The role of 3' poly(A) tail metabolism in tumor necrosis factor-alpha regulation.

In unstimulated RAW 264.7 macrophage-like cells, tumor necrosis factor-alpha (TNF-alpha) mRNA was transcribed and accumulated in the cytoplasm, but the TNF-alpha transcripts failed to associate with polysomes, and TNF-alpha protein was not detected. Stimulation with lipopolysaccharide (LPS) induced an increase in TNF-alpha transcription, cytoplasmic TNF-alpha mRNA accumulation, polysome association, and secretion of TNF-alpha protein. This process was associated with a 200-nucleotide increase in the apparent length of the TNF-alpha mRNA. The difference in TNF-alpha mRNA size was caused by marked truncation of the 3' poly(A) tail in unstimulated cells. Fully adenylated TNF-alpha mRNA appeared within 15 min of LPS stimulation. We speculate that removal of the poly(A) tail blocks initiation of TNF-alpha translation in unstimulated macrophages. LPS inactivates this process, allowing synthesis of translatable polyadenylated TNF-alpha mRNA.

Warming macrophages to febrile range destabilizes tumor necrosis factor-alpha mRNA without inducing heat shock.

We have previously reported that sustained tumor necrosis factor (TNF)-alpha expression is suppressed by temperatures in the febrile range in human macrophages. In this study, we examined the mechanisms of high-temperature-induced macrophage TNF suppression in the RAW 264.7 macrophage cell line. Incubating lipopolysaccharide (LPS)-stimulated RAW 264.7 cells at 40 degrees C reduced TNF secretion by 92% and peak TNF mRNA levels by 43% compared with cells incubated at 37 degrees C (P < 0.05) but did not affect levels of glyceraldehyde-3-phosphate dehydrogenase, beta-actin, or interleukin-6 mRNA. TNF mRNA half-life, measured after transcriptional arrest with actinomycin D, was reduced from 21.8 +/- 3.6 min in LPS-stimulated RAW 264.7 cells at 37 degrees C to 16.0 +/- 1.8 min at 40 degrees C (P < 0.03), but these cells at 40 degrees C did not alter transcription rate or TNF mRNA polysome association. TNF mRNA destabilization occurred at temperatures below the threshold (43 degrees C) for the generalized heat shock response in these cells. We conclude that heating macrophages to febrile-range temperatures attenuates sustained TNF expression by modulating posttranscriptional processing, including acceleration of TNF mRNA decay.

Differential effects of hyperthermia on macrophage interleukin-6 and tumor necrosis factor-alpha expression.

The pyrogenic cytokines tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) appear in the circulation during infections and injuries, but TNF-alpha and IL-6 are regulated differently in macrophages. We compared the effects of elevated temperatures within the usual febrile range on the expression of TNF-alpha and IL-6 in vitro in lipopolysaccharide (LPS)-stimulated human macrophages derived from peripheral blood monocytes (HuMoM phi). During an 18-h incubation at 37 degrees C with 5 ng/ml LPS, these cells released 5,030 +/- 1,460 pg TNF-alpha/10(6) cells (means +/- SE) and 1,380 +/- 280 pg IL-6/10(6) cells. In LPS-stimulated HuMoM phi incubated at 40 degrees C, TNF-alpha release was almost completely inhibited (76 +/- 76 pg TNF-alpha/10(6) cells; P < 0.01 compared with LPS-stimulated HuMoM phi at 37 degrees C), but release of IL-6 was preserved (1,600 +/- 780 pg IL-6/10(6) cells). Western and Northern analyses showed that levels of TNF-alpha mRNA and cell-associated and secreted TNF-alpha protein were decreased, but IL-6 expression was unchanged at 40 degrees C in LPS-stimulated macrophages. Incubating HuMoM phi at 40 degrees did not alter their viability after 18 h but induced a 75-fold increase in levels of the inducible heat-shock protein 72 (HSP-72) mRNA in the face of a 56% inhibition in total protein synthesis. Our results show that IL-6 expression persisted at incubation temperatures in the upper end of the physiological range that induced heat shock and attenuated the expression of functionally active TNF-alpha in LPS-stimulated HuMoM phi.

