Detection of bacterial pyrogens on the basis of their effects on gamma interferon-mediated formation of neopterin or nitrite in cultured monocyte cell lines

Diagnostic Accuracy of Synovial Neopterin, TNF-α and Presepsin in Periprosthetic Joint Infection: A Prospective Study

BACKGROUND

Due to the lack of specificity of conventional diagnostic tools, the prediction of periprosthetic joint infections (PJI) remains challenging. The purpose of this study was to evaluate the accuracy of synovial fluid neopterin, presepsin, and TNF-α as diagnostic parameters and to compare it to the biomarkers recommended in the 2018 definition of periprosthetic hip and knee infection.

METHODS

Between August 2018 and July 2019, a prospective cohort study was conducted in 80 patients with painful hip, shoulder, and knee arthroplasty. In addition to medical history, clinical and laboratory data were gathered. PJI was diagnosed based on the 2018 definition of periprosthetic hip and knee infection. Synovial joint fluid was analyzed for biomarker measurement using standard quantitative enzyme immunoassay kits.

RESULTS

Fifty-three patients (66%) were classified as the aseptic group and twenty-seven patients (34%) as the PJI group. The mean levels of synovial fluid neopterin were significantly higher (p < 0.01) in the PJI group than those in the aseptic group (aseptic 8.3 ± 6.9 vs. PJI 20.9 ± 21.4 nmol/L). The average values of synovial fluid TNF-α and presepsin were not significantly higher in the PJI group than those in the aseptic group (presepsin: aseptic 0.13 ± 0.19 vs. PJI 0.11 ± 0.32 ng/mL, p = 0.08; TNF-α: aseptic 6.6 ± 7.3 vs. PJI 46.3 ± 123.2 pg/mL, p = 0.17). Synovial fluid neopterin was 59% specific and 74% sensitive with a cut-off value of 7.2 nmol/L. The sensitivity and specificity of synovial fluid TNF-α were 63 and 51% with a cut-off value of 3.9 pg/mL. Synovial fluid presepsin was 51% specific and 29% sensitive with a cut-off value above 0.06 ng/mL.

CONCLUSION

Synovial fluid neopterin appears to a reliable diagnostic marker for detection of PJI. In contrast, synovial fluid TNF-α and presepsin are not suitable to exclude or diagnose PJI.

The role of urinary pteridines as disease biomarkers

Pteridines and their derivatives function as intermediates in the metabolism of several vitamins and cofactors, and their relevance to disease has inspired new efforts to study their roles as disease biomarkers. Recent analytical advances, such as the emergence of sensitive mass spectrometry techniques, new workflows for measuring pteridine derivatives in their native oxidation states and increased multiplexing capacities for the simultaneous determination of many pteridine derivatives, have enabled researchers to explore the roles of urinary pteridines as disease biomarkers at much lower levels with greater accuracy than with previous technologies or methods. As a result, urinary pteridines are being increasingly studied as putative cancer biomarkers with promising results being reported from exploratory studies. In addition, the role of urinary neopterin as a universal biomarker for immune system activation is being investigated in new diseases where it is anticipated to become a useful supplementary marker in clinical diagnostic settings. In summary, this review provides an overview of recent developments in the clinical study of urinary pteridines as disease biomarkers, covers the most promising aspects of advanced analytical techniques being developed for the determination of urinary pteridines and discusses the major challenges associated with implementing pteridine biomarkers in clinical laboratory settings.

Implementing the in vitro pyrogen test: one more step toward replacing animal experimentation

The MAT-based in vitro pyrogen test should be GLP-accredited for use in medical device testing

Gingival crevicular fluid as a periodontal diagnostic indicator- II: Inflammatory mediators, host-response modifiers and chair side diagnostic aids

There has been a steady growing trend during the last few decades to develop tools to monitor periodontitis, in the field of oral disease diagnosis. Since GCF has the chance of being closely approximated to the periodontal tissues where periodontal disease begins, it seems to provide more information than markers in saliva. In response to bacterial infection, host production of inflammatory mediators, may be the trigger for periodontal disease progression. Existing paradigms in the biology of periodontitis have supported the detection of elevated levels of these mediators in GCF. This article is the Part II of the review that deals with inflammatory mediators and host-response modifiers as the potential biomarkers present in gingival crevicular fluid (GCF) and the chair side point-of-care diagnostic aids applicable to monitor periodontal inflammation.

