Sortase-catalyzed in vitro functionalization of a HER2-specific recombinant Fab for tumor targeting of the plant cytotoxin gelonin

We report on the preparation of a new type of immunotoxin via in vitro ligation of the αHer2 antigen binding fragment (Fab) of the clinically-validated antibody trastuzumab to the plant toxin gelonin, employing catalysis by the bacterial enzyme sortase A (SrtA). The αHer2 Fab was fused with the extended SrtA recognition motif LPET↓GLEH₆ at the C-terminus of its heavy chain, thereby preventing interference with antigen binding, while the toxin was equipped with a Gly₂ sequence at its N-terminus, distant to the catalytically active site in the C-terminal region. Site-specific in vitro transpeptidation led to a novel antibody-toxin conjugate wherein gelonin had effectively replaced the Fc region of a conventional (monomerized) immunoglobulin. After optimization of reaction conditions and incubation time, the resulting Fab-Gelonin ligation product was purified to homogeneity in a two-step procedure by means of Strep-Tactin affinity chromatography—utilizing the Strep-tag II appended to gelonin—and size exclusion chromatography. Binding activity of the immunotoxin for the Her2 ectodomain was indistinguishable from the unligated Fab as measured by real-time surface plasmon resonance spectroscopy. Specific cytotoxic potency of Fab-Gelonin was demonstrated against two Her2-positive cell lines, resulting in EC₅₀ values of ~1 nM or lower, indicating a 1000-fold enhanced cell-killing activity compared with gelonin itself. Thus, our strategy provides a convenient route to the modular construction of functional immunotoxins from Fabs of established tumor-specific antibodies with gelonin or related proteotoxins, also avoiding the elevated biosafety levels that would be mandatory for the direct biotechnological preparation of corresponding fusion proteins.     

The antileukemic efficacy of an immunotoxin composed of a monoclonal anti-Thy-1 antibody disulfide linked to the ribosome-inactivating protein gelonin

Summary We prepared an immunoconjugate consisting of a monoclonal antibody recognizing the Thy-1 antigen and the ribosome-inactivating protein gelonin linked by a disulfide bond. This immunotoxin preparation was judged to contain less than 5% free antibody or gelonin. It was highly toxic in vitro in an antigen-specific fashion to the Thy-1 expressing RADA leukemia of A/J mice. The IC₅₀ of this preparation on RADA in vitro was 10^–12 M, while the IC₅₀ on the Thy-1 negative S1509a fibrosarcoma of A/J mice was 10^–7 M. The toxicity of this immunoconjugate was also measured in a direct proliferation assay and it was found that a 4-h exposure and a 24-h exposure of RADA cells to a 1 nM concentration of immunotoxin killed 90% and 99.9% of cells, respectively. Furthermore, efficacy in vitro was not due to the intrinsic susceptibility of RADA cells to tis type of immunotoxin, as one prepared with gelonin and an antibody recognizing the TL^a determinant on this leukemia had no efficacy in vitro. Clearance of the anti-Thy-1-gelonin immunoconjugate from the circulation of A/J mice after i.v. injection was rapid, especially during the first 8 h after injection, possibly because of binding to Thy-1 expressing tissue. Delivery of immunoconjugate to ascitic tumor in vivo was substantially better if the immunoconjugate was given by i.p. injection, rather than by the i.v. route. When given either i.v. or i.p. at the time of i.p. tumor inoculation in vivo, the anti-Thy-1-gelonin immunotoxin showed potency in an antigen-specific fashion; while this immunoconjugate prolonged survival and frequently cured RADA-inoculated mice, neither anti-Thy-1 antibody, gelonin, a combination of the two, nor immunotoxin of irrelevant specificity had any significant effect on survival. Anti-Thy-1-gelonin also had no effect on survival of A/J mice inoculated i.p. with S1509a. Furthermore, it was determined that a single i.p. dose of anti-Thy-1-gelonin killed 90% to 99% cells in vivo, and that the immunoconjugate was about as effective in this model as either adriamycin or cytoxan.This work was supported by ImmunoGen Inc. and in part by a grant from the National Institutes of Health, CA-14723     

Sequence variation in the β7–β8 loop of bacterial class A sortase enzymes alters substrate selectivity

