NF-kappaB activation provides the potential link between inflammation and hyperplasia in the arthritic joint

The transcription factor NF-kappaB is a pivotal regulator of inflammatory responses. While the activation of NF-kappaB in the arthritic joint has been associated with rheumatoid arthritis (RA), its significance is poorly understood. Here, we examine the role of NF-kappaB in animal models of RA. We demonstrate that in vitro, NF-kappaB controlled expression of numerous inflammatory molecules in synoviocytes and protected cells against tumor necrosis factor alpha (TNFalpha) and Fas ligand (FasL) cytotoxicity. Similar to that observed in human RA, NF-kappaB was found to be activated in the synovium of rats with streptococcal cell wall (SCW)-induced arthritis. In vivo suppression of NF-kappaB by either proteasomal inhibitors or intraarticular adenoviral gene transfer of super-repressor IkappaBalpha profoundly enhanced apoptosis in the synovium of rats with SCW- and pristane-induced arthritis. This indicated that the activation of NF-kappaB protected the cells in the synovium against apoptosis and thus provided the potential link between inflammation and hyperplasia. Intraarticular administration of NF-kB decoys prevented the recurrence of SCW arthritis in treated joints. Unexpectedly, the severity of arthritis also was inhibited significantly in the contralateral, untreated joints, indicating beneficial systemic effects of local suppression of NF-kappaB. These results establish a mechanism regulating apoptosis in the arthritic joint and indicate the feasibility of therapeutic approaches to RA based on the specific suppression of NF-kappaB.

Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-alpha

Tumour-necrosis factor-alpha (TNF-alpha) is a cytokine that contributes to a variety of inflammatory disease states. The protein exists as a membrane-bound precursor of relative molecular mass 26K which can be processed by a TNF-alpha-converting enzyme (TACE), to generate secreted 17K mature TNF-alpha. We have purified TACE and cloned its complementary DNA. TACE is a membrane-bound disintegrin metalloproteinase. Structural comparisons with other disintegrin-containing enzymes indicate that TACE is unique, with noteable sequence identity to MADM, an enzyme implicated in myelin degradation, and to KUZ, a Drosophila homologue of MADM important for neuronal development. The expression of recombinant TACE (rTACE) results in the production of functional enzyme that correctly processes precursor TNF-alpha to the mature form. The rTACE provides a readily available source of enzyme to help in the search for new anti-inflammatory agents that target the final processing stage of TNF-alpha production.

Cross-desensitization of receptors for peptide chemoattractants. Characterization of a new form of leukocyte regulation

The formylpeptide (fMLP) and C5a chemoattractants were previously shown to cross-desensitize each other's ability to mobilize Ca2+ in leukocytes but not to affect nonchemoattractant Ca(2+)-mobilizing receptors, and vice versa. Our data show that all receptors studied underwent homologous desensitization. Interestingly, peptide chemoattractants (fMLP, C5a, and IL-8) desensitized each other's Ca(2+)-mobilizing responses, but had no effect on a Ca(2+)-mobilizing purinergic receptor. Lipid chemoattractant receptors (PAF and leukotriene B4) were also desensitized by peptide chemoattractants but not vice versa. In the presence of cytochalasin B, only fMLP and C5a caused the activation of phospholipase D in intact leukocytes and enhanced desensitization of IL-8 and C5a but not fMLP receptors. To measure receptor/G protein interactions, agonist-stimulated GTP gamma S binding to leukocyte membranes was measured. Whereas all peptide receptors underwent homologous desensitization, C5a and IL-8, but not fMLP, receptors were cross-desensitized by other peptide chemoattractants. Furthermore, PMA caused inhibition of C5a- and IL-8- but not fMLP-stimulated GTP gamma S binding. These data suggest that in addition to homologous desensitization, peptide chemoattractant receptors cross-desensitize one another by at least two processes. One can be detected at the level of receptor/G-protein interaction and possibly involves receptor phosphorylation by protein kinase C. The fMLP receptor is resistant to this process. The second process is distal to receptor/G-protein interaction and utilizes an undefined pathway to cross-desensitize the Ca2+ mobilization response to all peptide chemoattractants. We propose that receptor cross-desensitization in leukocytes is orchestrated at several levels by mechanisms with selectivity for types of chemoattractant receptors.

