Characterization of the 14 kda fragment of human tumor-necrosis-factor-alpha

We report the characterization of a 14 kDa degradation fragment from recombinant human tumour necrosis factor-alpha (TNF alpha) by N-terminal sequencing and mass spectrometry. A single site between the dibasic residues Arg(31)-Arg(32) of the mature recombinant 17 kDa protein has been identified as the target site that generates the 14 kDa fragment. The observation that a maximum of 33% degradation occurs suggests that only one monomer per TNF trimer is cleaved. E. coli proteases specific for dibasic residues are thought to be responsible for this cleavage. A strategy has been developed which completely inhibits proteolysis. This strategy has been used to reduce the 14 kDa degradation fragment obtained from approximately 33% of the total purified protein to zero.     

Resistance to cytolysis by tumor necrosis factor alpha in malignant gynecological cell lines is associated with the expression of protein(s) that prevent the activation of phospholipase A2 by tumor necrosis factor alpha.

Although there are a limited number of cell lines that are sensitive to cytolysis by tumor necrosis factor alpha (TNF alpha), the vast majority are resistant. The analysis of TNF alpha-sensitive cells has shown that phospholipase A2 is activated by TNF alpha in these cells and that the activity of phospholipase A2 is required for their cytolysis. Many cell lines that are resistant to TNF alpha-mediated cytolysis are dependent on the maintenance of protein synthesis for their resistance. We have recently shown that this is also true for TNF alpha-resistant cell lines derived from cervical (ME-180 and SiHa) and ovarian (SK-OV-3 and OVCAR-3) carcinomas, in that they are sensitive to cytolysis by TNF alpha only in the presence of protein synthesis inhibitors. Here we show that the TNF alpha-mediated cytolysis of these resistant cell lines in the presence of the protein synthesis inhibitor emetine is similar to that of sensitive cells, in that cytolysis is inhibited by the inhibitors of phospholipase A2. The measurement of the release of radiolabeled material from cervical and ovarian carcinoma cell lines prelabeled with [3H]arachidonic acid showed that not only was phospholipase A2 required for the cytolysis of these cells by TNF alpha in the presence of protein synthesis inhibitors, but more importantly, phospholipase A2 was not activated by TNF alpha unless protein synthesis was inhibited. These results indicate that a protein synthesis-dependent resistance mechanism expressed by these cell lines blocks TNF alpha-mediated cytolysis by preventing the activation of phospholipase A2 by TNF alpha.     

Selective enhancement of the tumour necrotic activity of TNF alpha with monoclonal antibody.

The binding and biological activity of human TNF alpha on endothelial and tumour cells has been studied in the presence of monoclonal antibodies (MAbs). In particular, one monoclonal antibody to TNF alpha (MAb 32) has been identified which failed to inhibit binding and cytotoxicity of TNF alpha on WEHI-164 tumour cells but which was a potent inhibitor of TNF alpha-induced endothelial cell procoagulant activity on bovine aortic endothelial cells. The ability of MAb 32 to inhibit selectively the actions of TNF alpha on endothelial cells but not on tumour cells suggests a mechanism for enhancement of the anti-tumour action of TNF alpha in vivo when in complex with this antibody. Treatment of tumour bearing mice (WEHI-164 and Meth A fibrosarcoma) with TNF alpha-MAb 32 complex resulted in a 5- to 10-fold enhancement in the potency of the cytokine in comparison to free TNF alpha. Complexes between this cytokine and other MAbs generally resulted in either no effect or inhibition of TNF alpha activity in vivo and in vitro. Neither intact MAb 32 nor FAb' fragments of MAb 32 showed any tumour regressive activity in the absence of TNF alpha. The FAb' fragments were equipotent to the bivalent form of the antibody in enhancing TNF alpha activity. These data provide evidence that it is possible to segregate the individual biological activities of TNF alpha with concomitant enhancement of the tumour regressive activity of the cytokine in vivo.     

Purification of a colon cancer cell growth inhibitor and its identification as an insulin-like growth factor binding protein.