Gram-positive bacteria contain sortase enzymes on their cell surfaces that catalyze transpeptidation reactions critical for proper cellular function. In vitro, sortases are used in sortase-mediated ligation (SML) reactions for a variety of protein engineering applications. Historically, sortase A from Staphylococcus aureus (saSrtA) has been the enzyme of choice to catalyze SML reactions. However, the stringent specificity of saSrtA for the LPXTG sequence motif limits its uses. Here, we describe the impact on substrate selectivity of a structurally conserved loop with a high degree of sequence variability in all classes of sortases. We investigate the contribution of this β7–β8 loop by designing and testing chimeric sortase enzymes. Our chimeras utilize natural sequence variation of class A sortases from eight species engineered into the SrtA sequence from Streptococcus pneumoniae. While some of these chimeric enzymes mimic the activity and selectivity of the WT protein from which the loop sequence was derived (e.g., that of saSrtA), others results in chimeric Streptococcus pneumoniae SrtA enzymes that are able to accommodate a range of residues in the final position of the substrate motif (LPXTX). Using mutagenesis, structural comparisons, and sequence analyses, we identify three interactions facilitated by β7–β8 loop residues that appear to be broadly conserved or converged upon in class A sortase enzymes. These studies provide the foundation for a deeper understanding of sortase target selectivity and can expand the sortase toolbox for future SML applications.     

Quantitative analysis of protein synthesis inhibition and recovery in CRM107 immunotoxin-treated HeLa cells

Previously a mathematical model was proposed that quantitatively related protein synthesis inhibition kinetics of antitransferrin receptor-gelonin immunotoxins to the cellular trafficking of the targeting agent. That work is here extended to describe protein synthesis inhibition kinetics of immunotoxins containing the diphtheria toxin mutant CRM107. CRM107 differs from gelonin in both translocation and ribosomal inactivation mechanisms. Targeting agents used were antitransferrin monoclonal antibodies 5E9 and OKT9, OKT9Fab, and transferrin. CRM107 conjugates inhibited protein synthesis at substantially lower concentrations than gelonin conjugates; this effect was attributed to substantially higher translocation rates for CRM107. However, under certain conditions, CRM107 immunotoxin-treated cells were able to recover completely; this behavior was never observed with gelonin immunotoxins. To quantitatively capture this phenomenon, extracellular and cytosolic degradation of the toxin as well as growth-related recovery from toxin-induced damage were incorporated into the mathematical model. Translocation and cytosolic degradation rate constants were determined for each immunotoxin. Unlike the gelonin conjugates, the translocation rate of CRM107 conjugates depended on the targeting molecule. This provided indirect evidence that CRM107 remains disulfide linked to the targeting agent for at least part of the translocation process. Although the CRM107 conjugates all had higher translocation rates and inhibited protein synthesis at lower concentrations than the gelonin conjugates, the cells' ability to recover from protein synthesis inhibition at low immunotoxin concentrations limits the utility of CRM107 conjugates for targeted cell killing.     

An antibody Fc engineered for conditional antibody-dependent cellular cytotoxicity at the low tumor microenvironment pH

Despite the exquisite specificity and high affinity of antibody-based cancer therapies, treatment side effects can occur since the tumor-associated antigens targeted are also present on healthy cells. However, the low pH of the tumor microenvironment provides an opportunity to develop conditionally active antibodies with enhanced tumor specificity. Here, we engineered the human IgG1 Fc domain to enhance pH-selective binding to the receptor FcγRIIIa and subsequent antibody-dependent cellular cytotoxicity (ADCC). We displayed the Fc domain on the surface of mammalian cells and generated a site-directed library by altering Fc residues at the Fc–FcγRIIIa interface to support interactions with positively charged histidine residues. We then used a competitive staining and flow cytometric selection strategy to isolate Fc variants exhibiting reduced FcγRIIIa affinities at neutral pH, but physiological affinities at the tumor-typical pH 6.5. We demonstrate that antibodies composed of Fab arms binding the breast cell epithelial marker Her2 and the lead Fc variant, termed acid-Fc, exhibited an ∼2-fold pH-selectivity for FcγRIIIa binding based on the ratio of equilibrium dissociation constants K_d,7.4/K_d,6.5, due to a faster dissociation rate at pH 7.4. Finally, in vitro ADCC assays with human FcγRIIIa-positive natural killer and Her2-positive target cells demonstrated similar activities for anti-Her2 antibodies bearing the wild-type or acid-Fc at pH 6.5, but nearly 20-fold reduced ADCC for acid-Fc at pH 7.4, based on EC₅₀ ratios. This work shows the promise of mammalian cell display for Fc engineering and the feasibility of pH-selective Fc activation to provide a second dimension of selective tumor cell targeting.     