Cross-desensitization of receptors for peptide chemoattractants. Characterization of a new form of leukocyte regulation

The formylpeptide (fMLP) and C5a chemoattractants were previously shown to cross-desensitize each other's ability to mobilize Ca2+ in leukocytes but not to affect nonchemoattractant Ca(2+)-mobilizing receptors, and vice versa. Our data show that all receptors studied underwent homologous desensitization. Interestingly, peptide chemoattractants (fMLP, C5a, and IL-8) desensitized each other's Ca(2+)-mobilizing responses, but had no effect on a Ca(2+)-mobilizing purinergic receptor. Lipid chemoattractant receptors (PAF and leukotriene B4) were also desensitized by peptide chemoattractants but not vice versa. In the presence of cytochalasin B, only fMLP and C5a caused the activation of phospholipase D in intact leukocytes and enhanced desensitization of IL-8 and C5a but not fMLP receptors. To measure receptor/G protein interactions, agonist-stimulated GTP gamma S binding to leukocyte membranes was measured. Whereas all peptide receptors underwent homologous desensitization, C5a and IL-8, but not fMLP, receptors were cross-desensitized by other peptide chemoattractants. Furthermore, PMA caused inhibition of C5a- and IL-8- but not fMLP-stimulated GTP gamma S binding. These data suggest that in addition to homologous desensitization, peptide chemoattractant receptors cross-desensitize one another by at least two processes. One can be detected at the level of receptor/G-protein interaction and possibly involves receptor phosphorylation by protein kinase C. The fMLP receptor is resistant to this process. The second process is distal to receptor/G-protein interaction and utilizes an undefined pathway to cross-desensitize the Ca2+ mobilization response to all peptide chemoattractants. We propose that receptor cross-desensitization in leukocytes is orchestrated at several levels by mechanisms with selectivity for types of chemoattractant receptors.

Differences in phosphorylation of formylpeptide and C5a chemoattractant receptors correlate with differences in desensitization

To define the regulation of chemoattractant receptors, epitope-tagged human formyl peptide and C5a receptor cDNAs (ET-FR and ET-C5aR) were stably expressed in rat basophilic leukemia, RBL-2H3 cells. An antibody (12CA5) specific to "ET" was used to immunoprecipitate ET-FR and ET-C5aR. fMLP and C5a caused time- and dose-dependent phosphorylation of their respective receptors. Phosphorylated ET-FR migrated as a single broad band between 50 and 70 kDa on SDS-polyacrylamide gel electrophoresis, whereas ET-C5aR exhibited both fast (39-45 kDa) and broadly (39-52 kDa) migrating forms. Fast form phosphorylation alone was observed at low concentrations of C5a (0.001-0.01 microM), or at early times (5-30 s) with a higher concentration of C5a (0.1 microM). Phorbol 12-myristate 13-acetate, thrombin, or antigen caused no phosphorylation of ET-FR but stimulated exclusively fast form phosphorylation of ET-C5aR. The protein kinase C inhibitor staurosporine did not inhibit phosphorylation of ET-FR but blocked the fast migrating component of phosphorylated ET-C5aR. Homologous desensitization correlated with ligand-induced phosphorylation of both receptors. Of note, ET-C5aR but not ET-FR underwent heterologous desensitization by antigen, phorbol 12-myristate 13-acetate, and thrombin. The data suggest that protein kinase C mediates heterologous phosphorylation and desensitization of C5aR but not FR, yet, both receptors are homologously desensitized by a staurosporine-resistant kinase.

Rac1, a low-molecular-mass GTP-binding-protein with high intrinsic GTPase activity and distinct biochemical properties

