed inEscherichia coli is secreted to the periplasmic space, where it is processed normally and folds into a native structure. When bacteria synthesizing the toxin are exposed to pH 5, they die rapidly. The toxin undergoes a conformational change that is believed to allow it to be inserted into the bacterial inner membrane and form channels, which proves lethal for the cell. The membrane insertion event mimics the process by which the toxin inserts into the endosomal membrane of mammalian cells, leading to release of the enzymic A fragment into the cytoplasm. The observation of pH-dependent bacterial lethality provides the basis for a positive genetic selection method for mutant forms of the toxin that are altered in ability to undergo membrane insertion or pore formation.     

Lipid interaction of diphtheria toxin and mutants. A study with phospholipid and protein monolayers

To study the structural change of diphtheria toxin (DT) induced by low pH and its influence on the interaction with membrane lipids, protein and lipid monolayers were formed and characterized. DT at neutral and acidic pH forms stable monolayers, whose surface-pressure-increase curves allow an estimation of the apparent molecular area of 29.5 nm2/molecule at pH 7.4 (corresponding to a radius of 3.06 nm) and 34.5 nm2/molecule at pH 5.0 (corresponding to a radius of 3.32 nm). DT at pH 7.4 does not insert into phospholipid monolayers, while at pH 5.0 it penetrates into the lipid layer with a portion of apparent molecular area of 21.0 nm2/molecule (corresponding to a radius of 2.6 nm). The low-pH driven lipid interaction of the toxin is favoured by the presence of acidic phospholipids, without an apparent requirement for a particular class of negative lipids. The DT mutants crm 45 and crm 197 are capable of hydrophobic interaction already at neutral pH and cause an increase of surface pressure with a further increase upon acidification.     

Interactions between diphtheria toxin entry and anion transport in Vero cells. II. Inhibition of anion antiport by diphtheria toxin

When cells with surface-bound diphtheria toxin were exposed to pH 4.5, the toxin became shielded against lactoperoxidase-catalyzed radioiodination, indicating that the toxin was inserted into the membrane. Cells thus treated had strongly reduced ability to take up 36Cl-, 35SO4(2-), and [14C]SCN-. The reduction of chloride uptake was strongest at neutral pH, whereas that of sulfate was strongest at acidic pH. Lineweaver-Burk plots indicated that the toxin treatment reduced the Jmax but not the Km for the anions. The toxin also inhibited the NaCl-stimulated efflux of 35SO4(2-), indicating that the toxin inhibits the antiporter. No inhibition was found when toxin-treated cells were not exposed to low pH, whereas exposure to pH 4.5 for 20 s induced close to maximal inhibition. Half-maximal inhibition was obtained after exposure to pH 5.4. The concentration of diphtheria toxin required to obtain maximal inhibition (0.3 micrograms/ml) was sufficient to ensure close to maximal toxin binding to the cells. Even in ATP-depleted cells and in the absence of permeant anions, low pH induced inhibition of anion antiport in toxin-treated Vero cells. There was no measurable inhibition of anion antiport in cells with little or no ability to bind the toxin.     

Identification of diphtheria toxin receptor and a nonproteinous diphtheria toxin-binding molecule in Vero cell membrane

Two substances possessing the ability to bind to diphtheria toxin (DT) were found to be present in a membrane fraction from DT-sensitive Vero cells. One of these substances was found on the basis of its ability to bind DT and inhibit its cytotoxic effect. This inhibitory substance competitively inhibited the binding of DT to Vero cells. However this inhibitor could not bind to CRM197, the product of a missense mutation in the DT gene, and did not inhibit the binding of CRM197 to Vero cells. Moreover, similar levels of the inhibitory activity were observed in membrane fractions from DT-insensitive mouse cells, suggesting the inhibitor is not the DT receptor which is specifically present in DT-sensitive cells. The second DT-binding substance was found in the same Vero cell membrane preparation by assaying the binding of 125I-labeled CRM197. Such DT-binding activity could not be observed in membrane preparation from mouse L cells. From competition studies using labeled DT and CRM proteins, we conclude that this binding activity is due to the surface receptor for DT. Treatment of these substances with several enzymes revealed that the inhibitor was sensitive to certain RNases but resistant to proteases, whereas the DT receptor was resistant to RNase but sensitive to proteases. The receptor was solubilized and partially purified by chromatography on CM-Sepharose column. Immunoprecipitation and Western blotting analysis of the partially purified receptor revealed that a 14.5-kD protein is the DT receptor, or at least a component of it.     

Selective translocation of the A chain of diphtheria toxin across the membrane of purified endosomes.

Translocation is a necessary and rate-limiting step for diphtheria toxin (DT) cytotoxicity. We have reconstituted DT translocation in a cell-free system using endosomes purified from lymphocytes and have demonstrated this using two different probe/cell systems, which provided identical results: 125I-DT/human CEM cells and 125I-transferrin-DT/mouse BW cells. The cell-free DT translocation process was found to be dependent on the presence of the pH gradient endosome (pH 5.3)/cytosol (pH 7). Among the pH equilibrating agents, nigericin (5 microM) was found to be the most effective, inhibiting DT translocation by 88%. An optimum pH value of 7 on the cytosolic side of the membrane (pH gradient approximately 1.7) was determined. ATP per se is not required for DT translocation. 125I-DT translocation was 3-fold more active from late than from early endosomes, probably because of their slightly more acidic pH. Only the A chain of the toxin was found to escape from either 125I-DT/CEM or 125I-transferrin-DT/BW endosomes. Translocation of control endosome labels (125I-transferrin and 125I-horseradish peroxidase) was never observed. We also show that DT receptors present on resistant (mouse) cells block the translocation of the toxin and are responsible for the resistance of these cells to DT.