A potent and specific immunotoxin for tumor cells expressing disialoganglioside GD2

Summary Monoclonal antibody 14G2a (anti-GD₂) reacts with cell lines and tumor tissues of neuroectodermal origin that express disialoganglioside GD₂. mAb 14G2a was coupled to the ribosome-inactivating plant toxin gelonin with the heterobifunctional cross-linking reagentN-succinimidyl-3(2-pyridyldithio)propionate. The activity of the immunotoxin was assessed by a cell-free translation assay that confirmed the presence of active gelonin coupled to 14G2a. Data from an enzyme-linked immunosorbent assay demonstrated the specificity and immunoreactivity of the 14G2a-gelonin immunotoxin, which was identical to that of native 14G2a. Assays for complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC) revealed that these functional properties of the native 14G2a antibody were also preserved in the 14G2a-gelonin immunotoxin. The gelonin-14G2a immunotoxin was directly cytotoxic to human melanoma (A375-M and AAB-527) cells and was 1000-fold more active than native gelonin in inhibiting the growth of human melanoma cells in vitro. The augmentation of tumor cell killing of 14G2a-gelonin immunotoxin was examined with several lysosomotropic compounds. Chloroquine and monensin, when combined with 14G2a-gelonin immunotoxin, augmented its cytotoxicity more than 10-fold. Biological response modifiers such as tumor necrosis factor and interferon and chemotherapeutic agents such as cisplatinum andN,N-bis(2-chloroethyl)-N-nitrosourea (carmustine) augmented the cytotoxicity of 14G2a-gelonin 4- to 5-fold. The results of these studies suggest that 14G2a-gelonin may operate directly by both cytotoxic efforts and indirectly by mediating both ADCC and CDC activity against tumor cells; thus it may prove useful in the future for therapy of human neuroectodermal tumors.Research conducted, in part, by the Clayton Foundation for Research     

Structure of the Bacillus anthracis Sortase A Enzyme Bound to Its Sorting Signal: A FLEXIBLE AMINO-TERMINAL APPENDAGE MODULATES SUBSTRATE ACCESS*

The endospore forming bacterium Bacillus anthracis causes lethal anthrax disease in humans and animals. The ability of this pathogen to replicate within macrophages is dependent upon the display of bacterial surface proteins attached to the cell wall by the B. anthracis Sortase A (^BaSrtA) enzyme. Previously, we discovered that the class A ^BaSrtA sortase contains a unique N-terminal appendage that wraps around the body of the protein to contact the active site of the enzyme. To gain insight into its function, we determined the NMR structure of ^BaSrtA bound to a LPXTG sorting signal analog. The structure, combined with dynamics, kinetics, and whole cell protein display data suggest that the N terminus modulates substrate access to the enzyme. We propose that it may increase the efficiency of protein display by reducing the unproductive hydrolytic cleavage of enzyme-protein covalent intermediates that form during the cell wall anchoring reaction. Notably, a key active site loop (β7/β8 loop) undergoes a disordered to ordered transition upon binding the sorting signal, potentially facilitating recognition of lipid II.     

Initial Steps in the Pathway for Bacterial Degradation of Two Tetrameric Lignin Model Compounds

We investigated the metabolic route by which a lignin tetramer-degrading mixed bacterial culture degraded two tetrameric lignin model compounds containing β—O—4 and 5—5 biphenyl structures. The α-hydroxyl groups in the propane chain of both phenolic and nonphenolic tetramers were first oxidized symmetrically in two successive steps to give monoketones and diketones. These ketone metabolites were decomposed through C_α(=O)—C_β cleavage, forming trimeric carboxyl acids which were further metabolized through another C_α(=O)—C_β cleavage. Dehydrodiveratric acid, which resulted from the cleavage of the carbon bonds of the nonphenol tetramer, was demethylated twice. Four metabolites of the phenolic tetramer were purified and identified. All of these were stable compounds in sterile mineral medium, but were readily degraded by lignin tetramer-degrading bacteria along the same pathway as the phenol tetramer. No monoaromatic metabolites accumulated. All metabolites were identified by mass and proton magnetic resonance spectrometry. The metabolic route by which the mixed bacterial culture degraded tetrameric lignin model compounds was different from the route of the main ligninase-catalyzed C_α—C_β cleavage by Phanerochaete chrysosporium.     