Rac1, a member of the family of low-molecular-mass GTP-binding proteins, functions in phagocytic leukocytes as a component necessary for activation of the respiratory burst. To characterize the biochemical properties of rac1, the protein was expressed as a fusion protein in Escherichia coli and purified to greater than 99% homogeneity by affinity chromatography. Rac1 protein bound maximally bound and hydrolyzed GTP under low free-Mg2+ concentrations. Under those conditions, (45 nm free Mg2+), purified rac1 exhibited a steady-state GTPase activity of 18 nmol.min-1.mg protein-1 (turnover number approximately 0.39 min-1 at 37 degrees C), which is 40-fold higher than H-ras. The high intrinsic GTPase activity of rac1 under low free Mg2+ was mainly due to an increased kcat, the rate constant for hydrolysis of bound GTP, which was 0.29 min-1 for rac1 vs 0.007 min-1 for H-ras (at 20 degrees C). Rac1 also released bound GDP faster than H-ras (koff.GDP = 1.02 min-1 for rac1 vs 0.33 min-1 for H-ras at 20 degrees C). In contrast, rac1 released bound guanosine 5'-[gamma-thio]triphosphate (GTP[S]) at a slower rate than H-ras (koff.GTP[S] approximately 0.04 min-1 for rac1 vs 0.31 min-1 for H-ras at 20 degrees C). Rac1 was a very good substrate for in vitro geranylgeranylation (C20) but not for farnesylation (C15), whereas the converse is true for H-ras. Surprisingly, rac1 was a very poor substrate for in vitro ADP-ribosylation by the C3 component of Clostridium botulinum toxin compared to rhoA. As a further characterization of rac1, a mutant was made in which the Thr115 was replaced by asparagine. This protein (referred to as [Thr115----Asn]rac1) contains the consensus amino acids of all four GTP-binding domains of H-ras. The koff.GDP of [Thr115----Asn]rac1 was reduced to that of H-ras, but [Thr115----Asn]rac1 exhibited essentially identical kcat (0.13 min-1 at 20 degrees C) and koff-GTP[S] (0.03 min-1 at 20 degrees C) values as the wild-type protein. Thus, the region(s) in rac1 which control the dissociation of GTP[S] (and presumably GTP) do not entirely coincide with those controlling GDP dissociation. Biochemical analysis of [Thr115----Asn]rac1 also suggests that the region responsible for the increased kcat of rac1 is not within the consensus amino acids of the four guanine-nucleotide-binding domains.

Purification and characterization of an ADP-ribosyltransferase produced by Clostridium limosum

We purified a novel ADP-ribosyltransferase produced by a Clostridium limosum strain isolated from a lung abscess and compared the exoenzyme with Clostridium botulinum ADP-ribosyltransferase C3. The C. limosum exoenzyme has a molecular weight of about 25,000 and a pI of 10.3. The specific activity of the ADP-ribosyltransferase is 3.1 nmol/mg/min with a Km for NAD of 0.3 microM. Partial amino acid sequence analysis of the tryptic peptides revealed about 70% homology with C3. The novel exoenzyme modifies selectively the small GTP-binding proteins of the rho family in human platelet membranes presumably at the same amino acid (asparagine 41) as known for C3. Recombinant rhoA and rhoB serve as substrates for C3 and the C. limosum exoenzyme. Whereas recombinant rac1 protein is only marginally ADP-ribosylated by C3 or by the C. limosum exoenzyme in the absence of detergent, in the presence of 0.01% sodium dodecyl sulfate rac1 is modified by C3 but not by the C. limosum exoenzyme. Recombinant CDC42Hs protein is a poor substrate for C. limosum exoenzyme and is even less modified by C3. The C. limosum exoenzyme is auto-ADP-ribosylated in the presence of 0.01% sodium dodecyl sulfate by forming an ADP-ribose protein bond highly stable toward hydroxylamine. The data indicate that ADP-ribosylation of small GTP-binding proteins of the rho family is not unique to C. botulinum C3 ADP-ribosyltransferase but is also catalyzed by a C3-related exoenzyme from C. limosum.

Differential regulation of cAMP by endogenous versus transfected formylpeptide chemoattractant receptors: implications for Gi-coupled receptor signaling

Endogenous neutrophil formylpeptide receptors do not inhibit adenylylcyclase activation. The ability of a cloned and transfected human formylpeptide receptor to mediate the inhibition of adenylylcyclase was assessed in the human embryonic kidney 293 TSA cell line. Inclusion of 1 microM fMetLeuPhe resulted in a ca. 50% inhibition of isoproterenol-stimulated cAMP in transfected cells. Activation of adenylylcyclase by isoproterenol was inhibited ca. 30% by fMetLeuPhe in membranes prepared from transfected cells but not in membranes prepared from neutrophils. Prior treatment of transfected cells with pertussis toxin abrogated the inhibitory effect of fMetLeuPhe. These data indicate that factors in addition to the primary structure of the formylpeptide receptor govern its transductional activities.

