Rate of sister chromatid exchanges in Bloom syndrome fibroblasts reduced by co-cultivation with normal fibroblasts

Six strains of Bloom syndrome (BlS) fibroblasts responded to co-cultivation with normal fibroblasts at a 1:2 ratio by a reduced rate of sister chromatid exchanges (SCE's) from a mean of 67.5 (range = 59--78) to 28.4 (range = 21--35). The response was dose-dependent in one strain tested at 1:2, 1:1, and 2:1 ratios. In addition, quadriradial exchange figures and other signs of increased chromosomal instability were not found in BlS cells following co-cultivation with control cells. Control cells did not respond to BlS cells and maintained a normal rate of SCEs. Culture medium conditioned for 48 hrs by normal fibroblasts could also reduce the rate of SCEs in BlS fibroblasts, but less than in co-cultivation. We suggest that the reduced rate of SCEs and the lack of chromosomal instability in BlS cells following co-cultivation represent a corrective effect that is related to the basic defect and not dependent on cell-to-cell contact.     

Chromosomal aberrations in lymphocytes predict human cancer: a report from the European Study Group on Cytogenetic Biomarkers and Health (ESCH)

Chromosomal aberrations (CAs), sister chromatid exchanges (SCEs), and micronuclei (MN) in peripheral blood lymphocytes have for decades been used as cytogenetic biomarkers to survey genotoxic risks in the work environment. The conceptual basis for this application has been the idea that increased cytogenetic damage reflects an enhanced cancer risk. Nordic and Italian cohorts have been established to evaluate this hypothesis, and analyses presented previously have shown a positive trend between CA frequency and increased cancer risk. We now report on a pooled analysis of updated data for 3541 subjects examined for CAs, 2703 for SCEs, and 1496 for MN. To standardize for interlaboratory variation, the results for the various cytogenetic end points were trichotomized on the basis of the absolute value distribution within each laboratory as "low" (1-33 percentile), "medium" (34-66 percentile), or "high" (67-100 percentile). In the Nordic cohort, there was an elevated standardized incidence ratio (SMR) for all cancer among subjects with high CA frequency [1.53; 95% confidence interval (CI), 1.13-2.05] but not for those with medium or low CA frequency. In the Italian cohort, a SMR in cancer of 2.01 (95% CI, 1.35-2.89) was obtained for those with a high CA frequency level, whereas the SMRs for those with medium or low did not noticeably differ from unity. Cox's proportional hazards models gave no evidence that the effect of CAs on total cancer incidence/mortality was modified by gender, age at test, or time since test. No association was seen between the SCEs or the MN frequencies and subsequent cancer incidence/mortality. The present study further supports our previous observation on the cancer predictivity of the CA biomarker, which seems to be independent of age at test, gender, and time since test. The risk patterns were similar within each national cohort. This result suggests that the frequency of CAs in peripheral blood lymphocytes is a relevant biomarker for cancer risk in humans, reflecting either early biological effects of genotoxic carcinogens or individual cancer susceptibility.     

Surface expression of CD63 antigen (AD1 antigen) in P815 mastocytoma cells by transfected IgE receptors

We have examined epidermal growth factor (EGF) signalling in a CHO cell line (CHO11) which expresses a human EGF receptor truncated at amino acid 990. Previous studies showed that EGF treatment of these cells failed to increase prostaglandin production or phospholipase A2 activity. In the current study EGF increased the tyrosine phosphorylation of the intracellular signalling protein Shc in CHO11 cells but did not activate either of the downstream signalling enzymes raf or mitogen activated protein kinase (MAPK). The uncoupling of Shc activation from distal signalling in CHO11 cells contrasts with other cells which express similar mutant EGF receptors. The failure of She to activate distal signalling may reflect qualitative differences in the way that this protein is activated or could result from the activation of an inhibitory signalling pathway.     

Hysteroscopy: a review

3 chemicals were selected for mutagenicity testing from a priority list, based on production volume and available mutagenicity data. Propargyl alcohol (PA), 2-nitroaniline (2NA), and 5-methyl-1H-benzo-triazole (MBT) were selected for testing using the approach recommended in the Health Protection Branch Genotoxicity Guidelines. The battery of tests included the Salmonella/mammalian microsome mutation assay, the in vitro chromosomal aberration assay, and the bone-marrow micronucleus assay. The results indicate that 2 of the 3 chemicals, PA and 2NA, were clastogenic in vitro. Both PA and 2NA induced chromosomal aberrations in CHO cells in vitro with and without metabolic activation, while none induced reverse mutations detectable with the Salmonella/mammalian microsome assay. Because PA and 2NA were found to be in vitro clastogens, they also were tested in the mouse bone-marrow micronucleus assay. 2NA induced a small increase in micronuclei in males but not females. PA did not induce an increase in micronuclei.     

Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes

The diverse physiological functions exerted by the neuropeptide galanin may be regulated by multiple G protein-coupled receptor subtypes and intracellular signaling pathways. Three galanin receptor subtypes (GalRs) have been recently cloned, but the G protein coupling profiles of these receptors are not completely understood. We have generated GalR1- and GalR2-expressing Chinese hamster ovary (CHO) cell lines and systematically examined the potential for these two receptors to couple to the Gs, Gi, Go, and Gq proteins. Galanin did not stimulate an increase in cAMP levels in GalR1/CHO or GalR2/CHO cells, suggesting an inability of either receptor to couple to Gs. Galanin inhibited forskolin-stimulated cAMP production in GalR1/CHO cells by 70% and in GalR2/CHO cells by 30%, suggesting a strong coupling of GalR1 to Gi and a more modest coupling between GalR2 and Gi. GalR1 and GalR2 both mediated pertussis toxin-sensitive MAPK activity (2-3-fold). The stimulation mediated by GalR1 was inhibited by expression of the C-terminus of beta-adrenergic receptor kinase (beta ARKct), which specifically inhibits G beta gamma signaling, but was not affected by the protein kinase C (PKC) inhibitor, bis[indolylmaleimide], or cellular depletion of PKC. In contrast, GalR2-mediated MAPK activation was not affected by beta ARKct expression but was abolished by inhibition of PKC activity. The data demonstrate that GalR1 is coupled to a Gibetagamma signaling pathway to mediate MAPK activation. In contrast, GalR2 utilizes a distinct signaling pathway to mediate MAPK activation, which is consistent with Go-mediated MAPK activation in CHO cells. Galanin was unable to stimulate inositol phosphate (IP) accumulation in CHO or COS-7 cells expressing GalR1. In contrast, galanin stimulated a 7-fold increase in IP production in CHO or COS-7 cells expressing GalR2. The GalR2-mediated IP production was not affected by pertussis toxin, suggesting a linkage of GalR2 with Gq/G11. Thus, the GalR1 receptor appears to activate only the Gi pathway. By contrast, GalR2 is capable of stimulating signaling which is consistent with activation of Go, Gq/G11, and Gi. The differential signaling profiles and the tissue distribution patterns of GalR1 and GalR2 may underlie the functional spectra of galanin action mediated by these galanin receptors and regulate the diverse physiological functions of galanin.     

Effect of polycyclic aromatic hydrocarbons on somatic chromosomes of coal tar workers

The genotoxic effect of polycyclic aromatic hydrocarbons (PAHs) on somatic human chromosomes obtained from lymphocytes of 49 coal tar workers exposed to 26 micrograms/m3 benzo(a)pyrene, 16 mg/m3 benzene and 0.04 mg/m3 H2S in the ambient air, compared to equal numbers of matched controls breathing air containing 1 microgram/m3 benzo(a)pyrene, 1.5 mg/m3 benzene and 0.02 mg/m3 H2S, was investigated. The mitotic index (MI), chromosome aberrations (CAs), sister chromatid exchanges (SCEs) and satellite associations (SAs) were analysed. All the parameters showed a significant increase (p < 0.01 and 0.05) in the exposed individuals compared with the controls: viz MI, 4.59-7.92; CAs, 0.77-3.0; SCEs, 5.89-6.80; and SAs, 8.18-14.26. The occurrence of the DG type of satellite associations were highest and the 3D type lowest. The frequency of SCEs was highest in coal tar workers with an exposure period of 6-10 years. It is suggested that these results show PAH is genotoxic for humans.     