A Colorimetric Technique for Detecting Trichothecenes and Assessing Relative Potencies

We tested a novel colorimetric toxicity test, based on inhibition of β-galactosidase activity in the yeast Kluyveromyces marxianus, for sensitivity to a range of mycotoxins. A variety of trichothecene mycotoxins could be detected. The order of toxicity established with this bioassay was verrucarin A > roridin A > T-2 toxin > diacetoxyscirpenol > HT-2 toxin > acetyl T-2 toxin > neosolaniol > fusarenon X > T-2 triol > scirpentriol > nivalenol > deoxynivalenol > T-2 tetraol. The sensitivity of detection was high, with the most potent trichothecene tested, verrucarin A, having a 50% effective concentration (concentration of toxin causing 50% inhibition) of 2 ng/ml. Other mycotoxins (cyclopiazonic acid, fumonisin B₁, ochratoxin A, patulin, sterigmatocystin, tenuazonic acid, and zearalenone) could not be detected at up to 10 μg/ml, nor could aflatoxins B₁ and M₁ be detected at concentrations up to 25 μg/ml. This test should be useful for trichothecene detection and for studies of relevant interactions—both between trichothecenes themselves and between trichothecenes and other food constituents.     

How Natalizumab Binds and Antagonizes α4 Integrins

Natalizumab antibody to α₄-integrins is used in therapy of multiple sclerosis and Crohn''s disease. A crystal structure of the Fab bound to an α₄ integrin β-propeller and thigh domain fragment shows that natalizumab recognizes human-mouse differences on the circumference of the β-propeller domain. The epitope is adjacent to but outside of a ligand-binding groove formed at the interface with the β-subunit βI domain and shows no difference in structure when bound to Fab. Competition between Fab and the ligand vascular cell adhesion molecule (VCAM) for binding to cell surface α₄β₁ shows noncompetitive antagonism. In agreement, VCAM docking models suggest that binding of domain 1 of VCAM to α₄-integrins is unimpeded by the Fab, and that bound Fab requires a change in orientation between domains 1 and 2 of VCAM for binding to α₄β₁. Mapping of species-specific differences onto α₄β₁ and α₄β₇ shows that their ligand-binding sites are highly conserved. Skewing away from these conserved regions of the epitopes recognized by current therapeutic function-blocking antibodies has resulted in previously unanticipated mechanisms of action.     

Homologous and Heterologous Overexpression in Clostridium acetobutylicum and Characterization of Purified Clostridial and Algal Fe-Only Hydrogenases with High Specific Activities

Clostridium acetobutylicum ATCC 824 was selected for the homologous overexpression of its Fe-only hydrogenase and for the heterologous expressions of the Chlamydomonas reinhardtii and Scenedesmus obliquus HydA1 Fe-only hydrogenases. The three Strep tag II-tagged Fe-only hydrogenases were isolated with high specific activities by two-step column chromatography. The purified algal hydrogenases evolve hydrogen with rates of around 700 μmol H₂ min⁻¹ mg⁻¹, while HydA from C. acetobutylicum (HydA_Ca) shows the highest activity (5,522 μmol H₂ min⁻¹ mg⁻¹) in the direction of hydrogen uptake. Further, kinetic parameters and substrate specificity were reported. An electron paramagnetic resonance (EPR) analysis of the thionin-oxidized HydA_Ca protein indicates a characteristic rhombic EPR signal that is typical for the oxidized H cluster of Fe-only hydrogenases.     

