Phase II study of single-agent bosutinib, a Src/Abl tyrosine kinase inhibitor, in patients with locally advanced or metastatic breast cancer pretreated with chemotherapy.

BACKGROUND: This phase II study evaluated single-agent bosutinib in pretreated patients with locally advanced or metastatic breast cancer. PATIENTS AND METHODS: Patients received oral bosutinib 400 mg/day. The primary end point was the progression-free survival (PFS) rate at 16 weeks. Secondary end points included objective response rate, clinical benefit rate, 2-year overall survival rate, safety, and changes in levels of bone resorption/formation biomarkers. RESULTS: Seventy-three patients were enrolled and treated. Median time from diagnosis of metastatic disease to initiation of bosutinib treatment was 24.5 months. For the intent-to-treat population, the PFS rate at 16 weeks was 39.6%. Unexpectedly, all responding patients (n = 4) were hormone receptor positive. The clinical benefit rate was 27.4%. The 2-year overall survival rate was 26.4%. The main toxic effects were diarrhea (66%), nausea (55%), and vomiting (47%). Grade 3-4 laboratory aminotransferase elevations occurred in 14 (19%) patients. Myelosuppression was minimal. No consistent changes in the levels of bone resorption/formation biomarkers were seen. CONCLUSIONS: Bosutinib showed promising efficacy in prolonging time to progression in chemotherapy-pretreated patients with locally advanced or metastatic breast cancer. Bosutinib was generally well tolerated, with a safety profile different from that of the Src/Abl tyrosine kinase inhibitor dasatinib in a similar patient population.

Apoptosis induced by a chimeric Fas/FLICE receptor: lack of requirement for Fas- or FADD-binding proteins.

Current models for Fas (CD95)-mediated apoptosis suggest that FLICE/caspase-8 is recruited and activated, which results in cell death. However, the role of additional molecules in Fas signaling and FLICE activation is not clear. A chimeric Fas/FLICE (F/F) receptor, containing the extracellular/transmembrane portion of Fas and the caspase region of FLICE, mediated anti-Fas apoptosis. FLICE protease subunits were generated from the F/F precursor. Killing induced by Fas, but not F/F, was blocked by a dominant negative FADD. Apoptosis triggered through Fas and F/F was inhibited by coexpression of CrmA and p35, but not Bcl-xL. F/F bypassed Fas resistance in COS-7 cells and blocking by the death effector domain (DED)-containing viral protein MC159. These results show that: 1) F/F induces cell death, indicating that FLICE activation is sufficient for apoptosis and does not require additional Fas- or FADD-binding proteins; and 2) F/F bypasses proximal defects in Fas signaling that prevent FLICE recruitment or activation.

Do CTL kill target cells by inducing apoptosis?

Inhibitors of ICE-family proteases (caspases) block many examples of apoptotic cell death in vivo and in vitro, including multiple apoptotic stimuli for T lymphocytes. We have tested whether cell death induced by cytotoxic T lymphocytes was also blocked by caspase inhibitors. We found that the rapid apoptotic target cell death induced by Fas ligand-bearing CTL using the target Fas death pathway was efficiently blocked by caspase inhibitors. In contrast, target lysis induced by the CTL granule exocytosis pathway is not detectably blocked by such inhibitors, although the accompanying apoptotic nuclear damage is efficiently blocked. Thus caspase inhibitors prevent the hallmark phenotype of apoptosis without measurably affecting target cell death as evidenced by lysis.

Target cell lysis by CTL granule exocytosis is independent of ICE/Ced-3 family proteases.

Activation of ICE/Ced-3 family proteases (caspases) has been proposed to mediate both the granule exocytosis and Fas-Fas ligand pathways of rapid target cell death by cytotoxic T lymphocytes. In agreement with this model, two peptide fluoromethyl ketone caspase inhibitors and baculovirus p35 blocked apoptotic nuclear damage and target cell lysis by the CTL-mediated Fas-Fas ligand pathway. The peptide caspase inhibitors also blocked drug-induced apoptotic cell death in tumor cells. In contrast, the caspase inhibitors blocked CTL granule exocytosis-induced target apoptotic nuclear damage, but did not inhibit target lysis. These results are consistent with recent demonstrations that granzyme B can activate caspases leading to apoptotic nuclear damage, but show that target cell lysis by CTL granule exocytosis occurs by a caspase-independent pathway.

