Irmela Jeremias

TRAIL/Apo‐2‐ligand‐induced apoptosis in human T cells

Members of the tumor necrosis factor (TNF) family such as CD95 (APO‐1/Fas) ligand (L) trigger apoptosis in lymphoid cells. Recently, a new member of apoptosis‐inducing ligands, TRAIL (TNF‐related‐apoptosis‐inducing‐ligand)/Apo‐2 ligand, was identified that might act in a similar way. We compared TRAIL and CD95L‐induced apoptosis in human lymphoid cells. Expression of TRAIL was found in CD4+ and CD8+ T cells following activation, suggesting that TRAIL participates in T cell‐mediated induction of apoptosis. Similar to CD95L, TRAIL‐induced apoptosis in target cells is mediated by activation of caspases (ICE/Ced‐3 proteases). However, different human lymphoid cell lines and peripheral T cells differ in sensitivity towards induction of apoptosis by TRAIL and CD95L. In addition, T cells are highly sensitive towards CD95L‐induced apoptosis after prolonged activation in vitro, but remain completely resistant to TRAIL‐induced apoptosis. In contrast, T cells from HIV‐1‐infected patients previously shown to exhibit increased CD95 sensitivity are even more susceptible towards TRAIL‐induced cell death. These data suggest that TRAIL might participate in CD95‐independent apoptosis of lymphoid cells and might be involved in deregulated apoptosis in diseases such as leukemias and HIV‐1 infection.

TRAIL induced survival and proliferation in cancer cells resistant towards trail-induced apoptosis mediated by NF-κB

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is a potent inducer of apoptosis in cancer cells. Examining primary cells of children with untreated acute leukemia, TRAIL induced apoptosis in 50% of cells, but to our surprise attenuated spontaneous apoptosis in the remaining samples or, most importantly, even mediated proliferation. We therefore examined tumor cell lines of leukemic and nonleukemic origin with apoptosis resistance towards TRAIL because of absent Caspase-8 or dysfunctional FADD. In all cell lines tested, TRAIL treatment increased cell numbers in average to 163% within 4 days and accelerated doubling time from 24 to 19 h. TRAIL-mediated proliferation was completely abrogated by blockade of NF-κB activation using proteasome inhibitors or in RIP-negative, IKKγ-negative cells or in cells overexpressing dominant-negative IκBα. Our data describe the biological significance of TRAIL-mediated activation of NF-κB in cancer cells resistant to TRAIL-mediated apoptosis: TRAIL leads to an increase in tumor cell count by a prosurvival and possibly mitogenic function. Given the promising therapeutic potential of TRAIL as a novel anticancer drug, TRAIL-mediated survival or proliferation of target cells may restrict its use to apoptosis-sensitive tumors.

Interleukin-1 receptor antagonist and interleukin-6 for early diagnosis of neonatal sepsis 2 days before clinical manifestation.

BACKGROUND Neonatal sepsis is a common and life-threatening disorder, particularly among preterm infants. Early initiation of antibiotic therapy is frequently delayed because the first clinical signs of sepsis are non-specific and there are no reliable early laboratory indicators. We investigated the time course of expression and the prognostic power of the early inflammatory mediators interleukin-1 receptor antagonist (IL-1ra), interleukin-6 (IL-6), and circulating intercellular adhesion molecule-1 (cICAM-1) before clinical diagnosis of sepsis. METHODS In a prospective multicentre study, we monitored 182 very-low-birthweight infants in six intensive-care units for occurrence of sepsis. During routine or clinically indicated blood sampling, an additional sample was collected for measurement of IL-1ra, IL-6, cICAM-1, and C-reactive protein (CRP). Infants were grouped into those with proven sepsis, no infection, or unclassified. The mean study duration was 34 days. Whenever sepsis occurred, a study period of 10 days was defined: day 0 was the day of clinical diagnosis of sepsis; days -4 to -1 were the 4 days before diagnosis; days +1 to +5 were the 5 days after. We compared the concentrations of the immune mediators during the 10-day study period with group-specific baseline values from before day -4. FINDINGS 101 infants were included in the analysis: 21 with proven sepsis, 20 with no infection, and 60 unclassified. We excluded 57 because of incomplete datasets and 24 who had early-onset sepsis. IL-1ra and IL-6 increased significantly 2 days before diagnosis of sepsis; maximum median increases within the study period were 15-fold for IL-1ra and 12-fold for IL-6. The diagnostic sensitivities of IL-1ra, IL-6, and CRP concentrations on day 0 of diagnosis were 93%, 86%, and 43%, respectively; corresponding values on day -1 were 64%, 57%, and 18%. The specificities of IL-1ra, IL-6, and CRP concentrations were 92%, 83%, and 93%. cICAM-1 had a specificity of only 64%. INTERPRETATION IL-1ra and IL-6 are superior to cICAM-1 and CRP as predictors of sepsis 1 or more days before clinical diagnosis. Ad-hoc measurement of these cytokines could allow earlier initiation of antibiotic therapy with corresponding improvement in outcome in very-low-birthweight infants with sepsis.