Molecular Anatomy of ParA-ParA and ParA-ParB Interactions during Plasmid Partitioning

Firmicutes multidrug resistance inc18 plasmids encode parS sites and two small homodimeric ParA-like (δ₂) and ParB-like (ω₂) proteins to ensure faithful segregation. Protein ω₂ binds to parS DNA, forming a short left-handed helix wrapped around the full parS, and interacts with δ₂. Protein δ₂ interacts with ω₂ and, in the ATP-bound form, binds to nonspecific DNA (nsDNA), forming small clusters. Here, we have mapped the ω₂·δ₂ and δ₂·δ₂ interacting domains in the δ₂ that are adjacent to but distinct from each other. The δ₂ nsDNA binding domain is essential for stimulation of ω₂·parS-mediated ATP hydrolysis. From the data presented here, we propose that δ₂ interacts with ATP, nsDNA, and with ω₂ bound to parS at near equimolar concentrations, facilitating a δ₂ structural transition. This δ₂ “activated” state overcomes its impediment in ATP hydrolysis, with the subsequent release of both of the proteins from nsDNA (plasmid unpairing).     

Identification and Validation of a Linear Protective Neutralizing Epitope in the β-Pore Domain of Alpha Toxin

The plethora of virulence factors associated with Staphylococcus aureus make this bacterium an attractive candidate for a molecularly-designed epitope-focused vaccine. This approach, which necessitates the identification of neutralizing epitopes for incorporation into a vaccine construct, is being evaluated for pathogens where conventional approaches have failed to elicit protective humoral responses, like HIV-1 and malaria, but may also hold promise for pathogens like S. aureus, where the elicitation of humoral immunity against multiple virulence factors may be required for development of an effective vaccine. Among the virulence factors employed by S. aureus, animal model and epidemiological data suggest that alpha toxin, a multimeric β-pore forming toxin like protective antigen from Bacillus anthracis, is particularly critical, yet no candidate neutralizing epitopes have been delineated in alpha toxin to date. We have previously shown that a linear determinant in the 2β2-2β3 loop of the pore forming domain of B. anthracis protective antigen is a linear neutralizing epitope. Antibody against this site is highly potent for neutralizing anthrax lethal toxin in vitro and for protection of rabbits in vivo from virulent B. anthracis. We hypothesized that sequences in the β-pore of S. aureus alpha toxin that share structural and functional homology to β-pore sequences in protective antigen would contain a similarly critical neutralizing epitope. Using an in vivo mapping strategy employing peptide immunogens, an optimized in vitro toxin neutralization assay, and an in vivo dermonecrosis model, we have now confirmed the presence of this epitope in alpha toxin, termed the pore neutralizing determinant. Antibody specific for this determinant neutralizes alpha toxin in vitro, and is highly effective for mitigating dermonecrosis and bacterial growth in a mouse model of S. aureus USA300 skin infection. The delineation of this linear neutralizing determinant in alpha toxin could facilitate the development of an epitope-focused vaccine against S. aureus.     

Amino Acid Composition of the Host-specific Toxin of Helminthosporium carbonum

The host-specific toxin of Helminthosporium carbonum (C₃₂H₅₀N₆O₁₀) was hydrolyzed by 6 n HCl to yield a number of α-amino acids. The common amino acids, proline and alanine, occurred in a ratio of 1:2. Two other unstable α-amino acids that produced lower color values with ninhydrin were also produced. One of these was tentatively identified as 2-amino-2,3-dehydro-3-methylpentanoic acid by electrolytic reduction to isoleucine. Additional ninhydrin-reacting substances were produced in low yield and probably represented secondary hydrolysis products of the unstable amino acids. The finding of an α,β-unsaturated linkage in H. carbonum toxin explains the instability of the compound and may also account for its specific toxicity.     

Resonance Raman Studies of Co—O2 and O—O Stretching Vibrations in Oxy-Cobalt Hemes

