granulocyte-macrophage colony-stimulating factor, tumour-necrosis

Effect of thyroid hormones and other iodinated compounds on the transition of monocytes into veiled/dendritic cells: role of granulocyte-macrophage colony-stimulating factor, tumour-necrosis factor-\g=a\and interleukin-6 P Mooij, P J Simons, M de H A Drexhage

Haan-Meulman,

Department of Immunology, Erasmus University, PO Box 1 738, (Requests for offprints should be addressed to H A Drexhage)

H

3000 DR

J de Wit and Rotterdam, The Netherlands

Abstract Stimulation of human peripheral blood monocytes with the thyroid hormones tri-iodothyronine (T3) and thyroxine (T4) enhanced their ability to mature into cytologically and functionally characteristic veiled/dendritic cells. Veiled/dendritic cell transition induced by T3 and T4 was dependent on the production of granulocyte-macrophage colony-stimulating factor (GM-CSF), tumour necrosis factor-\g=a\(TNF\g=a\)and interleukin-6 (IL-6) in the culture, since the addition of antibodies specific for GM-CSF, TNF\g=a\and IL-6 to the culture system had blocking effects. The addition of antibodies to macrophage colony\x=req-\ stimulating factor and IL-1 had no effects. Contaminating T cells and B cells did not contribute to the transition of monocytes to veiled/dendritic cells, and it is therefore likely that the GM-CSF, TNF\g=a\and IL-6 produced in the culture system were derived from the monocytes themselves.

Introduction called veiled cells when free in body al. 1979) are excellent antigencells presenting (Hart & Fabre 1981, van Voorhis et al. 1983, King & Katz 1990). The cells play a crucial role in the initiation of normal (Austyn 1987, Metlay et al. 1989) and autoimmune responses. Knight et al. (1988) showed that autoimmune thyroiditis could be induced in Balb/c mice by the transfer of as few as 10 dendritic cells pulsed with thyroglobulin (Tg). Furthermore, the development of autoimmune lesions, such as arthritis (Poulter et al. 1983, Zvaifler et al. 1985, Harding & Knight 1986), thyroiditis (Kabel et al. 1987, Voorbij et al. 1990) or insulitis (Voorbij et al. 19896) starts with an increase in the number of dendritic cells and a local clustering of these cells with other lymphoid ceDs before autoantibodies can be detected in the circulation. Dendritic cells

(also

fluids; Drexhage

et

Stimulation of the blood monocytes with an optimal concentration of metrizamide (14\m=.\5%), reverse T3 (rT3; 2 \m=x\10\m=-\10m) or highly iodinated thyroglobulin (Tg; 2 \m=x\10\m=-\11m) also resulted in an increased transition of

monocytes

to

veiled/dendritic cells, but

to a

lesser

extent

in comparison with the thyroid hormones (T3, 31 \m=+-\6% and T4, 25 \m=+-\5%vs rT3, 22 \m=+-\8% and Tg with an iodination grade of 0\m=.\37%:20 \m=+-\4% veiled/dendritic cells). Administration of anti-GM-CSF, anti-TNF\g=a\ and anti\x=req-\ IL-6 to the culture system also had blocking effects on the transition from monocytes to veiled/dendritic cells

induced by the iodinated compounds. The mechanisms by which such iodinated compounds act on the monocyte to veiled/dendritic cell transition can only be speculated on

(interference H2O2-generating system?).

journal of Endocrinology (1994) 140,

503\p=n-\512

The precursor of the dendritic cells has not yet been fully characterized. We (Kabel et al. 1989a) and others (Peters et al. 1987) have reported that cells with a veiled or dendritic morphology, a major histocompatibility complex (MHC) class-II positivity and weak to absent acid phosphatase positivity, can mature from cells belonging to the blood monocyte pool. It was possible to induce 30-40% of veiled/dendritic cells from blood monocyte fractions purified via elutriation centrifugation (over 98% non-specific esterase-positive cells) after exposure of these cells to 14-5% metrizamide for 30 min, followed by an overnight culture under non-adhering conditions (Kabel et al. 1989a). These monocyte-derived veiled/dendritic cells had an in¬ creased ability to act as stimulator cells in a mixed leucocyte reaction (MLR.) when compared with control non-pulsed monocyte populations (Kabel et al. 1989a). The veiled/dendritic cells were also less phagocytic and

expressed less CD14. Since metrizamide is an iodinated compound structurally related to the thyroid hormones tri-iodothyronine (T3) and thyroxine (T4) we speculated that thyroid hormones and maybe iodinated com¬ pounds in general play a role in enhancing the -

transition of monocytes to veiled/dendritic cells. The study reported here describes the effect of the thyroid hormones T3 and T4 and a few iodinated com¬ pounds besides metrizamide, such as reverse T3 (rT3, the hormonally inactive form of T3) and Tg, with high, low or no iodination on the maturation of human peripheral blood monocyte fractions into functionally active veiled/ dendritic cells. Since cytokines and, in particular, granulocyte-macrophage colony-stimulating factor (GMCSF) and tumour necrosis factor- (TNFa) have been described as essential for the differentiation and survival of blood dendritic cells in vitro (Witmer-Pack et al. 1987, Markowicz & Engleman 1990, Santiago-Schwarz et al. 1992), we also measured, in a series of experiments, the effect of the administration of antibodies specific for -