Betulinic acid: A new cytotoxic agent against malignant brain-tumor cells

Malignant brain tumors are the most common solid tumors in children. The overall prognosis for this group of patients is still poor, emphasizing the importance of more effective therapies. Betulinic acid (Bet A) has been described as a novel cytotoxic compound active against melanoma and neuroblastoma cells. Here we report that Bet A was active against medulloblastoma and glioblastoma cell lines. In addition, Bet A exerted cytotoxic activity against primary tumor cells cultured from patients in 4 of 4 medulloblastoma‐tumor samples tested and in 20 of 24 glioblastoma‐tumor samples. Since a small percentage of primary‐glioblastoma‐tumor cells (4/24) did not respond to Bet‐A treatment, resistance to Bet A might occur. Induction of apoptosis by Bet A involved mitochondrial perturbations, since inhibition of the mitochondrial permeability transition by the mitochondrion‐specific inhibitor bongkrekic acid (BA) reduced Bet‐A‐induced apoptosis. In addition, mitochondria undergoing Bet‐A‐induced permeability transition triggered DNA fragmentation in isolated nuclei. Cytochrome c was released from mitochondria of Bet‐A‐treated cells, and might be involved in activation of caspases. Following treatment with Bet A, caspase‐8, caspase‐3 and PARP were proteolytically processed. Inhibition of caspase cleavage by the broad‐range caspase inhibitor zVAD.fmk strongly reduced Bet‐A‐induced apoptosis, indicating that apoptosis was mediated by activation of caspases. Since Bet A did not exhibit cytotoxicity against murine neuronal cells in vitro, these findings suggest that Bet A may be a promising new agent for the treatment of medulloblastoma and glioblastoma cells that clearly warrants further pre‐clinical and clinical evaluation. Int. J. Cancer 82:435–441, 1999.

Small molecule XIAP inhibitors cooperate with TRAIL to induce apoptosis in childhood acute leukemia cells and overcome Bcl-2-mediated resistance

Defects in apoptosis contribute to poor outcome in pediatric acute lymphoblastic leukemia (ALL), calling for novel strategies that counter apoptosis resistance. Here, we demonstrate for the first time that small molecule inhibitors of the antiapoptotic protein XIAP cooperate with TRAIL to induce apoptosis in childhood acute leukemia cells. XIAP inhibitors at subtoxic concentrations, but not a structurally related control compound, synergize with TRAIL to trigger apoptosis and to inhibit clonogenic survival of acute leukemia cells, whereas they do not affect viability of normal peripheral blood lymphocytes, suggesting some tumor selectivity. Analysis of signaling pathways reveals that XIAP inhibitors enhance TRAIL-induced activation of caspases, loss of mitochondrial membrane potential, and cytochrome c release in a caspase-dependent manner, indicating that they promote a caspase-dependent feedback mitochondrial amplification loop. Of note, XIAP inhibitors even overcome Bcl-2-mediated resistance to TRAIL by enhancing Bcl-2 cleavage and Bak conformational change. Importantly, XIAP inhibitors kill leukemic blasts from children with ALL ex vivo and cooperate with TRAIL to induce apoptosis. In vivo, they significantly reduce leukemic burden in a mouse model of pediatric ALL engrafted in non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice. Thus, XIAP inhibitors present a promising novel approach for apoptosis-based therapy of childhood ALL.

Betulinic acid-induced apoptosis in leukemia cells

Betulinic acid (BA), a natural component isolated from Birch trees, effectively induces apoptosis in neuroectodermal and epithelial tumor cells and exerts little toxicity in animal trials. Here, we show that BA-induced marked apoptosis in 65% of primary pediatric acute leukemia cells and all leukemia cell lines tested. When compared for in vitro efficiency with conventionally used cytotoxic drugs, BA was more potent than nine out of 10 standard therapeutics and especially efficient in tumor relapse. No crossresistances were found between BA and any cytotoxic drug. Intracellular apoptosis signaling in leukemia tumor cells paralleled the pathway found in neuroectodermal cells involving caspases, but not death receptors. In isolated mitochondria, BA induced release of both cytochrome c and Smac. Taken together, BA potently induces apoptosis in leukemia cells and should be further evaluated as a future drug to treat leukemia.