Lack of a role for Jun kinase and AP-1 in Fas-induced apoptosis.

Cross-linking of Fas (CD95) induces apoptosis, a response that has been reported to depend upon the Ras activation pathway. Since many examples of apoptosis have been reported to involve AP-1 and/or the AP-1-activation pathway. Since many examples of apoptosis have been reported to involve AP-1 and/or the AP-1-activating enzyme Jun kinase (JNK), downstream effectors of Ras or Ras-like small GTP-binding proteins, we evaluated the role of these molecules in Fas-mediated apoptosis. Although cross-linking of Fas on Jurkat T cells did result in JNK activation, increased activity was observed relatively late, being detectable only after 60 min of stimulation. Expression of a dominant negative form of SEK1 that blocked Fas-mediated induction of JNK activity had no effect on Fas-mediated apoptosis. Furthermore, maximally effective concentrations of anti-Fas did not cause JNK activation if apoptosis was blocked by a cysteine protease inhibitor, suggesting that under these conditions, activation of JNK may be secondary to the stress of apoptosis rather than a direct result of Fas engagement. Despite the activation of JNK, there was no induction of AP-1 activity as determined by gel shift assay or induction of an AP-1-responsive reporter. The lack of a requirement for AP-1 induction in Fas-mediated death was further substantiated with Jurkat cells that were stably transfected with a dominant negative cJun, TAM-67. While TAM-67 effectively prevented AP-1-dependent transcription of both the interleukin-2 and cJun genes, it had no effect on Fas-induced cell death, even at limiting levels of Fas signaling. Thus, induction of JNK activity in Jurkat cells by ligation of Fas at levels sufficient to cause cell death is likely a result, rather than a cause, of the apoptotic response, and AP-1 function is not required for Fas-induced apoptosis.

Apoptosis signaling pathways in normal T cells: differential activity of Bcl-2 and IL-1beta-converting enzyme family protease inhibitors on glucocorticoid- and Fas-mediated cytotoxicity.

Fas-mediated apoptosis plays an important role in regulating the immune response in peripheral T cells. Restimulation of T cell blasts up-regulates Fas and Fas ligand expression, with subsequent interaction leading to cell death. Overexpression of Bcl-2 in tumor cells blocks apoptosis induced by many stimuli, but inhibition of Fas-mediated killing has not been consistently observed. To examine the behavior of Bcl-2 in normal cells, T cell blasts were transiently transfected with Bcl-2 and related gene products to determine the effect on apoptotic signaling. Transient overexpression of Bcl-2 in mouse and human T cell blasts did not block Fas-mediated apoptosis, whereas etoposide- and glucocorticoid-induced cytotoxicity was potently inhibited. Expression of Bcl-xL and adenovirus E1B 19K did not interfere with anti-Fas killing. In contrast, interleukin-1beta-converting enzyme family protease inhibitors Ac-DEVD-CHO and CrmA blocked Fas-mediated apoptosis. These results suggest that peripheral T cells use distinct apoptosis signaling pathways with differential sensitivity to Bcl-2 and interleukin-1beta-converting enzyme family protease inhibitors. Since T cells normally express Bcl-2 and Bcl-xL following activation, their inability to block Fas-mediated apoptosis may allow for the elimination of self-reactive cells and the appropriate regulation of immune responses.

Recombinant mouse Bcl-2(1-203). Two domains connected by a long protease-sensitive linker.

Bcl-2 is a cytoplasmic integral membrane protein with potent anti-apoptotic activity but whose mechanism of action is poorly understood. The purpose of this paper was to obtain large amounts of soluble Bcl-2 protein for structural and functional studies. Mouse Bcl-2(1-203) (missing the COOH-terminal hydrophobic tail) was produced in bacterial inclusion bodies, solubilized in guanidine, and refolded by dialysis. The resulting protein was monomeric in nondenaturing solution and was active in protecting mouse T hybridoma cells from glucocorticoid-induced apoptosis. Refolded Bcl-2(1-203) showed no tendency to homodimerize by gel filtration or analytical ultracentrifugation. Limited proteolysis experiments identified a region between the BH3 and BH4 homology domains of Bcl-2(1-203) which was extremely susceptible to digestion by several common proteases, but not by a cell extract known to contain CPP-32-like (interleukin-1beta-converting enzyme family) protease activity. The protease-sensitive sites were located within a 50-residue stretch that contained most of the nonconserved and proline residues of Bcl-2(1-203). Trypsin-cleaved Bcl-2(1-203) eluted in the same position as the undigested protein on gel filtration in nondenaturing solution, indicating that the two portions of the molecule connected by the protease-sensitive region associate stably and noncovalently. The solution properties of Bcl-2(1-203) suggest that it consists of two noncovalently associated domains connected by a long protease-sensitive linker and that its structure is similar to that of Bcl-xL, which has been determined by x-ray and NMR analysis.