GM-CSF, TNFa, interleukm-6 (IL-6), macrophage factor (M-CSF) and IL-lß veiled/dendritic cell transition.

colony-stimulating monocyte

to

on

the

Materials and Methods blood monocytes Peripheral blood monocytes from healthy volunteers (who gave written informed consent) were isolated by Ficoll— Isopaque density gradient centrifugation, followed by Percoli density gradient centrifugation according to meth¬ ods described in detail (Kabel et al. 1989a). The percentage of monocytes was determined by positive staining for the enzyme non-specific esterase (NSE) according to Mullink et al. (1979). The cell suspension contained 60—80% NSE-positive monocytes, the contaminating cells mainly consisted of CD3-positive cells. In some experiments, monocyte suspensions were obtained via elutriation centrifugation (Beckman J21 cen¬ trifuge with a JE-6 elutriation rotor) with a purity >98% NSE-positive monocytes (for technical details see Kabel et al. 1989a). Alternatively, Percoll-purified monocyte preparations were used. These had been freed from contaminating and cells using a magnetic cell separ¬ ation system (MACS; Miltenyi Biotec GmbH, Bergisch Gladback, Germany) with mouse anti-human CD3 (Becton and Dickinson, San Jose, CA, USA) and CD19 (Coulter, Electronics Nederland, Mijdrecht, The Netherlands) monoclonal antibodies coated onto biotinlabelled magnetic beads (Becton and Dickinson) via streptavidin (Zymed Laboratories Ine, San Francisco, CA, USA) and a goat anti-mouse biotin-labelled conjugate (Dakopatts, Copenhagen, Denmark). The MACS separ¬ ation yielded monocyte suspensions with a purity of >95% NSE-positive monocytes. The isolation

of human peripheral

Incubation of human peripheral blood monocyte suspensions with hormones or iodinated compounds to obtain veiled or dendritic cells From the purified suspensions, a cell suspension contain¬ 1 x 106 cells/ml was made in serum-free modified

ing

(Gibco Europe Ltd, Paisley, Strathclyde, UK) supplemented with SF-1 mixture (Costar Europe Ltd, Badhoevedorp, The Netherlands) and

Dulbecco's culture fluid

antibiotics. A total of 3-4 x 105 NSE-positive monocytes was incubated with culture fluid alone or stimulated under non-adhering conditions (polypropylene tubes; Falcon labware division of Becton and Dickinson, Lincoln Park, NY, USA; 37 °C, 5% C02, 100% humidity) for 30 min with the following stimuli: (a) metrizamide (14-5% (w/v),

Serva, Heidelberg, Germany) according to Kabel et al. 1989a; (b) T4, 3, rT3 (Sigma Chemicals Corporation, St Louis, MO, USA), (c) Tg with high (0-37%), low (0-2%)

or no

(0%) iodination (I), kindly provided by JJ M de

Vijlder (Academical Medical Centre, Amsterdam, The Netherlands), or (d) KI, Nal, NaI03 or KI04 (as controls). The concentrations of the thyroid hormones, the iodinated compounds and the iodine salts ranged from 2x 10~,8M to 2x 10"9m. After exposure to the hormones or iodinated com¬ the cells were washed three times in RPMI 1640 (5 min, 500 g) and further cultured in the serum-free culture fluid for a period of 16 h (overnight) under non-adhering conditions (polypropylene tubes; 37 °C, 5% C02, 100% humidity). To determine the influence of endogenously produced cytokines on the transition from monocyte to veiled/ dendritic cell, several concentrations of monoclonal anti¬ bodies, specific for GM-CSF (Genzyme Corporation, Kent, UK), TNFa (Genzyme), IL-6 (kindly provided by Dr W A Buurman, University of Limburg, Maastricht, The Netherlands), M-CSF (Genzyme) and IL-lß (Glaxo, Geneva, Switzerland) were added 2 h after starting the overnight culture period. The antibody remained present in the culture fluid during the further culture period (in a series of preliminary experiments this time-schedule proved to be optimal in order to reach the effects described in the Results section).

pounds,

Determination dendritic cells

of the percentage of monocyte-derived veiled

or

After the 16-h overnight culture period, cells were centri¬ fuged (5 min, 500 g), and the supernatant was removed and stored 20 °C for the measurement of cytokines produced (data not shown). The cells were resuspended and the morphology of the living cells was studied using light microscopy at a magnification of x 400. Large cells, with actively moving veiled cytoplasmic processes and/or dendritic extensions were considered to be veiled/ dendritic cells. —

The number of

veiled/dendritic cells counted verified using cytocentrifuge directly suspension preparations: 50 µ of the cell suspensions at a concen¬ tration of 0-5 x 10 /ml were cytocentrifuged (3 mm, 800 in

living

were

r.p.m.) using a Cytospin apparatus (Nordic Immunological Laboratories, Tilburg, The Netherlands). Cytocentrifuge preparations were air-dried for at least 1 h and fixed in 100%

acetone

for 10 min.