We have purified a protein from serum-free conditioned medium of the HT29 human colon adenocarcinoma cell line based on its ability to inhibit the proliferation of the same cell line. The purification procedure consisted of acid gel permeation, semipreparative, and analytical reversed-phase chromatographies. The high-pressure liquid chromatography-purified colon cancer cell growth inhibitor migrates as a single band of 27 and 34 kDa on sodium dodecyl sulfate/polyacrylamide gels under nonreducing and reducing conditions, respectively. NH2-terminal amino acid sequence analysis of the first 32 residues has demonstrated that this protein belongs to the insulin-like growth factor-binding protein (IGFBP) family. More precisely, this growth inhibitor appeared to be identical to the recently cloned human IGFBP-4. This IGFBP (HT29-IGFBP) has been characterized by performing ligand blotting and competitive binding experiments. The affinity of HT29-IGFBP for insulin-like growth factor (IGF) II (approximately 3.4 x 10(10) M-1) is slightly greater than its affinity for IGF-I (approximately 1.4 x 10(10) M-1). HT29 cells also produce two other isoforms (28 and 31 kDa, nonreduced) of the HT29-IGFBP having the same partial NH2-terminal amino acid sequence as the 27-kDa protein. The monoclonal antibody alpha IR-3 is known to block the mitogenic actions of IGFs. alpha IR-3 inhibited the growth of HT29 cells, thus suggesting that IGFs are required for the growth of these colon cancer cells.     

Correlation of tumor necrosis factor alpha (TNF alpha) with high Caspase 3-like activity in myelodysplastic syndromes.

Increased intramedullary apoptotic death of hematopoietic cells is thought to contribute to the ineffective hematopoiesis in myelodysplastic syndromes (MDS). Furthermore, high amounts of tumor necrosis factor alpha (TNF alpha) have previously been correlated with apoptosis in MDS marrows. The present studies were undertaken to examine the status of two key downstream effectors of TNF alpha signaling, i.e. Caspase 1 and Caspase 3 enzymes, using a fluorometric assay in the bone marrow aspirate mononuclear cells (BMMNC) in relation to apoptotic DNA fragmentation detected by in situ end-labeling (ISEL) of DNA and with localization of TNF alpha in the corresponding biopsies from 14 MDS patients. Both Caspase 1 and 3 were detectable in freshly harvested BMMNC, albeit median Caspase 3 levels (47.5 units/mg protein) being almost 10 times higher than Caspase 1 (4.0 units/mg protein). Upon short-term culture for 4 h in a serum-supplemented medium in vitro a significant increase was seen in Caspase 3 activity (58.8 +/- 13.9 at 0 h vs. 177.8 +/- 55.2 units/mg protein at 4 h, n = 14, P = 0.017) and in percent cells labeled by ISEL (apoptotic index or AI%: 0.76% +/- 0.25% vs. 3.99% +/- 1.1%, n = 14, P = 0.004, respectively). Caspase 1 activity increased after 15 min in culture. Interestingly, TNF alpha levels measured by immunohistochemistry correlated with the net increase in Caspase 3 activity after 4 h (p = 0.517, n = 13, P = 0.07) and the starting levels of Caspase 1 at 0 h correlated with the Caspase 3 levels attained at 4 h (p = 0.593, n = 13, P = 0.033). Additionally when TNF alpha-positive bone marrows (8/14) were compared with the negative marrows (6/14) the Caspase 3 levels were significantly higher in the TNF alpha-positive marrows (189.6 +/- 66.2 vs. 25.0 +/- 14.6 units/mg protein, respectively, P = 0.043). The increase in AI%, though not statistically significant, was also higher in the TNF alpha-positive marrows. Finally in HL60 cells the effects of different Caspase inhibitors and pentoxifylline (PTX) (interferes with lipid signaling of cytokines) on TNF alpha-induced apoptosis were evaluated. TNF alpha treatment significantly increased AI% (P < 0.003) as compared to the untreated controls. A co-treatment with three Caspase inhibitors, zVAD.FMK (inhibitor of Caspases 1 and 3, 10 microM/l), Ac.YVAD.FMK (Caspase 1 inhibitor, 1 microM/l), Ac.DEVD.FMK (Caspase 3 inhibitor, 10 microM/l) as well as PTX (250 microM/l) significantly curtailed the AI% induced by TNF alpha. The present studies thus identify the downstream effectors of TNF alpha-inducible apoptosis in MDS and so also the suppressors of TNF alpha apoptotic signaling. These results may have significant clinical implications in the therapy of MDS in the future.     