Transition from caspase-dependent to caspase-independent mechanisms at the onset of apoptotic execution

We have compared cytoplasmic extracts from chicken DU249 cells at various stages along the apoptotic pathway. Extracts from morphologically normal "committed stage" cells induce apoptotic morphology and DNA cleavage in substrate nuclei but require ongoing caspase activity to do so. In contrast, extracts from frankly apoptotic cells induce apoptotic events in added nuclei in a caspase-independent manner. Biochemical fractionation of these extracts reveals that a column fraction enriched in endogenous active caspases is unable to induce DNA fragmentation or chromatin condensation in substrate nuclei, whereas a caspase-depleted fraction induces both changes. Further characterization of the "execution phase" extracts revealed the presence of an ICAD/DFF45 (inhibitor of caspase-activated DNase/DNA fragmentation factor)- inhibitable nuclease resembling CAD, plus another activity that was required for the apoptotic chromatin condensation. Despite the presence of active caspases, committed stage extracts lacked these downstream activities, suggesting that the caspases and downstream factors are segregated from one another in vivo during the latent phase. These observations not only indicate that caspases act in an executive fashion, serving to activate downstream factors that disassemble the nucleus rather than disassembling it themselves, but they also suggest that activation of the downstream factors (rather than the caspases) is the critical event that occurs at the transition from the latent to active phase of apoptosis.     

Management of acid-peptic esophageal strictures

A new, nontoxic material extracted from Aspergillus fumigatus, is described with complement-depletion capacity both in vivo and in vitro. Extracts of the mycelium of A. fumigatus were found to haemolyse sheep red blood cells, to be lethal for mice and to convert C3 into its faster electrophoretic form when incubated with normal human serum. Conversion did not take place when purified C3 was used, suggesting that serum factors were required in the reaction. Adsorption of the extracts with activated carbon or heating at 100 degrees C for 30 min yielded detoxified materials devoid of both toxic and haemolytic properties but still capable of converting serum C3. In guinea pigs, administration of detoxified extracts was followed by complement depletion lasting not less than 48 h. Heating (100 degrees C for 30 min) extracts which had been previously detoxified by carbon adsorption resulted in a partial loss (40%) of their complement-inhibitory capacity indicating that part of the anticomplementarity of AFE was due to a heat-stable substance.     

Ligation of CD28 receptor by B7 induces formation of D-3 phosphoinositides in T lymphocytes independently of T cell receptor/CD3 activation

The co-stimulatory role of B7/CD28 interactions is important in promoting T cell activation. Very little is known about the intracellular events that follow CD28 engagement although recent evidence has implicated coupling of CD28 to a protein tyrosine kinase signal transduction pathway. In this study we have investigated the putative role of D-3 phosphoinositides as mediators of CD28 receptor signaling, since phosphoinositide (PI) 3-kinase, the enzyme responsible for D-3 phosphoinositide formation, is a known substrate for protein tyrosine kinases associated with certain T cell surface receptors such as CD4 and interleukin-2 receptor. The lipid products of PI 3-kinase activity have been suggested to play a role in mitogenic signaling and growth regulation in other cells. Chinese hamster ovary cells (CHO) previously transfected with B7 cDNA, induced time-dependent elevation above basal levels of phosphatidylinositol(3,4)-bisphosphate (PtdIns(3,4)P2) and PtdIns(3,4,5)P3, while parental CHO cells that did not express B7 had no effect on these lipids. Moreover, the elevation of these same lipids by CD3 ligation was potentiated in an additive manner by CHO-B7+ but not by CHO-B7- cells. CHO-B7+ and CHO-B7- cells did not activate phospholipase C as evidenced by their inability to modulate basal or CD3-induced changes in the levels of phosphatidic acid or D-4 and D-5 phosphoinositides. These data imply that PI 3-kinase but not phospholipase C, may be an important signal transduction molecule with respect to CD28-mediated co-stimulation and T cell activation following ligation by B7.     

Addition of a bone marrow "facilitating cell" population increases stem cell-derived cobblestone area formation in impaired long-term bone marrow culture stroma

Treatment of mouse bone marrow (BM) with rabbit anti-mouse brain serum (RAMBS) plus complement (C') depletes several cell types, including T cells and facilitating cells (FCs), that is, cells that facilitate engraftment of sorted allogeneic stem cells (SCs) in vivo. In the present study, treatment of BM with RAMBS+C' resulted in the depletion of approximately half of the late cobblestone area (CA)-forming stem cells as assayed on irradiated long-term bone marrow culture (LTBMC) stroma. In addition, LTBMC of RAMBS+C'-treated BM produced functionally impaired stroma with reduced ability to support CA formation by nontreated exogenous SCs. This stromal impairment was not due to depletion of TCRalphabeta T cells in the BM, because BM cultures from TCR alpha-chain knockout mice supported normal numbers of exogenous CAs. Because CD8+/TCR- cells are enriched for FCs, we tested the effect of adding these cells back to the treated BM prior to culture. The sorted FCs alone did not produce CAs, but did improve the ability of the impaired stroma to support late CA formation by sorted SCs. These studies provide a new model for dissecting the roles of different cellular components of BM in producing functional stroma that supports CA formation by SCs, and show that the number of CAs formed depends on the "quality" of the stroma as well as the number of SCs seeded. These findings further suggest that CD8+/TCR- BM cells may be important for the establishment of functional stroma.     