Functional high efficiency expression of cloned leukocyte chemoattractant receptor cDNAs

Human kidney 293 TSA cells were transfected by a calcium phosphate method with human formylpeptide and C5a receptor cDNAs with high efficiency. Formylpeptide receptor positive transfectants expressed a total of 968,000 +/- 34,000 receptors per cell with two affinity states (Kds of ca. 0.43 nM and 39 nM), which in the presence of 100 microM GTP gamma S decreased by ca. 4-fold the number of high-affinity sites. The ligand binding pharmacology of cloned and expressed formylpeptide receptors were indistinguishable from endogenous receptors on human neutrophils. Expressed formylpeptide and C5a receptors were functionally active in mobilizing intracellular calcium via a pertussis toxin sensitive mechanism with an ED50 for formylpeptide of ca. 0.5-1.0 nM. This expression system, in which receptor expression can be monitored by flow cytometric methods and in which intracellular calcium responses are measurable, unlike in the more popular COS-7 cell expression system, will provide a useful basis for the analysis of chemoattractant receptor structure-function relationships.

Receptor class desensitization of leukocyte chemoattractant receptors

To better define their regulation, formylpeptide and C5a chemoattractant receptor cDNAs were transiently expressed with high efficiency (approximately 35-54%) in human kidney cells. As in neutrophils, both receptors were active in elevating intracellular calcium (ED50 approximately 0.5-1 nM). Agonist-specific desensitization for calcium elevation was observed for both chemoattractant receptors at doses of approximately 1 nM. Heterologous desensitization of formylpeptide, C5a, and alpha 1-adrenergic receptors required high doses of phorbol ester (100 nM phorbol 12-myristate 13-acetate). To further study the phenomenon of desensitization, formylpeptide and C5a receptor cDNAs were cotransfected resulting in approximately 80% of receptor-positive cells expressing both receptors. These cells also possessed endogenous alpha 1-adrenergic receptors. Interestingly, chemoattractant receptors were cross-desensitized by pretreatment with low doses of either C5a or formylmethionylleucylphenylalanine (10 nM) but not by the alpha 1-adrenergic agonist norepinephrine (up to 10 microM). Neither chemoattractant desensitized alpha 1-adrenergic receptors. This phenomenon was reproduced in human neutrophils. These data suggest a previously uncharacterized mechanism of receptor regulation, which is intermediate between homologous and heterologous desensitization. Class desensitization of chemoattractant receptors is less selective than homologous desensitization but is far more efficient and specific than heterologous desensitization. Receptor class desensitization may affect functional classes of receptors via modification of either the receptor or the shared guanine nucleotide-binding regulatory protein.

Isoprenylation of the low molecular mass GTP-binding proteins rac 1 and rac 2: possible role in membrane localization

Ras proteins can be modified at their COOH-terminal cysteine in the motif Cys-Ali-Ali-Xaa by a farnesyl isoprenoid. This modification is essential for membrane association and biological activity of ras proteins. A similar COOH-terminal amino acid sequence, Cys-Xaa-Ali-Xaa, exists in the ras-related GTP-binding proteins rac 1 and rac 2. To determine whether these proteins were similarly modified, COS cells were transfected with rac 1 and rac 2 cDNA and expressed proteins were labeled with [3H]mevalonic acid. We report here that both rac 1 and rac 2 are post-translationally modified by addition of an isoprenoid group, the likely site of which is the COOH-terminal cysteine. Isoprenylation was found only in racs associated with particulate cell fractions, suggesting that this modification may be associated with membrane localization of the proteins. These data specifically identify mammalian low molecular mass GTP-binding proteins other than ras that undergo post-translational modification and further define the COOH-terminal consensus sequence, Cys-Ali-Ali-Xaa, as an isoprenylation signal. This sequence may identify a larger family of low molecular mass GTP-binding proteins which are isoprenylated.