JNK/SAPK activity contributes to TRAIL-induced apoptosis

We report here that JNK/SAPKs are activated by TRAIL in parallel to induction of apoptosis in human T and B cell lines. Death signaling as well as JNK/SAPK activation by TRAIL in these cells is FADD- and caspase-dependent since dominant-negative FADD or the caspase inhibitor zVAD prevented both, apoptosis and JNK/SAPK activity. JNK/SAPK activity in response to triggering of CD95 by an agonistic antibody (αAPO-1) was also diminished by dominant-negative FADD or zVAD. Correspondingly, a cell line resistant to αAPO-1-induced death exhibited crossresistance to TRAIL-induced apoptosis and did not upregulate JNK/SAPK activity in response to TRAIL or αAPO-1. Inhibition of JNK/SAPK activity, by stably transfecting cells with a dominant-negative JNKK-MKK4 construct, reduced apoptosis in response to TRAIL or αAPO-1. Therefore, activation of JNK/SAPKs by TRAIL or αAPO-1 occurs downstream of FADD and caspases and contributes to apoptosis in human lymphoid cell lines.

Tumor Necrosis Factor–Related Apoptosis-Inducing Ligand–Mediated Proliferation of Tumor Cells with Receptor-Proximal Apoptosis Defects

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) might represent a future cytotoxic drug to treat cancer as it induces apoptosis in tumor cells without toxicity in animal trials. We recently described that in contrast to apoptosis, TRAIL mediates tumor cell survival and proliferation in certain tumor cells. Here we studied the effect of TRAIL on 18 cell lines and 53 primary leukemia cells and classified these tumor cells into four groups: TRAIL, anti-DR4 or anti-DR5 induced apoptosis in group A cells, whereas they had no effect on group 0 cells and mediated proliferation in group P cells. To our surprise, TRAIL induced simultaneous apoptosis and proliferation in group AP cells. More than 20% of all cells tested belonged to group P and showed TRAIL-mediated proliferation even in the presence of certain cytotoxic drugs but not inhibitors of nuclear factor-kappaB. Transfection with B-cell leukemia/lymphoma protein 2 transformed group A cells into group 0 cells, whereas transfection with Fas-associated polypeptide with death domain (FADD)-like interleukin-1-converting enzyme-inhibitory protein (FLIP) transformed them into group AP cells. Loss of caspase-8 or transfection of dominant-negative FADD transformed group A cells into group P cells. Taken together, our data suggest that proliferation is a frequent effect of TRAIL on tumor cells, which is related to receptor-proximal apoptosis defects at the level of the death-inducing signaling complex and should be prevented during antitumor therapy with TRAIL.

Blockade of the PD-1/PD-L1 axis augments lysis of AML cells by the CD33/CD3 BiTE antibody construct AMG 330: reversing a T-cell-induced immune escape mechanism

Bispecific T-cell engagers (BiTEs) are very effective in recruiting and activating T cells. We tested the cytotoxicity of the CD33/CD3 BiTE antibody construct AMG 330 on primary acute myeloid leukemia (AML) cells ex vivo and characterized parameters contributing to antileukemic cytolytic activity. The E:T ratio and the CD33 expression level significantly influenced lysis kinetics in long-term cultures of primary AML cells (n=38). AMG 330 induced T-cell-mediated proinflammatory conditions, favoring the upregulation of immune checkpoints on target and effector cells. Although not constitutively expressed at the time of primary diagnosis (n=123), PD-L1 was strongly upregulated on primary AML cells upon AMG 330 addition to ex vivo cultures (n=27, P<0.0001). This phenomenon was cytokine-driven as the sole addition of interferon (IFN)-γ and tumor necrosis factor-α also induced expression. Through blockade of the PD-1/PD-L1 interaction, AMG 330-mediated lysis (n=9, P=0.03), T-cell proliferation (n=9, P=0.01) and IFN-γ secretion (n=8, P=0.008) were significantly enhanced. The combinatorial approach was most beneficial in settings of protracted AML cell lysis. Taken together, we have characterized a critical resistance mechanism employed by primary AML cells under AMG 330-mediated proinflammatory conditions. Our results support the evaluation of checkpoint molecules in upcoming clinical trials with AMG 330 to enhance BiTE antibody construct-mediated cytotoxicity.

The Public Repository of Xenografts Enables Discovery and Randomized Phase II-like Trials in Mice

More than 90% of drugs with preclinical activity fail in human trials, largely due to insufficient efficacy. We hypothesized that adequately powered trials of patient-derived xenografts (PDX) in mice could efficiently define therapeutic activity across heterogeneous tumors. To address this hypothesis, we established a large, publicly available repository of well-characterized leukemia and lymphoma PDXs that undergo orthotopic engraftment, called the Public Repository of Xenografts (PRoXe). PRoXe includes all de-identified information relevant to the primary specimens and the PDXs derived from them. Using this repository, we demonstrate that large studies of acute leukemia PDXs that mimic human randomized clinical trials can characterize drug efficacy and generate transcriptional, functional, and proteomic biomarkers in both treatment-naive and relapsed/refractory disease.