JOURNAL OF CELLULAR PHYSIOLOGY 142552-656 (1990)

Tumor Necrosis Factor-Alpha Induces Vitamin D-1-Hydroxylase Activity in Normal Human Alveolar Macrophages A. M. PRYKE, C. DUCCAN, C. P. WHITE, S. POSEN, AND R. S. MASON* Departments of Medicine and Physiology, University of Sydney, N e w South Wales 2006, Australia (A.M.P., C.P.W., S.P., R.S.M.), Department of Respiratory Medicine, Royal North Shore Hospital, St. Leonards, New South Wales 2065 Australia (C.D.) The induction of 1-hydroxylase in alveolar macrophages by tumor necrosis factor-alpha (TNF) was examined in view of recent evidence suggesting that local production of 1,25-(OH),D3 may play a role in the regulation of immune functions. Incubation of pulmonary alveolar macrophages from normal human subjects with recombinant TNF caused a 2- to 10-fold increase in 25-hydroxyvitamin D,-1 -hydroxylase activity. The dose-response curve was linear over the range 0.05-5.0 IUhnl, and no further increase was seen at higher concentrations. The increase in 1 -hydroxylase activity was present after 12 h and reached a maximum after 3 days. The effect of TNF was inhibited in a dose-dependent manner by the presence of 1,25(OH),D3 (1 O-’o-10p8 M) in the incubation media for 5 days but was unaffected by 1Op9 M 1,25(OH),D, after 12 h. The enhancement of macrophage 1-hydroxylase activity by TNF was comparable to that induced by gamma interferon (IFN) but the effects of maximal doses of both agents were not additive. The presence of antibody to TNF resulted in a 76% inhibition of TNFinduced 1-hydroxylase but had no significant effect on IFN-induced 1 -hydroxylase activity.

Several cytokines cause alterations in macrophage function. One of these, gamma interferon (IFN), which is produced mainly by T lymphocytes but under some circumstances by macrophages themselves (Robinson et al., 19851, increases the capacity of alveolar and bone marrow macrophages to metabolize 25-hydroxyvitamin D, [25(OH)D,l to the active compound 1,25-dihydroxyvitamin D, [1,25(OH),D,] (Adams and Gacad, 1985; Koeffler et al., 1985; Reichel e t al., 1987a). Tumor necrosis factor-alpha (TNF), a peptide produced by activated macrophages (Mannel et al., 19801, which is cytotoxic for a range of tumor cells but not normal cells (Ruff and Gifford, 1981; Old, 1985), may also affect the function of immune cells in a number of ways. TNF treatment of undifferentiated myeloid cells increases nonspecific esterase activity, enhances antibody-dependent cell-mediated cytotoxicity, and increases the ability of these cells to reduce nitro blue tetrazolium (Trinchieri et al., 1986). In view of the suggestion that IFN and TNF activate macrophages through a similar mechanism (Carlsen and Prydz, 1988), we investigated (1)the effects of TNF on 25-hydroxyvitamin D-1-hydroxylase activity in cultured alveolar macrophages, and (2) whether the effects of TNF and IFN on induction of 1-hydroxylase were interdependent.

MATERIALS AND METHODS We obtained 25(OH)-[3H-26,27-methyl]D, (20.6 Ci/ mmol) and 1,25(0H),-[3H-26,27-methyllD, from Amersham Australia Pty. Ltd. (Sydney, N.S.W., Australia); 0 1990 WILEY-LISS, INC

25(OH)D, and 1,25(OH),D, were a gift from Dr. D. Kingston (Roche Pty. Ltd., Dee Why, N.S.W., Australia). All sterols were stored at -20°C under argon in spectroscopic grade ethanol (Ajax Chemicals, Sydney, N.S.W., Australia). Other solvents were high-performance liquid chromatography (HPLC) grade and were obtained from Millipore-Waters (North Ryde, N.S.W., Australia). Recombinant human gamma interferon was obtained from Amersham Australia Pty. Ltd., and diluted in 0.50 M glucose. Recombinant tumor necrosis factor-alpha and a neutralizing antibody to TNF (1U of anti-TNF neutralizes 1 U TNF) were obtained from Bohringer Mannheim. Dulbecco’s modified Eagle’s medium (DMEM), Iscove’s modification of Dulbecco’s medium, glutamine, and fetal calf serum (FCS) were obtained from Flow Laboratories Inc. (Sydney, N.S.W., Australia). Gentamycin was obtained from David Bull Laboratories Pty. Ltd. (Mulgrave, Vic., Australia), streptomycin and crystalline penicillin from Glaxo Australia Pty. Ltd. (Boronia, Vic., Australia), transferrin from Calbiochem-Behring Company (Kingsgrove, N.S.W., Australia), and ethanolamine from Sigma Chemical Co. (St Louis, MO, U S A . ) .