Regulation of avoidant behaviors and pain by the anti-inflammatory tyrosine phosphatase SHP-1

The protein tyrosine phosphatase SHP-1 is a critical regulator of cytokine signaling and inflammation. Mice homozygous for a null allele at the SHP-1 locus have a phenotype of severe inflammation and are hyper-responsive to the TLR4 ligand LPS. TLR4 stimulation in the CNS has been linked to both neuropathic pain and sickness behaviors. To determine if reduction in SHP-1 expression affects LPS-induced behaviors, responses of heterozygous SHP-1-deficient (me/+) and wild-type (+/+) mice to LPS were measured. Chronic (4-week) treatment with LPS induced avoidant behaviors indicative of fear/anxiety in me/+, but not +/+, mice. These behaviors were correlated with a LPS-induced type 2 cytokine, cytokine receptor, and immune effector arginase profile in the brains of me/+ mice not found in +/+ mice. Me/+ mice also had a constitutively greater level of TLR4 in the CNS than +/+ mice. Additionally, me/+ mice displayed constitutively increased thermal sensitivity compared to +/+ mice, measured by the tail-flick test. Moreover, me/+ glial cultures were more responsive to LPS than +/+ glia. Therefore, the reduced expression of SHP-1 in me/+ imparts haploinsufficiency with respect to the control of CNS TLR4 and pain signaling. Furthermore, type 2 cytokines become prevalent during chronic TLR4 hyperstimulation in the CNS and are associated positively with behaviors that are usually linked to type 1 pro-inflammatory cytokines. These findings question the notion that type 2 immunity is solely anti-inflammatory in the CNS and indicate that type 2 immunity induces/potentiates CNS inflammatory processes.

A potential test system for detecting contaminations by bacterial lipoproteins

Biological specimens are often contaminated with bacteria-derived products such as LPS or lipoproteins (LP), which trigger unwanted inflammatory responses in hosts. Whereas a contamination by LPS can be determined by various test systems, a contamination by LP can as yet not be determined. TLR4 and TLR2 are key components of the LPS and the LP receptor complex, respectively. It was tested in this study whether HEK293 cell stably transfected with bovine TLR2 have the ability to react to low concentrations of diacylated and triacylated synthetic LP. The stable cell lines we present here recognize low concentrations of synthetic LP resembling LP of different bacteria. Therefore, these cells are suitable to detect low contaminations present in probes. For example, HEK293 cells stably transfected with bovine TLR2 recognized an egg albumin preparation as contaminated, as evidenced by copious production of IL-8. In contrast, these cells did not respond to a highly purified human serum albumin (HSA) preparation used in the clinic but responded to HSA containing small amounts of diacylated synthetic LP. The TLR4 ligand LPS is often said to activate TLR2. Here we present evidence that LP contaminations are responsible for TLR2 activity. HEK293 cells stably transfected with bovine TLR2 and TLR1 (e.g. clone 1) did not respond to ultra-pure Escherichia coli LPS preparations but acquired responsiveness when stimulated with differently purified commercial LPS. Thus, the described system involving HEK293 cells stably transfected with bovine TLR2 and TLR1 is the first test system described attempting to measure a contamination by LP.

Stable transduction of bovine TLR4 and bovine MD-2 into LPS-nonresponsive cells and soluble CD14 promote the ability to respond to LPS