Bcl-2 blocks glucocorticoid- but not Fas- or activation-induced apoptosis in a T cell hybridoma.

Overexpression of Bcl-2 can prevent or markedly delay cell death induced by a variety of apoptotic stimuli. Although Fas and Fas ligand (FasL) interactions play a major role in the elimination of self-reactive T cells in the periphery, inhibition of Fas-mediated killing by Bcl-2 has not been consistently observed. The mouse T hybridoma 2B4.11 (2B4) has been a useful model to study glucocorticoid- and activation-induced apoptosis, which is mediated through Fas and FasL. Using both stable transfectants and transient transfections, overexpression of Bcl-2 or Bcl-xL readily blocked glucocorticoid-induced but not activation-induced apoptosis of 2B4 cells. Bcl-2 expression did not inhibit Fas-mediated cytotoxicity triggered by cells expressing FasL or by the transient transfection of human Fas. Similarly, overexpression of Bcl-2 in the mouse T hybridoma A1.1 did not block activation-induced/Fas-mediated apoptosis. In Jurkat cells, however, expression of Bcl-2 partially inhibited anti-Fas-induced cell death. A Bcl-2-related protein that can interfere with anti-Fas killing, the adenoviral E1B 19K, also did not block activation-induced/Fas-mediated apoptosis in 2B4 cells. In contrast, expression of CrmA, a cowpox virus protein that inhibits ICE-like protease activity, blocked activation-induced apoptosis in 2B4 cells but had little effect on Dex-mediated cytotoxicity. These results show that: 1) Bcl-2 can have strikingly different anti-cell death activity in the same cell depending upon the apoptotic stimulus, 2) distinct apoptosis signaling pathways may exist with differential sensitivity to Bcl-2 and ICE-like protease inhibitors.

A bispecific antibody prolongs survival in mice bearing lung metastases of syngeneic mammary adenocarcinoma.

In the present study we tested whether T cells retargeted with a bispecific antibody (bsAb) could block the growth of lung metastases of syngeneic mammary adenocarcinoma in immunocompetent mice. BALB/c mice were injected i.v. with tumor and i.p. with a genetically engineered bispecific F(ab')2 [bs(Fab')2] having specificity for murine CD3 epsilon chain and for the gp52 mouse mammary tumor viral glycoprotein, which is expressed on the tumor cells. The bs(Fab')2 was physically stable in blood and serum, was removed from the body with a half-time of 12-15 h, and accumulated in lymphoid tissue where it bound to T cells. We show that treatment of tumor bearing mice with the bs(Fab')2 significantly prolonged their survival relative to untreated controls. Two other genetically engineered bs(Fab')2s having specificity for murine CD3 epsilon chain and irrelevant antigens did not inhibit tumor growth. In addition, survival was not affected by bsAb therapy using a variant tumor cell line that expressed low levels of the gp52 target antigen. Inhibition of tumor growth was even more evident by histologic analysis. Treatment with the relevant bs(Fab')2 resulted in a marked reduction of tumor burden in lung sections taken on days 7, 9 and 11. This is the first report demonstrating that a bsAb can inhibit the growth of syngeneic solid tumor metastases in mice without addition of T cell activators.

Bispecific antibodies retarget murine T cell cytotoxicity against syngeneic breast cancer in vitro and in vivo.