Flow cytometric analyses of the specific activation of peripheral blood mononuclear cells from healthy donors after in vitro stimulation with a fermented mistletoe extract and mistletoe lectins

Immunostimulatory properties of mistletoe extracts derived from Viscum album L. (VAL) are well described, demonstrating activation especially of T, T-helper cells and monocytes/macrophages. In order to characterise in detail the communication between different cell populations, we studied mistletoe-induced expression of co-stimulatory signals and their ligands by flow cytometry. Peripheral blood mononuclear cells (PBMC) from 15 healthy controls were incubated for 7 days with a fermented VAL extract. VAL significantly upregulated the expression of the co-stimulatory molecule B7.1 (CD80) on monocytes/macrophages, but not B7.2 (CD86). No significant changes in the expression of either molecules on B cells could be found, suggesting that only monocytes/macrophages act as antigen presenting cells (APCs) in this in vitro system. Purified mistletoe lectins, components of most VAL extracts were also analysed, but did not induce similar responses of monocytes/macrophages. The receptor for B7 molecules, CD28, but not CTLA-4 (CD152), was also found to be significantly enhanced on CD4+ cells after VAL simulation. There was no evidence for activation of a B cell response via the CD40/CD40L pathway. Our data support the concept that stimulation by VAL extracts induces a specific T-helper cell reaction with monocytes/macrophages acting as APCs and purified lectins do not exert the same effects.     

Chinese hamster ovary cells expressing a novel type of acetylated low density lipoprotein receptor. Isolation and characterization

Macrophage scavenger receptors mediate the recognition of a wide range of negatively charged macromolecules including acetylated low density lipoproteins (AcLDL). Chinese hamster ovary (CHO) cells were cultured in the presence of increasing concentrations of simvastatin, a cholesterol biosynthesis inhibitor, and AcLDL as the sole source of exogenous lipoproteins. The cells surviving under these conditions specifically bound 125I-labeled AcLDL with high affinity and degraded them via an endocytic pathway. Unexpectedly, the association and degradation of 125I-labeled AcLDL by these CHO cells were not inhibited by dextran sulfate, fucoidan, and polyinosinic acid, competitors of macrophage scavenger receptors, but were completely inhibited by maleylated bovine serum albumin. Furthermore, these cells effectively took up negatively charged liposomes containing acidic phospholipids such as phosphatidylserine and phosphatidic acid, whereas CHO cells expressing macrophage scavenger receptors did not. AcLDL and negatively charged liposomes were cross-competed with each other. Northern blot analysis using the cDNA for the macrophage scavenger receptor revealed that these CHO cells did not express this receptor. From these observations, we conclude that the isolated CHO cells express a novel type of AcLDL receptor, which is distinct from macrophage scavenger receptors with respect to ligand specificity and competitor sensitivity.     

Alpha 1,6-linked fucose affects the expression and stability of polysialic acid-carrying glycoproteins in Chinese hamster ovary cells

To determine the effect of alpha1,6-linked fucose modification of N-glycans on the expression of polysialic acids (PSAs), the expression of PSAs in a fucose-lacking mutant of Chinese hamster ovary (CHO) cells, Lec13, was compared with that in CHO K1 cells. PSA synthase activity in these cells and the antennary structures of N-glycans associated with the neural adhesion molecule (NCAM), which is a major PSA-carrying glycoprotein, did not differ between the two types of cells. Metabolic labeling of cells with [3H]glucosamine for 48 h followed by immunoprecipitation with anti-PSA monoclonal antibodies revealed that the amount of labeled PSA-carrying glycoproteins obtained from Lec13 cells was 10-times less than that from K1 cells, although the incorporation of [3H]glucosamine into total extracts and NCAM was almost the same. In contrast, when cells were pulse labeled with [35S]methionine followed by a 1 h chase, there was not such a great difference in PSA-carrying protein synthesis between K1 and Lec13 cells. However, during a prolonged chase period, PSA-carrying proteins rapidly decreased in Lec13 cells, whereas those in K1 cells did not change. The degradation of PSA-carrying glycoproteins in Lec13 cells was partly prevented when the cells were grown in fucose-containing medium. Therefore, fucose modification of core N-glycans may affect the efficient expression of PSAs through the intracellular stability of PSA-carrying glycoproteins.     