The interaction of bovine cells with lipopolysaccharide (LPS) was explored using human embryo kidney (HEK) 293 cell line stably transduced with bovine toll-like receptor-4 (TLR4) alone or in combination with bovine MD-2. These lines and mock-transduced HEK293 cells were tested by flow cytometry for LPS-fluorescein isothiocyanate (LPS-FITC) binding, nuclear factor kappa B (NFkappaB) activation, interleukin-8 (IL-8) production and interferon-beta mRNA expression/interferon (IFN) type I production. Whereas bovine TLR4 was sufficient to promote binding of high concentrations of LPS-FITC, both bovine TLR4 and MD-2 were required for activation by LPS, as assessed by NFkappaB activation and IL-8 production. Induction of IFN bioactivity was not observed in doubly transduced HEK293 cells, and no evidence for IFN-beta mRNA induction in response to LPS was obtained, although cells responded by IFN-beta mRNA expression to stimulation by Sendai virus and poly-inosinic acid-poly-cytidylic acid (poly(I:C)). Cells stably transduced with both bovine TLR4 and bovine MD-2 responded to LPS by IL-8 production, in decreasing order, in the presence of fetal bovine serum (FCS), of human serum, and of human serum albumin (HSA). The reduced activity in the presence of HSA could be restored by the addition of soluble CD14 (sCD14) but not of LPS binding protein (LBP). This is in contrast to macrophages which show a superior response to LPS in the presence of HSA when compared with macrophages stimulated by LPS in the presence of FCS. This suggests that macrophages but not HEK293 cells express factors rendering LPS stimulation serum-independent. Stably double-transduced cells reacted, in decreasing order, to LPS from Rhodobacter sphaeroides, to LPS from Escherichia coli, to synthetic lipd-IVa (compound 406), to diphosphoryl-lipid-A (S. minnesota) and to monophosphoryl-lipid-A (S. minnesota). They failed to react to the murine MD-2/TLR4 ligand taxol. This resembles the reactivity of bovine macrophages with regard to sensitivity (ED(50)) and order of potency but is distinct from the reactivity pattern of other species. This formally establishes that in order to react to LPS, cattle cells require serum factors (e.g. sCD14) and cell-expressed factors such as MD-2 and TLR4. The cell lines described are the first of a series expressing defined pattern recognition receptors (PRR) of bovine origin. They will be useful in the study of the interaction of the bovine TLR4-MD-2 complex and Gram-negative bovine pathogens, e.g. the agents causing Gram-negative bovine mastitis.

Gamma interferon is dispensable for neopterin production in vivo

Previous studies have indicated that neopterin is synthesized in vitro by human monocyte-derived macrophages and dendritic cells upon stimulation with gamma interferon (IFN-gamma). Neopterin production under specific conditions in vitro has also been obtained upon stimulation with IFN-alpha and/or IFN-beta. However, it is unknown if any IFN-gamma-independent neopterin synthesis is possible in vivo. In the present study we investigated the serum neopterin concentrations in patients affected by the syndrome of Mendelian susceptibility to mycobacterial disease (MSMD). Indeed, this syndrome is characterized by deeply impaired or absent IFN-gamma production or function due to severe mutations in molecules involved in IFN-gamma/interleukin-12 (IL-12)/IL-23-dependent pathway. Serum neopterin levels were measured by an enzyme-linked immunosorbent assay in 27 patients with MSMD. We found that serum neopterin levels are elevated in the complete absence of IFN-gamma activity due either to a complete deficiency of its receptor or to deleterious mutations of IL-12 or its receptor. These data clearly indicate that, as reported from in vitro studies, other stimuli are able to induce neopterin synthesis in vivo. Consequently, neopterin cannot be used as means of diagnosis of MSMD due to IFN-gamma-, IL-12-, and IL-23-dependent pathway defects.

Neopterin concentrations in blood donors differ between AB0 blood group phenotypes

OBJECTIVES

Neopterin is produced by human monocyte-derived macrophages upon stimulation with interferon-gamma and is therefore a sensitive indicator for cellular immune activation. Common factors like age, diastolic blood pressure, body mass index, or smoking habits were found to be associated with neopterin concentrations in humans.

DESIGN AND METHODS

In order to find possible genetic determinants which might influence neopterin production, we investigated 8288 consecutive blood donors after exclusion of samples suspicious of infections.

RESULTS

Donors with blood group phenotype 0 had moderately, but significantly (P < 0.0001) higher neopterin concentrations (mean +/- SD: 6.94 +/- 1.52 nmol/L) than those with phenotype A (6.75 +/- 1.50 nmol/L), phenotype B (6.73 +/- 1.48 nmol/L), and phenotype AB (6.68 +/- 1.57 nmol/L).

CONCLUSIONS

Neopterin levels are higher in donors with blood group phenotype 0 than in other phenotypes. Data point to a genetic background of different neopterin concentrations. However, alterations of neopterin levels were much less expressed than the changes known to occur during diseases with an activated immune response.

In vitro bioassay of endotoxin using fluorescein as a pH indicator in a macrophage cell culture system

Based on the biological activity of endotoxin, we propose a possible new method for detecting endotoxin using a pH-indication system of macrophage culture media. After RAW 264.7 macrophage cells were treated with lipopolysaccharide (LPS), the addition of fluorescein to the LPS-treated media reproductively reduced its absorption and emission spectra (it was a dose-dependent reduction). The advantages of this LPS-detection method were compared with the Limulus Amebocyte Lysate (LAL) test by using purified bacterial LPS (Salmonella minnessota, Escherichia coli, and Pseudomonas aeruginosa). Additionally, the absorption and fluorescence intensity of fluorescein, following treatment of RAW 264.7 cells with a high concentration of Staphylococcus aureus (Gram-positive, lysed bacteria), could not generally be detected by the LAL test, but they were found to be reduced, in a dose-response relationship, with this new system. The macrophage culture system-method might be a good supplement to the LAL assay for detection of LPS, Gram-negative and Gram-positive bacteria.