Thereafter, indirect immuno¬

peroxidase staining was performed according to a tech¬ nique described previously (Kabel et al. 1989a) using the monoclonal antibody HLA-DR at a dilution of 1:50 (specific for MHC class-II antigens; Becton and Dickinson). Cells with long cytoplasmic protrusions, a reniform nucleus, strong MHC class-II positivity, but absent or weak acid phosphatase reactivity were consid¬ ered to be the veiled/dendritic cells (this criterium has previously been used by us in fixed histological prepara¬ tions; Wilders et al. 1983, Kabel et al. 1987, 1989a, Voorbij et al. 19896, 1990). Purification oj the veiled Idendritic cell suspensions In a few experiments veiled/dendritic cells were enriched using the following technique which is based on the

characteristics of veiled/dendritic cells not to adhere spontaneously to surfaces but to move around in the culture fluid actively (Drexhage et al. 1979). After the overnight culture period under non-adhering conditions, the cell suspension containing monocytes and induced veiled/dendritic cells was washed (5 min at 500 g in polypropylene tubes) and cells were resuspended in RPMI 1640 supplemented with 10% fetal calf serum (FCS; Integro BV, Zaandam, The Netherlands). The cell suspension was placed in a Boyden chamber system divided by a polycarbonate microfilter (5 µ ; Nucleopore; polyvinylpyrrolidone-free; Costar Corpora¬ tion) in which the top chamber (polypropylene) contained the washed cell suspension in RPMI 1640, 10% FCS, and the bottom chamber RPMI 1640, 10% FCS and 10 nM of the chemoattractant n-formyl-methionyl-leucyl-

veiled/dendritic cells

capable of stimulating cells to proliferate, allogeneic MLR was performed (for details see Kabel et al. 1989a), using the veiled/dendritic cells as an

stimulator cells. To obtain the responder population, lymphocyte isolation was performed with subsequent Ficoll—Isopaque and Percoli density gradient centrifugation (the pellet containing and lymphocytes and NK cells) and nylon wool adherence (Polysciences Ine, Warrington, PA, USA) (Julius et al. 1973). The non-adhering cells were >90% positive for the marker CD3. The mixed leucocyte cultures were performed in triplicate in 96-well flat-bottomed microtitre plates (Falcon). The following were used as stimulator cells (X-irradiated, 2000 RAD), (a) Non-cultured freshly isolated Percoli monocytes, kept at 4 °C, not contain¬ ing any cells with a veiled/dendritic morphology; (b) Percoll-purified monocytes cultured overnight at 37 °C under non-adhering conditions (containing some veiled/ dendritic cells); (c) Percoll-purified monocytes cultured overnight at 37 °C under non-adhering conditions after a pulse with T3, T4, rT3 or the iodinated compounds (containing raised numbers of veiled/dendritic cells), in some instances these populations were enriched for veiled/ dendritic cells by the Boyden chamber method (see above); (d) blood veiled/dendritic cell populations ob¬ tained according to the well-established separation method described by Knight et al. (1986) (control). Stimulator cells (3 x 104) were cultured with 1-5 x 105 allogeneic lymphocytes in a total volume of 200 µ RPMI 1640 supplemented with 10% human blood group A serum and antibiotics for 5 days at 37 °C, 5% C02. X-irradiated monocytes/dendritic cells (3x10) alone or 1-5 x 10 lymphocytes in the presence of 10—50 µg phytohaemagglutinin/ml (PHA; Welcome Diagnostics, Dartford, Kent, UK) served as controls. Stimulator and

responder populations

volunteers.

were

from the various

healthy

Subsequently, the cells were collected from the top and bottom chambers. The percentage of cells with a veiled/ dendritic morphology was counted in both cell suspensions as described above under light microscopy (magnification 400 x ). The top chamber contained very high percentages of veiled/dendntic cells (see Results).