Received June 6, 1989; accepted November 8, 1989.

*To whom reprint requestskorrespondence should be addressed.

TNF INDUCES VITAMIN D-l ACTIVITY IN MACROPHAGES

BRONCHOALVEOLAR LAVAGE Bronchial lavage procedures were performed on nine normal subjects with a fiberoptic bronchoscope (Pentax, Sydney, Australia) on nonsmoking normal volunteers. All subjects gave informed consent. The project conformed to the guidelines set down by the National Health and Medical Research Council of Australia for human experimentation and had the approval of the Medical Ethics Committee at Royal North Shore Hospital. Lavage samples, consisting of 4 x 50 ml aliquots of 0.9% w/v saline were obtained from the right middle lobe. Blood samples were taken on the day of lavage from all volunteers. Serum calcium, serum 25(OH)D3, and serum 1,25(OH),D3 concentrations, estimated as previously described (Seshadri et al., 1985), were normal in all subjects.

653

counted in 5 ml of Instanel scintillation fluid (Canberra Packard Internatiocal S.A., Zurich, Switzerland) on a Tri-Carb 1500 Liquid Scintillation Analyzer (Canberra Packard). This methanol/methylene chloride (3 : 97, v/v) solvent system, which was used routinely in this study, has been shown to distinguish 1,25(OH),D, from 19- nor-10 keto derivatives of 25(OH)DZ1(Cohen and Gray, 1985). Material co-eluting with authentic 1,25(OH),D, in this system was further identified in two other HPLC systems: (1)isopropanol/hexane solvent (9 : 91, v/v) and a 4.6-mm ID x 25-cm Beckman Si Ultrasphere (5-pm) silica column (Beckman); and (2) methanoliwater (80 : 20, v/v) and a 4.6-mm ID x 25-cm Zorbax ODS reversephase octadecyl-bonded column (Du Pont Company, Wilmington, DE, U.S.A.).

MACROPHAGE ACTIVATION WITH IFN AND TNF CELL CULTURE Macrophages were treated daily for up to 6 days prior The alveolar washings were centrifuged (2,OOOg for to 1-hydroxylase assay with either IFN (1,000-2,500 10 min); resuspended in 10 ml culture medium consist- IUlml) or TNF (0.05-50 IUlml). For time-course studing of DMEM containing 10% FCS, supplemented with ies, cells were pretreated with 50 IU TNF for 12-96 h 40 pg/ml gentamycin sulfate, 150 pgiml streptomycin, before 1-hydroxylase activity was measured. To deter150 pg/ml cr stalline penicillin, 30 pg/ml transferrin, mine whether the effects of TNF (50 IU/ml for 5 da s) and 2 x 10- BM ethanolamine; and plated into 24 well were inhibited by the exogenous 1,25(OH),D3, 10- culture plates (Flow) a t a density of approximately 1.75 M (final concentrations) of this metabolite were x lo5 cells/cm2. After 24 h, nonadherent cells were added daily for 5 days (long-term inhibition) or 12 h removed from cultures by gentle washing in culture (short-term inhibition) before the 1-hydroxylase assay. medium. Culture medium was changed daily. To idenTo determine whether the action of TNF was additify macrophages, representative wells were stained af- tive to that of IFN, macrophages were preincubated ter 5 days using a mononuclear cell-specific alpha with IFN (2,500 IUlml) and TNF (50 or 500 IU TNFlml) naphthylacetate esterase stain kit (Sigma). At this for 6 days before 1-hydroxylase activity was measured. time, >98% of cells were esterase positive. In order to determine whether IFN induced l-hydroxylase activity through TNF production, macro1-HYDROXYLASE ASSAY phages activated with either 5 IU TNF/ml or 2,500 IU After 8 days of culture, macrophages were washed IFNlml were incubated with and without 10 IU of a three times with Iscove’s modification of Dulbecco’s neutralizing antibody to TNF. Untreated macrophages medium, then incubated with 500 pl of Iscove’s con- were also incubated both with and without this antitaining 20 nM (0.25 pCi) L3H]25(0H)D3(specific activ- body. ity 20.5 pCi/nmol) for 5 h a t 37°C in a n atmosphere of RESULTS 5% carbon dioxide in 95% air. At the end of the incubation period, the medium was removed and subjected Incubation of normal alveolar macrophages for 6 to lipid extraction as previously described (Mason et days with 50 IU/ml TNF resulted in a n average foural., 1984). The extract was dried under nitrogen, recon- fold increase in 1,25(0H),D3 production (Fig. 1A). The stituted in 400 pl of spectroscopic grade ethanol, and response to TNF in the nine subjects examined ranged stored under argon a t -20°C. In order to determine the from a 2- to 10-fold increase in 1-hydroxylase activity. rate of nonspecific tracer breakdown, “blanks” (incuba- The incubation of these cells with 2,500 IU/ml IFN tions carried out in the absence of cells) were run in resulted in a n average fivefold increase in 1,25(OH),D3 each assay. production (Fig. 1A). Macrophages from one subject treated with both TNF (50 IU/ml or 500 IU/ml) and HIGH-PERFORMANCE LIQUID IFN (1,000 IU/ml) produced no more 1,25(OH),D3 than CHROMATOGRAPHY did macrophages treated with IFN alone (Fig. 1B). In High-performance liquid chromatography was used macrophages from a second subject, the addition of 100 to analyze the reaction products. Extract (200 pl) was IU/ml TNF to cells pretreated with IFN (1,000 IUlml) dried under nitrogen and reconstituted in 200 pl of 3% for 6 days did not increase 1,25(OH),D3 production afmethanol/97% methylene chloride (v/v). Samples were ter 12 h. Production rates of 1,25(OH),D3 in these cells chromatographed using a Waters HPLC system with were 218 20 fmol/h/106 cells and 236 * 70 fmol/h/106 automated sample injection (Millipore-Waters). Then, cells, respectively. A 50% increase in 1-hydroxylase activity was detect185 pl of sample was applied to a 4.6-mm ID x 25 cm Beckman Si Ultrasphere (5-pm) silica column able after 12-h pretreatment with 50 I U TNF, while (Beckman Instruments, Inc., San Ramon, CA, U.S.A.). the maximal 1-hydroxylase activity was reached after The column was eluted with methanoUmethylene chlo- 72 h (Fig. 2). In macrophages from a second subject, ride (3 : 97, v/v) a t a flow rate of 1 ml/min; 1-ml frac- 1-hydroxylase activity increased from 59 * 4 fmol/h/ tions were collected, evaporated overnight, and lo6 cells to 122 t 9 fmol/h/106 cells after 12-h pretreat-