Bispecific antibodies with specificity for CD3 and a tumor antigen can redirect cytolytic T cells to kill tumor targets, regardless of their natural specificity. To assess the clinical potential of bispecific antibodies for treatment of human cancers we have, in the present study, adapted a totally synergeic mouse model to the targeting of mouse T cells against mouse tumors in immunocompetent mice. We show that gp52 of the mouse mammary tumor virus (MTV) can serve as a tumor-specific antigen for redirected cellular cytotoxicity. Chemically crosslinked and genetically engineered bispecific antibodies with specificities for gp52 and murine CD3 epsilon-chain induced activated mouse T cells to specifically lyse mouse mammary tumor cells from cultured lines and primary tumors from C3H-MTV+ mice. Retargeted T cells also blocked the growth of mammary tumors in vitro as well as their growth in syngeneic mice. These findings identify murine MTV-induced mammary adenocarcinomas as a solid-tumor, animal model for retargeting T cells with bispecific antibodies against syngeneic breast cancer.

A simple assay for examining the effect of transiently expressed genes on programmed cell death.

Programmed cell death (PCD) has been observed in a wide variety of cell types in response to physiologic signals or types of stress. How these stimuli trigger PCD, and whether there is a common PCD signal transduction pathway, is not clear. As more genes are described that may participate in or regulate PCD, an assay system in which gene products can easily be introduced and/or modulated would be of great value. To avoid the generation and screening of multiple individual stable cell transfectants, a simple transient transfection death assay has been developed. 2B4.11, a murine T cell hybridoma, was transfected by electroporation with a constitutively active beta-galactosidase reporter gene and the cells were incubated in culture medium or with a PCD-inducing stimulus. The amount of beta-galactosidase activity remaining in the intact cells at the end of the culture period represented only viable transfected cells. Bcl-2 was chosen to examine whether this system would be useful to study the effect of transiently transfected genes since it blocks PCD in a number of experimental systems. Consistent with data obtained using stable transfectants, transient expression of Bcl-2 in 2B4.11 completely protected cells from glucocorticoid- and cytotoxic agent-induced PCD. This protection from death was confirmed at the individual cell level by the transient co-expression of a class I Ld surface antigen and flow cytometric analysis. Some of the advantages of the transient transfection death assay described here are; (1) the simple and sensitive beta-galactosidase assay, (2) the rapidity of the assay, (3) the ability to perform conventional viability assays to monitor treatment-induced cytotoxicity, (4) multiple gene products can be tested alone, and in combination, (5) antisense or dominant negative approaches can be used, and (6) the adaptability of this assay system to other cell types, transfection techniques, or reporter and expression vectors. The transient transfection death assay should make it easier to identify and order important steps in the PCD signal transduction pathways.

Dominant negative mutant of c-Jun inhibits NF-AT transcriptional activity and prevents IL-2 gene transcription.

Expression of the transcription complex AP-1, composed of Jun and Fos family members, can be induced by a variety of stimuli. In lymphocytes, AP-1 transcriptional activity increases after TCR ligation and plays an important role in T cell activation events such as lymphokine secretion. To explore the requirements for AP-1 in IL-2 production, the AP-1 complex was targeted with a dominant negative mutant c-Jun protein, TAM-67, from which the transactivation domain has been deleted. In transient transfections of Jurkat cells, TAM-67 efficiently inhibited endogenous AP-1 transcriptional activity and blocked the activity of a reporter construct containing the 5' regulatory region of the IL-2 gene. TAM-67 also inhibited the transcriptional activity of nuclear factor-AT (NF-AT), whereas the NF-kappa B, NF-IL-2A, and the proximal TRE-like sites were relatively unaffected. The use of this dominant negative transcription factor suggests that: 1) transactivation-defective nuclear factors represent a novel approach to study the functional consequences of nuclear protein interactions on gene transcription; 2) the proximal TRE-like site from the IL-2 promoter is different from the consensus TRE; and 3) AP-1 plays an important role in the transcriptional activation mediated by the NF-AT binding complex.

Mechanism of macrophage-mediated cytotoxicity: production of a soluble cytotoxic factor.

Peritoneal macrophages from C3HeB/FeJ mice became cytotoxic for 6C3HED lymphosarcoma cells, P815 mastocytoma cells, and L-929 fibroblasts when treated with the calcium ionophore, A23187, at concentrations ranging from 1.0 to 20 microM. The effect of A23187 on other activation processes was also tested. It was found that A23187 and lipopolysaccharide (LPS) acted synergistically, but no consistent synergy with macrophage-activating factor (MAF) was observed. Cytotoxic activity (M phi-CF) was found in cellfree supernatants from M phi activated by A23187 or LPS. Furthermore, these two activating agents synergize in the production of M phi-CF. The cytotoxic activity of the crude material was not blocked by catalase or protease inhibitors. Fractionation of supernatants by high pressure liquid chromatography has shown that there was a peak of cytotoxic activity with a m.w. of approximately 45,000. Interestingly, L-929 cells were 30-fold more sensitive to M phi-CF than a lymphotoxin-resistant subline of L-929.