During the last 16 h, 0-5 µ [ HJthymidine (Amer¬ sham Nederland BV, s'Hertogenbosch, The Netherlands), was added to each well; the cells were harvested on filter paper and [3H] thymidine incorporation was measured in a liquid scintillation analyser (Packard Tricarb 2500 TR). MLR data are given as proliferation indices (PI). These values were calculated by dividing the amount of incor¬ porated [' H]thymidine (c.p.m.) in lymphocyte suspen¬ sions which had been stimulated with cell suspensions containing monocytes and/or veiled/dendritic cells as stimulator cells by the amount of incorporated ['HJthymi¬ dine (c.p.m.) in the same lymphocyte suspensions in the absence of any stimulating cell suspension.

Mixed

Statistical

phenylalanine (fMLP; Sigma). Cells in the top chamber

were

allowed

to

adhere

to

the

polycarbonate filter and migrate through the filter towards the fMLP gradient in the bottom chamber during a 2-5-h incubation period (37 °C, 5% C02, 100% humidity).

leucocyte reaction To verify whether the monocyte-derived cells with a veiled or dendritic morphology were functionally active

analysis

Differences in monocyte to veiled/dendritic cell transition were analysed using the Wilcoxon's rank sum test.

Differences in the stimulating capabilities of monocyte or veiled/dendritic cell populations in the MLR were tested using a paired binomial test. Values of P90% (trypan blue exclusion). When monocytes were kept as a control at 4 °C overnight, veiled/dendritic cells could not be identified in the cell suspension. An exposure of the monocytes to the thyroid hormones T3 and T4 prior to the overnight culture at 37 °C enhanced the capability of the blood monocytes to mature into veiled/dendritic cells in a dose-dependent manner (Fig. 1). A stimulation of the blood ~10 monocytes with x 10 concentrations or m) resulted in (2 T3 T4 optimal 30-5 ±5-7% or 24-5 ±4-8% veiled/dendritic cells respec¬ tively. Thyronine (T0; the non-iodinated backbone mol¬ ecule of thyroid hormones) was unable to enhance the transition of monocytes into veiled/dendritic cells (Fig. 1). At best (at a T0 concentration of 2 x 10 ~10 m) 13-5 ± 3-4% veiled/dendritic cells were found; this value was not significantly different from the spontaneous maturation during incubation with serum-free culture fluid (repre¬ sented by the shaded area in Fig. 1). In four separate experiments, KI, Nal, NaI03 and KI03 also failed to stimulate the maturation of mono¬ cytes into veiled/dendritic cells when tested at various

concentrations (2-200 x 10"n m). In two series of experiments, Percoll-purified monocyte suspensions that had been freed of contaminating cells and cells (using MACS, yielding over 95% NSE-positive

monocytes), or elutriator-purified monocytes (yielding over 98% NSE-positive) were used. In these experiments, the thyroid hormone T3 also gave a clear enhanced monocyte

to

veiled/dendritic cell transition and 41%

(MACS separation) and 29 ±10% (n 10, elutriator puri¬ fication) veiled/dendritic cells were found respectively, =

m being the optimal T3 concentration (data not These experiments using almost pure monocyte suspensions showed that and cells contaminating the monocyte population or products of these and cells are not necessary for the transition of monocytes to veiled/ dendritic cells. Pretreatment of the monocytes with metrizamide prior to the overnight culture increased the monocyte to veiled/dendritic cell transition as expected, and 26-8 ±2-9% veiled/dendritic cells were found (Fig. 1), as has been reported previously (Kabel et al. 1989a) (although in cultures containing 10% FCS). rT3 and both iodinated forms of Tg -were also capable of enhancing the monocyte to dendritic cell transition, although to a somewhat lesser extent when compared with the thyroid hormones and metrizamide (Fig. 1). At ° m rT3 (optimal concentra¬ a concentration of 2 x 10 tion), 22 ±7% veiled/dendritic cells were found. With regard to the iodinated forms of Tg, at a concentration of 2xl0"nM Tg 0-2% I (optimal concentration) 22 ± 2% veiled/dendritic cells were found, and at a concentration of 2x10~um Tg 0-37% I (optimal concentration) 20 ± 4% veiled/dendritic cells were found. Non-iodinated Tg was almost incapable of stimulating the transition from monocyte to veiled/ dendritic cells, and low values of 19 ±3% and 17 ±2% veiled/dendritic cells were the maximum levels found (Fig. 1). Note, however, that the transition was induced 5 and at much lower concentrations (namely 2x10 2 xlO-12 m Tg). At these concentrations the iodinated Tgs gave similar values, and one may wonder whether this transition was induced, for instance, by the ingestion of the large Tg molecule.

shown).

Immunocytochemistry verify the character of the veiled/dendritic cells directly in the wet cell suspensions, immuno¬ cytochemical studies were performed on fixed cytocentri¬ fuge preparations of the cell populations after overnight culture. Cells with cytoplasmic protrusions, strong MHC class-II positivity and weak or absent acid phosphatase activity were considered as veiled/dendritic cells in the To

counted

fixed and stained preparations. A strong correlation was found (r=0-87, P