Interaction of Oct-1 and automodification domain of poly(ADP-ribose) synthetase

Ataxia telangiectasia (AT) cells display a profound sensitivity to ionizing radiation, exhibiting more frequent chromosomal breaks, increased micronuclei formation and abnormal DNA repair kinetics following exposure. Despite the recent cloning of the ATM gene there remains a need for a simple and rapid means of discriminating AT heterozygotes from normal individuals. Caffeine (1,3,7-trimethyl xanthine), known to inhibit the repair of double-strand DNA breaks following ionizing radiation, increases the frequency of radiation induced chromosomal breaks in normal cells. Here we report that caffeine potentiates the induction of chromosomal breaks in G2 arrested AT heterozygote and normal lymphoblastoid cells, but not in homozygous AT lymphoblastoid cells. This observation parallels the findings reported by others that caffeine fails to potentiate the effect of ionizing radiation in radiation-sensitive yeast strains and radiation sensitive CHO cells. It also suggests that caffeine may somehow mimic the effect of the ATM gene product in normal cells. We also report that caffeine is unlikely to be useful in helping to discriminate AT heterozygotes from normal individuals.     

Evidence for linkage between the development of asthma in childhood and the T-cell receptor beta chain gene in Japanese

GTPase-activating proteins (GAPs) stimulate the hydrolysis of GTP bound to small G-proteins and regulate the signal transduction pathway. Changes in the expression of p21-Ras p120-GAP induced by growth factor treatment were examined in cultured Chinese hamster ovary (CHO) and human choriocarcinoma (BeWo) cells. Expression of p120-GAP and GAP activity were measured. Fetal bovine serum induced a significant increased level of GAP in CHO cells, but did not increase GAP in BeWo cells. The results suggest that growth factors affect Ras GAP expression in CHO cells, while they do not in other cells such as BeWo cells.     

Induction of sister-chromatid exchanges by styrene analogues in cultured human lymphocytes

Styrene and 11 styrene analogues were tested for their ability to induce sister-chromatid exchanges (SCEs) by a 48-h treatment in human whole-blood lymphocyte cultures (72 h). Styrene, its methyl-substituted derivatives (substituted at 2-, 3-, 4-, 3,5- or beta-position) and two styrene oxides (3,5-dimethylstyrene-7,8-oxide and 4-nitrostyrene-7,8-oxide) induced a distinct dose-dependent increase in SCEs. Also, 4-methylstyrene-7,8-oxide and alpha-methylstyrene showed a positive effect, but were not able to double the mean number of SCEs/cell at the concentration ranges available. Ethylbenzene had a marginal effect on SCEs at the highest dose tested. 2-Phenylethanol did not increase SCEs. The results indicate that styrene and methylstyrenes are converted into reactive metabolites in the whole-blood lymphocyte cultures. The negative or weak effects of styrene analogues without a double bond in the side-chain (ethylbenzene and 2-phenylethanol) suggest that the reactive metabolites are derived from the conversion of the vinyl group and are styrene-7,8-oxides.     

Lysophosphatidic acid-induced association of SHP-2 with SHPS-1: roles of RHO, FAK, and a SRC family kinase

SHPS-1 is an approximately 120 kDa glycosylated receptor like protein that contains three immunoglobulin-like domains in its extracellular region as well as four potential tyrosine phosphorylation and SRC homology 2 (SH2) domain binding sites in its cytoplasmic region. Lysophosphatidic acid (LPA) stimulated the rapid tyrosine phosphorylation of SHPS-1 and its subsequent association with SHP-2, a protein tyrosine phosphatase containing SH2 domains in Rat-1 fibroblasts. LAP-induced tyrosine phosphorylation of SHPS-1 was inhibited by Clostridium botulinum C3 exoenzyme (which inactivates RHO) but not by pertussis toxin. The protein kinase C activator phorbol ester, 12-O-tetradecanoylphorbol 13-acetate (TPA) also stimulated tyrosine phosphorylation of SHPS-1; however, down-regulation of protein kinase C by prolonged exposure of cells to TPA did not affect LAP-induced tyrosine phosphorylation of SHPS-1. LPA-induced tyrosine phosphorylation of SHPS-1 was markedly reduced in either focal adhesion kinase (FAK)-deficient mouse cells or CHO cells overexpressing the tyrosine kinase CSK. Overexpression of a catalytically inactivate SHP-2 markedly inhibited MAP kinase activation in response to low concentrations of LPA in CHO cells, whereas overexpression of a wild-type SHPS-1 did enhance this effect of LPA. Furthermore, MAP kinase activation in response to a low concentration of LPA was inhibited by botulinum C3 exoenzyme. These results indicate that LPA-induced tyrosine phosphorylation of SHPS-1 and its association with SHP-2 may be mediated by a RHO-dependent pathway that includes FAK and a SRC family kinase. Thus, in addition to its role in receptor tyrosine kinase-mediated MAP kinase activation, the formation of a complex between SHPS-1 and SHP-2 may, in part, play an important role in the activation of MAP kinase in response to low concentrations of LPA.     