Level of neopterin, a marker of immune cell activation in gingival crevicular fluid, saliva, and urine in patients with aggressive periodontitis

BACKGROUND

Neopterin, a marker of cellular immune activation, is produced by human macrophages after induction by interferon gamma that is secreted by T lymphocytes. Neopterin concentrations in diverse body fluids have been reported to increase in parallel with bacteria in the clinical course of infections. Therefore, determination of neopterin in body fluids was thought to be useful for predicting the prognosis and diagnosis of aggressive forms of periodontal disease, in which the cell-mediated immune response plays an important role in immunopathogenesis. The aim of the present study was to observe the role of neopterin in the pathogenesis of aggressive periodontitis (AgP).

METHODS

Thirteen individuals who were systemically and periodontally healthy and 16 systemically healthy individuals diagnosed with AgP were recruited for this study. Mixed saliva and urine samples were collected from each subject. Gingival crevicular fluid (GCF) samples were obtained from 6 teeth with > or =5 mm probing depth (PD). After evaluation of GCF amount from paper strips, enzyme-linked immunosorbent assay (ELISA) was employed to determine the amount of neopterin in urine, saliva, and GCF.

RESULTS

The amount of neopterin in urine and saliva measured 235.77+/-405.31 micromol neopterin/mol creatinine and 9.85+/-7.66 nmol/l, respectively, for the AgP group and 225.45+/-100.72 micromol neopterin/mol creatinine and 5.25+/-5.76 nmol/l, respectively, for controls. The present data demonstrate that, while salivary neopterin levels were found to be significantly different between periodontitis and control subjects, there were non-significant differences in urine neopterin levels. The amount and concentration of neopterin in GCF measured was 18+/-12.75 nmol/l and 3.67+/-2.40 nmol/ml for the AgP group and 2.51+/-1.72 nmol/l and 3.88+/-4.50 nmol/ml for the control group. When total amounts of neopterin are taken into consideration, a significant difference between AgP and controls is shown; however, no significant difference in net concentration of neopterin was found between both groups.

CONCLUSIONS

This study is the first report to evaluate the involvement of neopterin in AgP and this might be considered of value in understanding periodontal disease mechanisms.

Interferon-gamma-primed monocytoid cell lines: optimizing their use for in vitro detection of bacterial pyrogens

In order to reduce animal testing for quality control of pharmaceutical agents intended for parenteral use, the Limulus amebocyte lysate (LAL) assay is now being accepted in many cases as an alternative to measuring pyrogenic activity of samples in rabbits. However, since the LAL test is specific for cell wall components from Gram-negative bacteria and is sometimes difficult to perform in samples containing large amounts of protein, this alternative still leaves a considerable diagnostic gap. Here, we have optimized a previously established test based on assessing the formation of neopterin or nitrite in interferon-gamma-treated human (THP-1) or murine (J774A.1, RAW264.7) monocytoid cell lines, respectively, in response to bacterial pyrogens. Optimal results were obtained either with THP-1 cells in serum-containing media and using a high concentration of interferon-gamma (IFN-gamma) or with RAW264.7 cells in serum-free media and independent of the IFN-gamma dose. Results were significantly correlated with those obtained by another cell-culture-based assay in which formation of tumor necrosis factor-alpha by THP-1 1G3 cells was assessed. Also in RAW264.7 murine monocytoid cells, formation of nitrite and of tumor necrosis factor-alpha in response to a variety of samples was correlated. Samples shown to be pyrogenic in rabbits in a previous study were unambiguously detected with the test presented here. As expected, the LAL test was negative with cell-free supernatants from Staphylococcus aureus66 kDa). Taken together, these results indicate that the use of monocytoid cell lines and the detection of metabolites which are triggered in the course of immunostimulation could fill the gap left by the LAL test and help to further reduce animal testing for pyrogens.