Strong evidence suggests that the stretching vibration of the bound oxygen can be perturbed by an accidentally degenerate porphyrin ring mode, resulting in two split frequencies. In the Co(II)(TpivPP) (pyridine) ¹⁸O₂ complex, we demonstrate that the ν(¹⁸O—¹⁸O) mode, after being shifted from its ν(¹⁶O—¹⁶O) value at 1,156 cm^-1, undergoes a resonance interaction with the 1,080 cm^-1 porphyrin mode, giving rise to two lines at 1,067 and 1,089 cm^-1. In the O₂ complex of Co(II) mesoporphyrin IX-substituted sperm whale myoglobin, we observed a dramatic intensity increase at 1,132 cm^-1 upon ¹⁶O₂ → ¹⁸O₂ substitution, which is due to the reappearance of the 1,132-cm^-1 porphyrin mode after the removal of resonance conditions. A decrease in O₂ binding affinity, caused by the proximal base tension, corresponds to an increase in the Co—O₂ stretching frequency. The ν(Co—O₂) at 527 cm^-1 for the low affinity Co(II)(TpivPP)(1,2-Me₂Im) O₂ complex is 11 cm^-1 higher than the 516-cm^-1 value for the high affinity complex (with N-MeIm replacing 1,2-Me₂Im). However, in the corresponding iron complexes the reverse behavior is observed, i.e., the ν(Fe—O₂) decreases for the (1,2-Me₂Im) complex. There is a 24-cm^-1 difference in the Co—O₂ stretching frequencies between Co(II)(TpivPP)(N-MeIm)O₂ (at 516 cm^-1) and oxy meso CoMb (at 540 cm^-1), suggesting a protein induced distortion of the Co—O—O linkage. However, the values for ν(Fe—O₂) are nearly identical between Fe(II)(TpivPP)(N-MeIm)O₂ (at 571 cm^-1) and oxy Mb (at 573 cm^-1), indicating that O₂ binds to myoglobin in the same manner as in the sterically unhindered “picket fence” complex. Evidence is presented that suggests the presence of two dioxygen stretching frequencies due to two different conformers in each of the N-MeIm and 1,2-Me₂Im complex of oxy Co(II)(TpivPP).     

Characterization of Two Novel α-Glucosidases from Bifidobacterium breve UCC2003

Two α-glucosidase-encoding genes (agl1 and agl2) from Bifidobacterium breve UCC2003 were identified and characterized. Based on their similarity to characterized carbohydrate hydrolases, the Agl1 and Agl2 enzymes are both assigned to a subgroup of the glycosyl hydrolase family 13, the α-1,6-glucosidases (EC 3.2.1.10). Recombinant Agl1 and Agl2 into which a His₁₂ sequence was incorporated (Agl1_His and Agl2_His, respectively) exhibited hydrolytic activity towards panose, isomaltose, isomaltotriose, and four sucrose isomers—palatinose, trehalulose, turanose, and maltulose—while also degrading trehalose and, to a lesser extent, nigerose. The preferred substrates for both enzymes were panose, isomaltose, and trehalulose. Furthermore, the pH and temperature optima for both enzymes were determined, showing that Agl1_His exhibits higher thermo and pH optima than Agl2_His. The two purified α-1,6-glucosidases were also shown to have transglycosylation activity, synthesizing oligosaccharides from palatinose, trehalulose, trehalose, panose, and isomaltotriose.     

Cellular resistance to the antimelanoma immunotoxin ZME-gelonin and strategies to target resistant cells

 The development of cellular resistance to immunotoxins has been demonstrated in a variety of models and can involve a number of mechanisms. For the present study, an immunotoxin was utilized composed of an antimelanoma antibody ZME-018 recognizing a 240-kDa surface glycoprotein (gp 240) and the plant toxin gelonin. Human melanoma cells (A375-M) were grown in the presence of increasing amounts of ZME-gelonin and a clonal variant (A-375-ZR) was developed that was 100-fold resistant to ZME-gelonin compared to parental cells. Scatchard analysis showed that the A375-M parental cells had 260×10³ ZME-gelonin-binding sites/cell with relatively low affinity (5 nM). In contrast, resistant A375-ZR cells demonstrated a reduced number of low-affinity sites (160×10³/cell), but showed a small number (47×10³) of higher-affinity sites (0.8 nM). Internalization rates and degradation rates of ¹²⁵I-labeled ZME-gelonin were identical in both the parental and resistant cells. A375-ZR cells were found to be more resistant to vincristine and doxorubicin than were parental cells. Both cell lines were almost equally sensitive to native gelonin, 5-fluorouracil (5-FU), cisplatin, melphalan, carmustine, interferon γ (IFNγ) and IFNα. In addition, both cell lines were equally sensitive to another gelonin-antibody conjugate that binds to cell-surface, GD₂ (antibody 14G₂A). However, resistant cells were twice as sensitive to the cytotoxic effects of etoposide than were parental cells. Finally, a variety of agents were tested in combination with ZME-gelonin against A375-ZR cells in an attempt to identify agents to augment immunotoxin cytotoxic effects against resistant cells. The agents 5-FU, cisplatin, IFNγ, IFNα, and etoposide were the most effective in augmenting the cytotoxicity of ZME-gelonin against resistant cells. These studies suggest that development of resistance to one immunotoxin does not cause development of cross-resistance to other gelonin immunotoxins. Further, specific biological response modifiers and chemotherapeutic agents may be effective in augmenting the effectiveness of immunotoxins and specifically targeting or reducing the emergence of immunotoxin-resistant cells. Received: 15 March 1995 / Accepted: 28 November 1995     