Methylthiol:coenzyme M methyltransferase from Methanosarcina barkeri, an enzyme of methanogenesis from dimethylsulfide and methylmercaptopropionate

During growth on acetate, Methanosarcina barkeri expresses catabolic enzymes for other methanogenic substrates such as monomethylamine. The range of substrates used by cells grown on acetate was further explored, and it was found that cells grown on acetate also converted dimethylsulfide (DMS) and methylmercaptopropionate (MMPA) to methane. Cells or extracts of cells grown on trimethylamine or methanol did not utilize either DMS or MMPA. During growth on acetate, cultures demethylated MMPA, producing methane and mercaptopropionate. Extracts of acetate-grown cells possessed DMS- and MMPA-dependent coenzyme M (CoM) methylation activities. The activity peaks of CoM methylation with either DMS or MMPA coeluted upon gel permeation chromatography of extracts of acetate-grown cells consistent with an apparent molecular mass of 470 kDa. A 480-kDa corrinoid protein, previously demonstrated to be a CoM methylase but otherwise of unknown physiological function, was found to methylate CoM with either DMS or MMPA. MMPA was demethylated by the purified 480-kDa CoM methylase, consuming 1 mol of CoM and producing 1 mol of mercaptopropionate. DMS was demethylated by the purified protein, consuming 1 mol of CoM and producing 1 mol of methanethiol. The methylthiol:CoM methyltransferase reaction could be initiated only with the enzyme-bound corrinoid in the methylated state. CoM could demethylate, and DMS and MMPA could remethylate, the corrinoid cofactor. The monomethylamine corrinoid protein and the A isozyme of methylcobamide:CoM methyltransferase (proteins homologous to the two subunits comprising the 480-kDa CoM methylase) did not catalyze CoM methylation with methylated thiols. These results indicate that the 480-kDa corrinoid protein functions as a CoM methylase during methanogenesis from DMS or MMPA.     

A role for PACE4 in the proteolytic activation of anthrax toxin protective antigen

Several bacterial protein toxins require activation by eukaryotic proteases. Previous studies have shown that anthrax toxin protective antigen (PA), Pseudomonas exotoxin A (PE), and diphtheria toxin (DT) are cleaved by furin C-terminal to the sequences RKKR, RQPR, and RVRR, respectively. Because furin-deficient cells retain some sensitivity to PA and DT, it is evident that other cellular proteases can activate these toxins. Whereas furin has been shown to require arginine residues at positions -1 and -4 for substrate recognition, another protease with an activity which could substitute for furin in toxin activation, the furin-related protease PACE4, requires basic residues in the -1, -2, and -4 positions of the substrate sequence. To examine the relative roles of furin and PACE4 in toxin activation, we used furin-deficient CHO cells (FD11 cells) transfected with either the furin (FD11/furin cells) or PACE4 (FD11/PACE4 cells) gene. Mutant PA proteins containing the cleavage sequence RAAR or KR were cytotoxic toward cells expressing only PACE4. In vitro cleavage data demonstrated that PACE4 can recognize RAAR and, to a much lesser extent, KR and RR. When extracts from PACE4-transfected cells were used as a source of proteases, PACE4 had minimal activity, indicating that it had been partially inactivated or did not remain associated with the cell membranes. Cleavage of iodinated PA containing the sequence RKKR or RAAR was detected on the surface of all cell types tested, but cleavage of a dibasic sequence was detected only intracellularly and only in cells that expressed furin or PACE4. The data provide evidence that PACE4 is present at the exterior of cells, that it plays a role in the proteolytic activation of anthrax toxin PA, and that PACE4 can activate substrates at the sequence RAAR or KR.