Identification of a guanine nucleotide-binding protein G(o) in human neuroblastoma

It has been found that neuroblastoma contains a relatively high immunoreactivity for the alpha subunit of guanine nucleotide-binding protein G(o) (G(o) alpha) (Kato et al., Cancer Res. 47, 5800-5805, 1987), which is predominantly localized in the nervous tissues and neuroendocrine cells. To make clear whether neuroblastoma indeed produces G(o) alpha, we purified a guanine nucleotide-binding protein from human neuroblastoma and compared it with G(o) alpha from human brain. A guanine nucleotide-binding protein serving as the specific substrate of islet-activating protein, pertussis toxin, was purified from a human neuroblastoma tissue obtained at surgical resection. The protein had a molecular mass of 39 kDa and reacted with rabbit antibodies to bovine G(o) alpha. Heat stability and kinetics of guanosine 5'-(3-O-thio)triphosphate binding of this protein were very similar to those of G(o) alpha purified from human brain. The results of peptide mapping analysis with a limited proteolysis and amino acid analysis indicated that there was no difference between these proteins. Therefore, it was concluded that this guanine nucleotide-binding protein isolated from neuroblastoma was very similar, if not identical, to brain G(o) alpha.     

Role of TNF alpha in bryostatin-induced inhibition of human hematopoiesis.

In previous studies bryostatin has been shown to cause dose-dependent stimulatory or inhibitory effects on colony formation in acute myeloid leukemias. In this study we investigated the inhibitory effect of high dose bryostatin-1 (bryo-1) on normal human bone marrow mononuclear cells (BMNC) colony-forming capacity. Preculturing BMNC for 24 h with 250 nM bryo-1 reduced colony formation by 66 +/- 11% whereas this treatment did not reduce clonogenic capacity of highly purified CD34+ BMNC. When precultures with bryo-1 were performed in the presence of several cytokine neutralizing antibodies abrogation of the inhibitory effect could only be demonstrated by anti-TNF alpha. However, preculturing of BMNC or CD34+ cells with a wide range of TNF alpha concentrations as well as TNF alpha neutralization in supernatant of bryo-1-stimulated BMNC failed to affect the inhibitory effect on CD34+ cells. Both indicate that TNF alpha was not the only factor responsible for the inhibitory effect. Depletion of CD14+ cells from BMNC cultures showed that upon bryo-1 exposure the monocytes served as the main source of TNF alpha but not as a source of the inhibitory cytokine(s): in CD14+-depleted cultures the combination of exogenous added TNF alpha and bryo-1 resulted in an inhibition of colony formation comparable to that found in crude BMNC. In contrast, purified CD34+ cultures were not directly affected by bryo-1 and TNF alpha. However, clonogenic growth of purified CD34+ cells was inhibited if mononuclear cells were preincubated with TNF alpha alone for 24 h, and the supernatant of these cultures was used together with bryo-1. These results show that bryo-1-induced inhibition of clonogenic cell growth is not mediated by a direct effect of bryo-1 on CD34 cells but is a result of a process involving production of TNF alpha by CD14+ cells upon bryo-1 stimulation together with the induction of (a) secondary factor(s) by TNF alpha, which together with bryo-1 itself is inhibitory towards clonogenic cell growth.     

Growth arrest of the breast cancer cell line, T47D, by TNF alpha; cell cycle specificity and signal transduction.