Dissipation of the Membrane Potential in Susceptible Corn Mitochondria by the Toxin of Helminthosporium maydis, Race T, and Toxin Analogs

We have tested directly the effect of Helminthosporium maydis T (Hmt) toxin and various analogs on the membrane potential formed in mitochondria isolated from a Texas (T) cytoplasmic male-sterile and a normal (N) corn. ATP, malate or succinate generated a membrane potential (negative inside) as monitored by the absorbance change of a cationic dye, safranine. The relative membrane potential (Δψ) could also be detected indirectly as ⁴⁵Ca²⁺ uptake. Hmt toxin added to T mitochondria dissipated the steady state Δψ similar to addition of a protonophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP). Toxin analogs (Cpd XIII: C₄₁H₆₈O₁₂ and Cpd IV: C₂₅H₄₄O₆), reduced native toxin (RT2C: C₄₁H₈₄O₁₃) and Pm toxin (band A: C₃₃H₆₀O₈, produced by the fungus, Phyllosticta maydis) were effective in dissipating Δψ and decreasing Ca²⁺ uptake with the following order: Pm (100) » HmT (23-30) > Cpd XIII (11-25) » RT2C (0-4−1.8) > Cpd IV (0.2−1.0). In contrast, the toxins and analogs had no effect on Δψ formed in N mitochondria. The striking similarities of the HmT toxin (band 1: C₄₁H₆₈O₁₃) and Cpd XIII on T mitochondrial activities provide strong evidence supporting the correctness of the polyketol structure assigned to the native toxin. Since the Δψ in energized mitochondria is caused mainly by the electrogenic extrusion of H⁺, the results support the idea that HmT toxin increases membrane permeability of T mitochondria to H⁺. The host specificity of the toxin suggests that an interaction with unique target site(s) on the inner mitochondrial membrane of T corn causes H⁺ leakage.     

High contrast tumor imaging with radio-labeled antibody Fab fragments tailored for optimized pharmacokinetics via PASylation

Although antigen-binding fragments (Fabs) of antibodies constitute established tracers for in vivo radiodiagnostics, their functionality is hampered by a very short circulation half-life. PASylation, the genetic fusion with a long, conformationally disordered amino acid chain comprising Pro, Ala and Ser, provides a convenient way to expand protein size and, consequently, retard renal filtration. Humanized αHER2 and αCD20 Fabs were systematically fused with 100 to 600 PAS residues and produced in E. coli. Cytofluorimetric titration analysis on tumor cell lines confirmed that antigen-binding activities of the parental antibodies were retained. The radio-iodinated PASylated Fabs were studied by positron emission tomography (PET) imaging and biodistribution analysis in mouse tumor xenograft models. While the unmodified αHER2 and αCD20 Fabs showed weak tumor uptake (0.8% and 0.2% ID/g, respectively; 24 h p.i.) tumor-associated radioactivity was boosted with increasing PAS length (up to 9 and 26-fold, respectively), approaching an optimum for Fab-PAS₄₀₀. Remarkably, 6- and 5-fold higher tumor-to-blood ratios compared with the unmodified Fabs were measured in the biodistribution analysis (48 h p.i.) for αHER2 Fab-PAS₁₀₀ and Fab-PAS₂₀₀, respectively. These findings were confirmed by PET studies, showing high imaging contrast in line with tumor-to-blood ratios of 12.2 and 5.7 (24 h p.i.) for αHER2 Fab-PAS₁₀₀ and Fab-PAS₂₀₀. Even stronger tumor signals were obtained with the corresponding αCD20 Fabs, both in PET imaging and biodistribution analysis, with an uptake of 2.8% ID/g for Fab-PAS₁₀₀ vs. 0.24% ID/g for the unmodified Fab. Hence, by engineering Fabs via PASylation, plasma half-life can be tailored to significantly improve tracer uptake and tumor contrast, thus optimally matching reagent/target interactions.