Suppressive effects of serum on the LPS-induced production of nitric oxide and TNF-alpha by a macrophage-like cell line, WEHI-3, are dependent on the structure of polysaccharide chains in LPS

The effect of serum on LPS-induced activation of a murine macrophage-like cell line, WEHI-3, was examined. Foetal calf serum strongly inhibited the production of nitric oxide (NO) and TNF-alpha by LPS-stimulated WEHI-3 cells, while it enhanced the production of both by other macrophage-like cell lines, J774.1 and BAM3, on treatment with LPS. This suppressive effect of serum on WEHI-3 cells was most remarkable when the cells were stimulated with rough-chemotype LPS, Ra LPS, Rc LPS and Rd2 LPS. Foetal calf serum also inhibited TNF-alpha production by the same cells stimulated with high concentrations of smooth-form LPS (S LPS; > 1000 ng/mL). Serum-mediated suppression was also observed for expression of the TNF-alpha gene in Rc LPS-stimulated WEHI-3 cells. This suppressive effect of FCS was most remarkable during the 1-2 h before the addition of LPS, but it was not observed when FCS was added at 1 h after the addition of LPS, suggesting dependence on the time of FCS addition to LPS-stimulated cells. No significant difference was observed in the expression of CD14 on WEHI-3 cells cultured in the presence and absence of serum, suggesting that CD14 is not involved in the serum-mediated suppression of these LPS-responses. On the contrary, FCS showed enhancing effects on the production of NO and TNF-alpha by WEHI-3 cells stimulated with low concentrations (< 100 ng/mL) of S LPS and rough mutant Salmonella minnesota Re LPS. These results suggest that the ability of FCS to suppress LPS-induced activation of WEHI-3 cells in mainly dependent on the structure of polysaccharide chains and also on the concentration of LPS employed.

Bacterial endotoxin alters kinetics of BK channels in rat cerebrovascular smooth muscle cells

Patch-clamp recordings were used to study the effects of Escherichia coli bacterial endotoxin (lipopolysaccharide, LPS) on the properties of large-conductance, Ca2+-dependent K+ channels (BK channels) in the membrane of enzymatically dispersed rat cerebrovascular smooth muscle cells (CVSMCs). LPS had negligible effects on the kinetic and conductance properties of BK channels when applied to the extracellular domain of these channels. However, acute application of LPS (10-100 microg/ml) to inside-out patches of CVSMC membrane isolated in a cell-free environment rapidly and reversibly increased the open probability of BK channels, leaving the conductance of these channels unaltered. The magnitude of this effect decreased as the concentration of free Ca2+ at the cytoplasmic membrane face was lowered, but was little affected by changes in membrane potential. Kinetic analysis showed that LPS accelerated reopening of BK channels while having little effect on mean channel open time. Detoxified E. coli LPS, from which the fatty acid chains of Lipid A were partially removed, showed slightly reduced activity when compared to the parent endotoxin molecule. A purified E. coli Lipid A had negligible effects on BK channel function. These results indicate that LPS activates BK channels in cerebrovascular smooth muscle cells when present at the cytoplasmic membrane face. This novel mechanism may provide insights into the regulation of BK channels by intracellular, membrane-associated elements.

Human monocytoid cell lines as indicators of endotoxin: comparison with rabbit pyrogen and Limulus amoebocyte lysate assay

The aim of this study was to develop an in vitro test system for pyrogenic substances. Three clones derived from human monocytoid cell lines, which were selected by their high sensitivity to lipopolysaccharide (LPS), were assessed for tumor necrosis factor (TNF) production. Their response to pyrogen-containing samples was compared with that in a Limulus amoebocyte lysate assay and the rabbit pyrogen test. We show here that the induction of TNF in these clones is a valid in vitro alternative to determine endotoxin in commercial preparations requiring pyrogenicity testing. Cell clones derived from Mono Mac 6 (MM6 2H8 and MM6 4B5) responded to sub-ng/ml concentrations of complete rough-strain and smooth-strain LPS, to ng/ml concentrations of diphosphoryl-lipid A, and to microgram/ml concentrations of monophosphoryl-lipid A and to detoxified LPS. Cells reacted to > or = 1 microgram/ml lipoteichoic acid by TNF production, and were relatively insensitive to toxic shock syndrome toxin-1 (TSST-1) and to muramyl dipeptide adjuvant peptide. The reaction pattern of a clone derived from THP-1 (THP-1 1G3) was in general, similar to that of the MM6 clones, except that THP-1 1G3 failed to react to diphosphoryl-lipid A. When tested on commercial samples destined for parenteral use, there was a close correlation between a sensitive Limulus amoebocyte lysate (LAL) test and the cell culture test on the one hand, and between the pyrogen test and the cell culture test on the other hand. The data suggest that this cell-based test is able to recognize pyrogens derived from gram-negative organisms in test samples with appropriate sensitivity and specificity. This test appears to be able to eliminate some of the false-positive data obtained in the LAL test.