The effects of tumour necrosis factor-alpha (TNF alpha) on the growth and DNA synthesis of the human breast cell line, T47D, were studied. A dose-dependent, reversible inhibition of thymidine incorporation and cell growth was observed in the range of 0.1 ng ml-1 to 100 ng ml-1 of TNF alpha. Cell viability was not impaired in any of the experiments. Flow-cytometric DNA analysis demonstrated that after 24 h exposure to TNF alpha, T47D cells accumulated in the G1 phase of the cell cycle, and were depleted in the G2/M and S phases, suggesting a block in the progression of the G1/S transition. The involvement of protein kinases (PK) and protein phosphatases in TNF alpha-induced signal transduction was also investigated. A transient and rapid 2-fold increase in total cellular protein kinase C (PKC) activity was detected after 10 min exposure to TNF alpha. To study the role of the observed PKC activation in the cytostatic effect of TNF alpha, T47D cells were exposed to the cytokine in the presence of the potent PKC inhibitor, H7. The inhibitory effect of TNF alpha on thymidine incorporation was not affected by exposure to H7 at concentrations sufficient to block the stimulation of thymidine up-take induced by the PKC agonist, phorbol-12-myristate-13-acetate (PMA). The involvement of other signalling pathways was addressed using the cyclic nucleotide-dependent PK inhibitor, H8; the calmodulin-dependent PK inhibitor, W7; and the inhibitor of protein phosphatases PP1 and PP2B, okadaic acid. Exposure of T47D cells to these enzyme inhibitors failed to antagonise the inhibition of thymidine incorporation by TNF alpha. Taken together, these results indicate that the cytostatic effect of TNF alpha on T47D cells occurs at the G1/S transition of the cell cycle, and is mediated by an intracellular pathway which does not involve the activity of protein kinases C and A, nor protein phosphatases PP1, PP2B.     

Phosphatase inhibitors modulate the growth-regulatory effects of human tumor necrosis factor on tumor and normal cells.

Tumor necrosis factor alpha (TNF-alpha) has been shown to inhibit the growth of tumor cells and stimulate the growth of certain normal cells in vitro. The mechanism by which TNF exerts its cell growth-regulatory effects is not known. In this report, we investigated the effects of phosphatase inhibitors on the cell growth-inhibitory effects of TNF on L-929, a highly sensitive murine connective tissue tumor cell line, and on the growth-stimulatory effects of TNF on normal human fibroblasts. The antiproliferative effects of TNF on L-929 cells were inhibited by orthovanadate, an inhibitor of phosphotyrosine phosphatases, in a dose-dependent manner. Okadaic acid, which is a specific inhibitor of phosphoserine- and phosphothreonine-specific phosphatases, also blocked the growth-inhibitory effects of TNF, suggesting that TNF may function through the activation of certain phosphatases. These inhibitors had no effect on TNF receptors. Addition of phosphatase inhibitor, even 12 h after the treatment of cells with TNF, was sufficient to block the antiproliferative effects of the cytokine, suggesting that the inhibitor is acting at a late event in the pathway of action of TNF. Cells were protected by orthovanadate from the cytotoxic effects of TNF even in the presence of actinomycin D or cycloheximide, thus indicating the lack of a requirement for de novo protein synthesis. Orthovanadate altered the cell morphology from flat spindle shapes to rounded ones. Besides anticellular effects, a phosphatase inhibitor also suppressed the proliferative effects of TNF on human fibroblasts. These results thus suggest that phosphatases may be needed for both proliferative and antiproliferative effects of this cytokine. This is the first report to suggest that phosphatases play a role in the growth-regulatory action of TNF.     

The zinc finger protein A20 interacts with a novel anti-apoptotic protein which is cleaved by specific caspases.

A20 is a Cys2/Cys2 zinc finger protein which is induced by a variety of inflammatory stimuli and which has been characterized as an inhibitor of cell death by a yet unknown mechanism. In order to clarify its molecular mechanism of action, we used the yeast two-hybrid system to screen for proteins that interact with A20. A cDNA fragment was isolated which encoded a portion of a novel protein (TXBP151), which was recently found to be a human T-cell leukemia virus type-I (HTLV-I) Tax-binding protein. The full-length 2386 bp TXBP151 mRNA encodes a protein of 86 kDa. Like A20, overexpression of TXBP151 could inhibit apoptosis induced by tumour necrosis factor (TNF) in NIH3T3 cells. Moreover, transfection of antisense TXBP151 partially abolished the anti-apoptotic effect of A20. Furthermore, apoptosis induced by TNF or CD95 (Fas/APO-1) was associated with proteolysis of TXBP151. This degradation could be inhibited by the broad-spectrum caspase inhibitor zVAD-fmk or by expression of the cowpox virus-derived inhibitor CrmA, suggesting that TXBP151 is a novel substrate for caspase family members. TXBP151 was indeed found to be specifically cleaved in vitro by members of the caspase-3-like subfamily, viz. caspase-3, caspase-6 and caspase-7. Thus TXBP151 appears to be a novel A20-binding protein which might mediate the anti-apoptotic activity of A20, and which can be processed by specific caspases.     