Identification of the ral and rac1 gene products, low molecular mass GTP-binding proteins from human platelets

Identification of the GTP-binding proteins from human platelet particulate fractions was attained by their purification via successive column chromatography steps followed by amino acid sequencing. To enhance the likelihood of identifying the GTP-binding proteins, two assays were employed to monitor GTP-binding activities: (i) guanosine 5'-(3-O-[35S]thio)triphosphate (GTP gamma S)-binding followed by rapid filtration and ii) [alpha-32P]GTP-binding following sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electroblotting onto nitrocellulose membranes. The latter assay permitted the isolation of a 28-kDa GTP-binding protein that bound [alpha-32P]GTP prominently but was only poorly detected with the GTP gamma S-binding assay. The amino acid sequences of three peptide fragments derived from the 28-kDa protein were identical to regions of the amino acid sequence deduced from a simian ral cDNA with the exception of one conservative substitution (Asp147----Glu). A full length human ral cDNA was isolated from a placental cDNA library, and its deduced amino acid sequence, compared with simian ral, also contained the Asp----Glu substitution along with two other substitutions and an additional three NH2-terminal amino acids. In addition to the 28-kDa protein, two distinct 25-kDa GTP-binding proteins were purified from platelets. One of these proteins has been previously characterized as G25K, an abundant low molecular mass GTP-binding protein. Partial amino acid sequence obtained from the second unidentified 25-kDa protein indicates that it is the product of the rac1 gene; a member of a newly identified gene family which encode for low molecular mass GTP-binding proteins (Didsbury, J., Weber, R.F., Bokoch, G. M., Evans, T., and Snyderman, R. (1989) J. Biol. Chem. 264, 16378-16382). These results identify two new GTP-binding proteins in human platelets, ral, the major protein that binds [alpha-32P]GTP on nitrocellulose transfers, and rac1, a substrate for botulinum C3 ADP-ribosyltransferase.

Molecular cloning of a new human G protein. Evidence for two Gi alpha-like protein families

The amino acid sequence of a novel G protein alpha subunit (Gx alpha) has been deduced from the nucleotide sequence of a human cDNA clone isolated from a differentiated HL-60 cDNA library. The cDNA encodes a polypeptide of 354 amino acids (Mr 40,519) which is closely related to Gi alpha proteins. The amino acid sequence homology between Gx alpha and human myeloid Gi alpha is 86% with 15 nonconservative substitutions. Gx alpha also shares 86% homology with both rat brain and mouse macrophage Gi alpha but is more homologous (94%) to bovine brain Gi alpha with only 5 nonconservative amino acid differences. G proteins previously termed Gi alpha may fall into at least two distinct groups, with one including human myeloid Gi alpha, rat brain Gi alpha and mouse macrophage Gi alpha; and other Gx alpha and bovine brain Gi alpha. One group probably contains true Gi and the other a new class of G protein whose function remains to be determined.

Differences in the regulation of messenger RNA for housekeeping and specialized-cell ferritin. A comparison of three distinct ferritin complementary DNAs, the corresponding subunits, and identification of the first processed in amphibia

The ferritin family is a widespread group of proteins that maintain iron in a soluble form and also protect against the toxic effects of excess iron. The structure and sequence of the proteins are highly conserved. However, the cell-specific features of structure which occur within the same organism indicate cell specificity of gene expression and may be related to variations in types of iron storage, i.e. specialized-cell ferritin (stored iron is for other cell types) versus housekeeping ferritin (stored iron is for intracellular purposes related to normal or stress metabolism); the protein structure may also affect rates of iron turnover. Iron induces ferritin synthesis and accumulation by recruiting stored ferritin mRNA that is efficiently translated in cells specialized for iron storage. For the first time we show the occurrence of three different cDNAs from bullfrog tadpoles, corresponding to three subunits of the protein: H, M, and L. Thus, ferritin can be encoded by at least three different mRNAs and probably three different genes, in contrast to the older idea of two, H and L; the subunits maintain the conserved sequences of known ferritins and have similar predicted masses, 20.5, 20.6, and 19.9 kDa, but have distinct mobilities in denaturing gels. Ferritin subunit expression is cell specific; more of the H and L chain mRNAs are expressed in red cells than in liver. Ferritin expression is regulated by transcription (or mRNA stability) in adult red cells; cellular levels of ferritin mRNA were 20% that of embryonic red cells, and L subunit mRNA increased 2.5 times with excess iron. Ferritin expression is also regulated during translation in adult red cells; iron recruits stored ferritin mRNA, but only during certain stages of red cell maturation, in contrast to embryonic red cells. The developmental differences in ferritin expression are discussed in relation to the shift from specialized-cell ferritin to housekeeping ferritin in red cells of the embryonic versus adult lines.