Characterization of tumor binding by the IC-21 macrophage cell line.

The purpose of this study was to determine if the SV40-transformed murine macrophage cell line IC-21 is a suitable model to study the selective high avidity binding of tumor cells by subpopulations of activated macrophages. IC-21 macrophages bound P815, RBL5, and EL-4 murine tumor cells with high avidity, as measured by the inverted centrifugation method. Tumor binding by IC-21 macrophages was competitively inhibited by crude membrane vesicles prepared from tumor cells but not by cell membranes prepared from nontransformed splenic leukocytes, suggesting that this process was mediated by tumor-specific binding sites. IC-21 macrophages and primary cultures of pyran copolymer-elicited peritoneal macrophages demonstrated similar tumor binding avidity, kinetics, saturability, and metabolic requirements for optimal high avidity tumor binding. However, compared with primary cultures of pyran copolymer-elicited peritoneal macrophages, IC-21 macrophages bound 4-fold more tumor cells and were more homogeneous for tumor binding capability. Finally, one third of maximal tumor cell binding by IC-21 macrophages was completed within 5 min of contact with tumor, suggesting that IC-21 macrophages constitutively expressed some high avidity tumor binding sites. Their stable and homogeneous capability for binding tumor cells and their ease of growth make the IC-21 macrophage cell line a potentially valuable model for elucidating the molecular mechanisms responsible for selective high avidity tumor binding by subpopulations of activated macrophages.

Effects of assay medium composition on macrophage-mediated tumor cell binding and lysis.

Activated macrophages express selective, high avidity tumor binding and cytotoxicity for tumor targets. The reported macrophage-mediated tumor binding and killing activities vary considerably among studies. Whereas, some of these studies utilized identical tumor target cells and similar sources of macrophages, the composition of the assay media used by different groups varied with respect to: type of medium (Eagle's minimal essential medium (EMEM), RPMI-1640, Dulbecco's modified Eagle's medium with 1.0 g/l glucose (DMEMLG), or with 4.5 g/l, glucose (DMEMHG)), concentration of serum (0-20%), and the addition of certain reagents (amphotericin B, pyruvate, and Hepes). The purpose of this study was to evaluate the effects of varying these parameters of medium composition on macrophage-mediated high (HA) and low avidity tumor binding and cytolysis. Tumor cytolysis was measured with an 18 h 51Cr release assay using peritoneal macrophages from C3H/HeN mice primed in vivo with pyran copolymer and further stimulated in vitro with lipopolysaccharide (10 ng/ml). P815 tumor cells were used as targets at a 10:1 effector:target ratio. Binding of targets to macrophages was determined by a method utilizing inverted centrifugation. Under optimal conditions, 2 x 10(5) macrophages bound as many as 17,600 +/- 3565 tumor cells and caused up to 50.4 +/- 3.6% cytolysis. Assays performed in DMEMHG compared with the other three media tested resulted in HA tumor binding and cytolysis which were decreased by up to 42.5% (P less than 0.05) and 64.3% (P less than 0.01), respectively, compared with the three other types of medium. The addition of pyruvate (1 mM) to EMEM with 5% fetal calf serum (FCS) stimulated 98.8% (P less than 0.01) and 50.6% (P less than 0.1) increases in tumor binding and cytolysis, respectively compared with EMEM/5% FCS alone, while Hepes (25 mM) stimulated 58.3% (P less than 0.01) and 37.5% (P less than 0.1) increases in these activities. Amphotericin B (2.5 micrograms/ml) completely abrogated tumor cytolysis, but it caused no change in tumor binding. Serum produced variable effects on macrophage-mediated tumor killing. Five of six lots of FCS inhibited tumor lysis, by 16 to 98% (32.6 +/- 28.6%; mean +/- SD). However, the same lot of FCS which inhibited cytolysis by 98% enhanced HA binding by 152% (P less than 0.05). Finally, several commercially available serumless medium preparations supported macrophage-mediated tumor binding; however, none of the serumless media tested supported macrophage-mediated tumor cytolysis. We conclude that common differences in assay medium composition can markedly alter macrophage-mediated tumor cell binding and cytolysis.