Transient increase of plasma interleukin-6 after infusion of recombinant tumor necrosis factor alpha in advanced cancer patients

Tumor Necrosis Factor alpha administered in the therapy of advanced cancer influences certain hormones and cytokines secretion. In turn, these also modulate the biological activity of Tumor Necrosis Factor alpha. It has been shown in several studies that the cytokine Interleukin -6 (IL)-6 is produced in response to various hormones and other cytokines eg. Tumor Necrosis Factor (TNF alpha). In our study we focused on the Interleukin-6 (IL-6) secretion in response to TNF alpha administration in 12 patients undergoing TNF alpha biotherapy due to advanced neoplastic disease. Plasma IL-6 was estimated prior and at various time points (2, 3, 5 a,d 12 hours) after TNF alpha intravenous infusion. IL-6 level was estimated with ELISA method. In conclusion, we suggest the stimulating influence of hrec TNF alpha administered as therapy for advanced cancer on IL-6 secretion.     

Phase II study of liposomal muramyl tripeptide in osteosarcoma: the cytokine cascade and monocyte activation following administration.

PURPOSE: A phase II trial that uses liposome-encapsulated muramyl tripeptide phosphatidylethanolamine (L-MTP-PE) in patients with relapsed osteosarcoma is underway. To determine if in vivo cytokine induction plays a role in the mechanism of action of L-MTP-PE, we investigated the circulating cytokine levels of 16 patients who were undergoing therapy. PATIENTS AND METHODS: Patients had histologically proven osteosarcoma and pulmonary metastases that developed either during adjuvant chemotherapy or that were present at diagnosis and persisted despite chemotherapy. Patients were rendered disease-free by surgery. The major goal of the study was to improve the disease-free interval in this high-risk group. L-MTP-PE 2 mg/m2 was infused during a 1-hour period twice a week for 12 weeks, then once a week for 12 weeks. Serial blood samples were collected after L-MTP-PE administration and were assayed for cytokine levels (tumor necrosis factor-alpha [TNF alpha] interleukin-1 alpha [IL-1 alpha], IL-1 beta, IL-6, interferon-gamma [IFN-gamma], neopterin, C-reactive protein). RESULTS: After the infusion of L-MTP-PE, there was rapid induction of circulating TNF alpha and IL-6. TNF alpha levels peaked 1 to 2 hours after infusion in 10 of 16 patients, whereas peak IL-6 levels were detected at 2 to 3 hours in all patients. Induction of circulating TNF alpha and IL-6 was evident only after the first dose of L-MTP-PE. Neither IL-1 alpha nor IL-1 beta was detected in the plasma. Neopterin levels increased at 24 hours postinfusion, which indicated macrophage activation, and were not related to the induction of circulating IFN-gamma. C-reactive protein was elevated in all patients at 24 hours and decreased by 72 hours. Unlike circulating TNF alpha and IL-6, elevations in C-reactive protein and neopterin could be detected throughout the treatment course. CONCLUSION: It is concluded that L-MTP-PE has specific biologic effects in patients with osteosarcoma that may be important to the drug's immunostimulatory capacity and its effectiveness as an antitumor agent.     

Purification of human chorionic gonadotropin from pregnancy urine by immunoaffinity chromatography using a monoclonal antibody anti-beta chain hCG

Human chorionic gonadotropin (hCG) is found in pregnancy urine and in urine from some cancer patients. It is found in a variety of forms, whose concentrations have clinical importance. A highly active preparation of hCG from pregnancy urine has been obtained by a simple method of immunoaffinity chromatography using a monoclonal antibody (MAb) against beta-subunit hCG. The recuperation of hCG from urine extract by this procedure was 84.18%, with a purification factor of 19.78. The specificity activity of the obtained hCG was 9705.36 IU/mg. The preparation appeared homogeneous by SDS-PAGE, analytical gel filtration, and double immunodiffusion. The molecular mass was 40 kDa by gel filtration on Superose 12 HR. Molecular masses estimated by electrophoresis were 14.6 and 23.5 kDa for the alpha and beta subunits, respectively.     