Human Gi protein alpha-subunit: deduction of amino acid structure from a cloned cDNA

The amino acid sequence of the alpha-subunit of Gi, the human adenylate cyclase inhibiting GTP-binding protein, has been deduced from the nucleotide sequence of a DNA clone complementary to Gi alpha mRNA from differentiated U937 cells. The cDNA encodes a polypeptide of 355 amino acids (Mr 40456). The amino acid sequence homology between human Gi alpha and rat, murine, and bovine Gi alpha is 98.6, 97.7 and 87.9% respectively. Differentiation of the U937 cells from monoblasts to monocyte-like cells resulted in a 3-fold increase in Gi alpha mRNA as well as a 3.6-fold increase in the 41 kDa pertussis toxin substrate presumed to be Gi alpha. Thus, increased levels of this G-protein are associated with monocyte differentiation and appear to be regulated transcriptionally.

Multiple red cell ferritin mRNAs, which code for an abundant protein in the embryonic cell type, analyzed by cDNA sequence and by primer extension of the 5'-untranslated regions

Ferritin maintains iron in a bioavailable, nontoxic form for vertebrates and invertebrates, higher plants, fungi, and bacteria; the protein is formed from two classes of subunits (H and L) in ratios which vary in different cell types. Ferritin may be an abundant, differentiation-specific protein or a "housekeeping" protein. The red cells of embryos are specialized for iron storage and have abundant ferritin; iron regulates the synthesis of ferritin in such cells translationally by recruitment of stored, ferritin mRNA and by translational competition. To characterize mRNA regulated in such a manner, we prepared cDNA from reticulocytes of bullfrog tadpoles, a readily available source of embryonic red cells; moreover, no protein sequence information was available for nonmammalian ferritin. An almost full-length (817 base pairs) cDNA (pJD5F12) was isolated and sequenced, the 5' end was analyzed by primer extension, and the cloned DNA was used as a hybridization probe. We have shown that ferritin mRNA is stored in the cytoplasm and that the 5' end of the mRNA is heterogeneous. The 5'-untranslated region of ferritin mRNA consisted of 143 nucleotides in the major (65%) species and 146 or 152 in the minor species (approximately 17% each). (Heterogeneity is characteristic of some other abundant mRNAs, e.g. globin, which is also translationally regulated.) Since excess iron had no detectable effect on the heterogeneity of the 5' end of ferritin mRNA, the feature is more likely associated with mRNA abundance and/or cell specialization than translational control. In the bullfrog, as in humans and rats, ferritin is encoded by multiple genomic sequences (four to eight) which specify proteins of considerable homology. For example, 75 of the 81 amino acids present in all mammalian ferritins sequenced are also present in the frog; the overall homology between frogs and humans or rats is 59-66%. Ferritin H and L subunits in humans are distinct (overall homology 56%) and appear to have diverged from a common precursor relatively recently. In contrast, ferritin H and L subunits have high homology in tadpole red cells, determined by hybrid select translation, which suggests that bullfrog red cell ferritin may be close to the primordial sequence.

Binding and uptake of diphtheria toxin by toxin-resistant Chinese hamster ovary and mouse cells

We investigated two phenotypically distinct types of diphtheria toxin-resistant mutants of Chinese hamster cells and compared their resistance with that of naturally resistant mouse cells. All are resistant due to a defect in the process of internalization and delivery of toxin to its target in the cytosol, elongation factor 2. By cell hybridization studies, analysis of cross-resistance, and determination of specific binding sites for 125I-labeled diphtheria toxin, we showed that these cell strains fall into two distinct complementation groups. The Dipr group encompasses Chinese hamster strains that are resistant only to diphtheria toxin, as well as mouse LM cells. These strains possess a normal complement of high-affinity binding sites for diphtheria toxin, but these receptors are unable to deliver active toxin fragment A to the cytosol. Cells of the DPVr group have a broader spectrum of resistance, including Pseudomonas exotoxin A and several enveloped viruses as well as diphtheria toxin. In these studies, which investigate the resistance of these cells to diphtheria toxin, we demonstrate that they possess a reduced number of specific binding sites for this toxin and behave, phenotypically, like cells treated with the proton ionophore monensin. Their resistance is related to a defect in a mechanism required for release of active toxin from the endocytic vesicle.