T*

*

654

PRYKE ET AL

B

A

untreated

2500IU IFNlrnl

50IU TNF/ml

T

T

T

-

-

-

10WIU IFN/ml

50IUTNF

+EN

500IUTNF +IFN

Fig. 1. 1-Hydroxylase activity of normal human alveolar macrophages after incubation with TNF, IFN, or combinations of the two agents. A: TNF 50 IUiml or IFN 2,500 IUiml were added daily for 6 days. Bar heights show the means ?SEM for results from 8 (IFN) or

9 (untreated, TNF) subjects. B: Macrophages from one subject treated with IFN 1,000 IUiml for 6 days also received TNF 50 IUiml for the final 4 days or 500 IUiml for 12 h prior to the 1-hydroxylase assay. Bar heights show the means ? 1 SD of triplicate determinations.

ment with TNF. The effect of TNF was dose dependent over the range 0.05-5 IU/ml (Fig. 3), but higher concentrations of TNF (5500 IU/ml) caused no further increase in 1-hydroxylase activity (data not shown). M) 24 h The addition of 1,25(OH),D, (lO-'O-lO-s after the addition of TNF (thus both agents were present for 5 days before 1-hydroxylase activity was measured) led to dose-dependent suppression of l-hydroxylase activity (Table 1).When 1,25(OH),D3 (lo-' M) was added 12 h before the assay, no decrease in 1-hydroxylase activity was noted, although lo-' M 1,25(OH)2D, caused a small and variable (0-30%) decrease within 12 h. No 24-hydroxylase activity could be detected in these cells, regardless of the incubation conditions. The stimulatory effect of TNF was diminished when 10-IU anti-TNF antibody was added at the same time as TNF (Table 2). Anti-TNF antibody had no effect on IFN-induced 1-hydroxylase activity (Table 2).

TABLE 1. Inhibition of macrophage 1-hydroxylase by 1,25-(OH),D,'

DISCUSSION It is now generally accepted that 1,25(OH),D3, which affects the function of several types of immune effector cells, including T and B lymphocytes, monocytes, and macrophages (Rigby, 1988), may be produced locally by

Subject Subject 1 fmoles/h/106 cells % inhibition Subject 2 fmoles/hi106 cells % inhibition

Untreated 288

?

51

-

198 t 3 -

[1,25-(OH),D31 M M

lo-''

f 74

lo-@M

244 ? 14 16

214

26

58

187 i 1

118 t 25'

120 2 202

6

41

40

123

f 50'

'TNF-treated macrophages (50 IUiml for 6 days) were incubated with and without 1,25-(OH),D3for 5 days prior to the measurement of 1-hydroxylaseactivity. Results presented are the means 2 1 SD of triplicate determinations. 'Significantly different from control values (P