Identification, characterization, and cloning of TIP-B1, a novel protein inhibitor of tumor necrosis factor-induced lysis.

Some cancer cells evade elimination by virtue of their insensitivity to agents that induce apoptosis. Conversely, the side effects of anticancer agents could be diminished if normal cells were more resistant. To further elucidate the factors that contribute to the susceptibility of a cell to apoptosis, these investigations were designed to identify proteins isolated from cells exposed to low concentrations of tumor necrosis factor (TNF) that, when incubated with normally TNF-sensitive cells, protect these cells from TNF-induced cytotoxicity. TIP-B1, a novel protein, has been identified, purified, and characterized from cytosolic extracts of TNF-treated human fibroblasts. The approximately 27 kDa pI-4.5 TIP-B1 protein is unique based on both the sequence of three internal peptides (comprising 51 amino acids) and the nucleotide sequence of the corresponding 783-bp cDNA partial clone. Western blot analyses using polyclonal antisera raised against both the purified native TIP-B1 and the approximately 14 kDa product of the cDNA partial TIP-B1 clone, as well as Northern blot analyses using the cDNA insert as a probe, indicate that TIP-B1 may belong to a family of proteins that are expressed in a number of cell lines from diverse tissues. TNF-sensitive cells, when exposed to 4-10 microg/ml concentrations of TIP-B1 prior to the addition of TNF, are completely protected from TNF-induced lysis. Furthermore, TIP-B1 protects cells from apoptotic lysis induced by TNF. Preincubation of TIP-B1 with TNF does not affect the ability of TNF to induce lysis. Moreover, TIP-B1 does not seem to interfere with the interactions between TNF and the TNF receptors, based on a preliminary flow cytometric analysis of the cellular binding of biotinylated TNF. On the basis of these characteristics, TIP-B1 is not a soluble TNF receptor, an anti-TNF antibody, nor a protease that degrades TNF; yet TIP-B1 functions when added exogenously to cells. These characteristics, its novel sequence, and its function when added exogenously to cells indicate that TIP-B1 is unique and is not one of the other proteins reported previously to be involved in resistance to TNF. The ability of TIP-B1 to function after exogenous incubation with target cells makes TIP-B1 a likely candidate for therapeutic manipulation of TNF-induced effects.     

Relationship between expression of topoisomerase II isoforms and intrinsic sensitivity to topoisomerase II inhibitors in breast cancer cell lines.

Topoisomerase II is a key target for many anti-cancer drugs used to treat breast cancer. In human cells there are two closely related, but differentially expressed, topoisomerase II isoforms, designated topoisomerase II alpha and beta. Here, we report the production of a new polyclonal antibody raised against a fragment of the C-terminal domain of the 180 kDa form of topoisomerase II (the beta isoform), which does not cross-react with the 170 kDa form (the alpha isoform). Using this antibody, together with a polyclonal antibody specific for the 170 kDa isoform of topoisomerase II, we have examined the relationship between the sensitivity of a panel of human breast cancer cell lines to different classes of topoisomerase II inhibitors and cellular levels of the topoisomerase II alpha and beta proteins. We found that sensitivity to amsacrine showed a correlation with the level of expression of topoisomerase II alpha protein, and that sensitivity to etoposide showed a similar correlation with the level of expression of topoisomerase II beta protein. There was also a relationship between sensitivity of these cell lines to mitoxantrone and the cellular level of both isoforms of topoisomerase II. No relationship was found between the level of mRNA for topoisomerase II alpha or beta, and either sensitivity of breast cancer cell lines to topoisomerase II inhibitors or the level of topoisomerase II protein expression.     