Serum factors, cell membrane CD14, and beta2 integrins are not required for activation of bovine macrophages by lipopolysaccharide

The role of serum factors such as lipopolysaccharide (LPS)-binding protein (LBP) and of macrophage-expressed CD14 and beta2 integrins in the activation of bovine macrophages by LPS was investigated. Macrophage activation was determined by measuring tumor necrosis factor production, NO generation, and upregulation of procoagulant activity by LPS (Escherichia coli O55:B5) at concentrations of 100 pg/ml to 100 ng/ml. The 50% effective dose for LPS was 1 order of magnitude higher than that for activating human macrophages. Macrophages were activated by LPS in the presence of serum or in the presence of albumin demonstrated to be free of LBP. The capacity to react to LPS in the absence of LBP was not due to the acquisition of LBP during a previous culture in serum. It was then established which CD14-specific antibodies block LPS binding to monocytes. Among the CD14-specific antibodies recognizing bovine mononuclear phagocytes (60bca, 3C10, My4, CAM36, VPM65, CMRF31, and TUK4), the first four blocked the binding of LPS-fluorescein isothiocyanate to bovine monocytes at low concentrations. Anti-CD14 antibodies did not block LPS-mediated activation of bovine bone marrow-derived macrophages, monocyte-derived macrophages, and alveolar macrophages. This was observed in experiments in which anti-CD14 concentrations exceeded the 50% inhibitory dose by >30-fold (3C10 and My4) or >300-fold (60bca), as defined in the binding assay described above. Monocyte-derived macrophages from an animal deficient in beta2 integrins and control macrophages were activated by similar concentrations of LPS, suggesting that beta2 integrins are not important bovine LPS receptors. Thus, in bovine macrophages, LPS recognition pathways which are independent of exogenous LBP, of membrane-expressed CD14, and of beta2 integrins may exist.

Treatment of an infected murine macrophage cell line (J774A.1) with interferon-gamma but not tumor necrosis factor-alpha or live Mycobacterium intracellulare alone modulates the expression of adhesion molecules

In the present study, we evaluated whether the activation of a murine macrophage cell line (J774.1A) by treatment with recombinant murine tumor necrosis factor-alpha (rTNF-alpha) or recombinant murine interferon-gamma (rIFN-gamma) before or simultaneous with infection with Mycobacterium intracellulare would affect their ability to express lymphocyte function-associated antigen-1 (LFA-1) and to restrict growth and kill the ingested M. intracellulare. The data showed that the effect of lipopolysaccharide (LPS) in increasing the level of LFA-1 was the same in the presence or absence of M. intracellulare. The inability of M. intracellulare to affect the level of expression of LFA-1 was irrespective of the M. intracellulare to J774A.1 ratio. A significant increase in the expression of LFA-1 was observed when J774A.1 cells were prestimulated with IFN-gamma 1 day before the addition of the bacteria. The addition of IFN-gamma with M. intracellulare simultaneously, however, did not affect the expression of the adhesion molecules as compared with the IFN-gamma alone. Our results indicated no change in the level of LFA-1 on J774A.1 following exposure with TNF-alpha. We observed that preexposure with 10-10(4) IU/ml of TNF-alpha can significantly decrease the number of ingested M. intracellulare. Simultaneous addition of 10(3) and 10(4) IU/ml of TNF-alpha, however, did not have any mycobactericidal effect. This indicates that the TNF-alpha-induced killing by J774A.1 cells was relatively selective, depending on the concentration and the time of presence of TNF-alpha. The data may suggest that the uptake of M. intracellulare is carried out via other adhesion receptors when M. intracellulare and IFN-alpha are present simultaneously and that in the presence of TNF-alpha other surface receptors are involved in the uptake of M. intracellulare. Flow cytometry analysis of the spleen cells removed at various times from M. intracellulare-infected mice also indicated no change in the level of LFA-1 beta or MAC-1, a finding comparable with that of the J774A.1 cells.