Macrophage-mediated cytotoxicity: role of a soluble macrophage cytotoxic factor similar to lymphotoxin and tumor necrosis factor.

Guinea pig peritoneal macrophages, when activated for cytotoxicity by the calcium ionophore A23187 or lipopolysaccharide, produce a cytotoxic factor [macrophage cytotoxic factor (M phi-CF)] that is not blocked by catalase or protease inhibitors. Fractionation of culture supernates containing M phi-CF by gel filtration revealed one peak of cytotoxic activity of Mr approximately 45,000, the same as guinea pig lymphotoxin (LT). Antiserum prepared against purified guinea pig LT completely neutralized the cytotoxic activity of M phi-CF. In addition, the cytotoxic factor in guinea pig tumor necrosis serum was found to have a Mr of 45,000 and was neutralized by anti-LT. Thus, M phi-CF is physicochemically and immunochemically similar to LT and tumor necrosis factor, if not identical. To investigate the role of M phi-CF in macrophage-mediated cytotoxicity, anti-LT was added to A23187- or lipopolysaccharide-activated macrophages before addition of L-929 target cells. In 10 of 16 experiments, the inhibition of macrophage-mediated cytotoxicity was 100%. In the others, cytotoxicity was blocked partially, the lowest inhibition being 49%. The effectiveness of inhibition appeared to be inversely related to the intensity of macrophage activation. These results indicate that M phi-CF plays a significant role in macrophage-mediated cytotoxicity but involvement of another mechanism cannot be excluded.

H-2 restriction: independent recognition of H-2 and foreign antigen by a single receptor.

We describe two situations in which the recognition of hapten can compensate for the lack of recognition of appropriate H-2 gene products in hapten-specific, H-2 restricted, T lymphocyte-mediated cytolysis. First, we show that although recognition of appropriate H-2 gene products is essential for the lysis of target cells bearing a low hapten density, significant hapten-specific lysis of H-2 inappropriate target cells is observed at high levels of target cell derivatization. Secondly, we show that hapten-conjugated anti-H-2 antibody inhibits cytolysis poorly even though its binding to target cell H-2 antigens is equivalent to that of underivatized antibody. These results suggest that hapten and H-2 are recognized independently and are therefore inconsistent with the altered-self model. Although our data do not exclude the dual-recognition model, we prefer to interpret them within the framework of a single-receptor model in which hapten and H-2 are recognized independently by receptors of identical idiotype on the T cell. We postulate that the affinity of these receptors for the relevant H-2 gene product is low enough so that the T cell is not activated by encounters with normal-self cells expressing that H-2 gene product. However, when self cells express in addition a foreign antigen that can also be recognized by the same receptor, then the force of T cell-target cell interaction may be increased sufficiently to activate T cell effector function.

Effect of trypsin on mouse mammary tumor virus.

Undisrupted mouse mammary tumor virus (MuMTV) derived from the milk of of RIII mice has been analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electron microscopy after treatment with insolubilized trypsin. No alterations were found in viral fine structure by either freeze-etch or negative-stain electron microscopy. No alterations were found in the ability of trypsinized virus to compete in a radioimmune assay for viral antigens. Infectivity experiments indicate no significant differences in the ability of treated virus to infect C57Bl mice. However, significant differences were observed in polypeptide composition. The intensely periodic acid-Schiff-positive band, gp140, was shown by galactose oxidase-borotritide labeling to be degraded into a fragment of 125,000 molecular weight. The major glycoprotein, gp55, was split into fragments of 36,000 and 23,000 molecular weight, both of which stained with periodic acid-Schiff stain. Gp68 was removed from the virus. Experiments with purified, iodinated gp55 showed that the trypsin-induced fragments of gp55 were immunologically active. We conclude that: (i) certain glycoproteins at the surface of MuMTV are accessible to an insoluble form of trypsin, (ii) the trypsin causes a nick in the polypeptide chain without affecting the configuration of the molecule; (iii) the nicked molecules remain bound to the virus; and (iv) the presence of these nicked molecules does not interfere with the biological or antigenic expression of virus function.