Anti-tumor effects of interferon in mice injected with interferon-sensitive and interferon-resistant Friend leukemia cells. V. Comparisons with the action of tumor necrosis factor

A number of similarities and dissimilarities in the anti-tumor effects of TNF and interferon alpha/beta have been observed in DBA/2 mice injected with Friend erythroleukemia cells (FLC). Mouse TNF exerted marked anti-tumor effects in mice injected either s.c. or i.p. with FLC lines 3C18 or 3 gamma R8 resistant in vitro to the cytotoxic effects of TNF. Likewise, mouse interferon alpha/beta had anti-tumor activity in mice injected with these FLC, resistant to the action of interferon alpha/beta or gamma in vitro. The results of histopathologic examination and 31P nuclear magnetic resonance analyses of 3C18 FLC s.c. tumors injected with TNF resembled the results previously obtained for 3C18 FLC tumors injected with interferon alpha/beta, although the effects of TNF occurred more rapidly. Injection of mice with antibody to mouse interferon alpha/beta or gamma did not abrogate the anti-tumor effects of TNF in mice injected i.p. with FLC. Our results suggest that in this experimental system the anti-tumor effects of TNF, like interferon alpha/beta, do not result from a direct effect on the tumor cells themselves but are host-mediated.     

Positive and negative effects of tumor necrosis factor on colony growth from highly purified normal marrow progenitors.

The effects of recombinant human tumor necrosis factor alpha (TNF alpha) on colony growth were studied using highly enriched progenitor cells from normal human bone marrow. Supplementation of TNF to culture resulted in a dose-dependent suppression of granulocyte colony-stimulating factor (G-CSF) induced granulocytic colony formation and also erythropoietin (Epo) induced erythroid burst formation. However, the number of erythroid bursts, stimulated by interleukin-3 (IL-3) plus Epo, increased when TNF was added at comparable concentrations. Further, TNF enhanced eosinophilic colony growth induced by IL-3 or granulocytic-macrophage colony-stimulating factor (GM-CSF). In GM-CSF cultures TNF (100-1000 U/ml) also induced granulocytic and macrophage colonies. The addition of neutralizing antibodies against G-CSF, GM-CSF, or interleukin-6 (IL-6) to culture did not abrogate the observed effects of TNF, so that stimulation of myeloid colony growth was unlikely to result from the secondary induction of G-CSF or GM-CSF. TNF therefore exerts favourable effects on hematopoietic progenitors responsive to the more primitive colony-stimulating factors (IL-3, GM-CSF) and potent negative effects on precursors reactive to the single lineage G-CSF and Epo. These contrasting effects of TNF suggest that TNF, when available to marrow progenitors at similar tissue concentrations, may drive hematopoiesis within the progenitor cell compartment into selected directions.     

Expression of mRNA for cytokines in tumor-infiltrating mononuclear cells in ovarian adenocarcinoma and invasive breast cancer.

Cytokine gene expression in tumor-infiltrating lymphocytes (TIL) in frozen-tissue sections of 2 types of human solid tumor--ovarian adenocarcinoma and invasive breast cancer--was examined by in situ hybridization with 35S-labeled cDNA probes for human cytokines. The proportion of cells containing mRNA able to hybridize to the antisense c-DNA probes for interleukin 2 (IL-2), tumor necrosis factor alpha (TNF alpha), interferon gamma (IFN gamma) or receptors for IL-2 (either p55 or p70) was also determined in human normal peripheral lymphoid tissues and inflammatory tissues. Few cells were positive for IL2 and TNF alpha mRNA in reactive human lymph nodes and tonsils. Inflammatory lesions, such as salpingitis or chronic active hepatitis, contained 10-20 times more cells positive for cytokine mRNA than reactive lymphoid tissue. In contrast, tumor-infiltrating lymphocytes (TIL) in the stroma of ovarian carcinomas or most ductal breast tumors only rarely expressed mRNA for TNF alpha, IL2 or IFN gamma. The intensity of mononuclear cell infiltration in these tumors correlated positively with the percentage of cells which expressed mRNA for IL-2, TNF alpha and IL-2R. In those ductal breast carcinomas which contained intracellular or intraductal mucins, up to 30% of lymphoid cells in the tumor stroma were positive for IL-2, TNF alpha, IFN gamma and IL-2R. Thus, strong evidence for local activation of mononuclear cells in situ, exemplified by the expression of genes for cytokines, was obtained only in inflammatory lesions and in mucin-producing breast carcinomas. In most carcinomas studied, few TIL expressed genes for cytokines as measured by in situ hybridization. Thus, human solid tumors appear to differ in their ability to induce gene expression for cytokines in TIL.