presentation of exogenous antigens by macrophages: analysis of major histocompatibility complex...

Eur. J. Immunol. 1994. 24: 2421-2428 Presentation of exogenous Ag with MHC I and I1 by M@ 2421

Magdalena Kovacsovics-Bankowsko and Kenneth L. Rock

Division of Lymphocyte Biology, Dana Farber Cancer Institute Boston, and Department of Pathology, Harvard Medical School Boston

Presentation of exogenous antigens by macrophages: analysis of major histocompatibility complex class I and I1 presentation and regulation by cytokines*

There is an antigen presenting cell (APC) in the lymphoid organs capable of presenting exogenous antigen (Ag) with major histocompatibility complex (MHC) class I molecules. This study was initiated to isolate clones of these APC to definitively establish their phenotype and to further study their properties. Murine bone marrow macrophages (BM MQ) were immortalized by overexpressing myc and ruf oncogenes. Five BM MQ cell lines were generated that are phagocytic and expressed at their surface M@ differentiation Ag. All five cell lines processed and presented exogenous ovalbumin (OVA) with MHC class I molecules. They all presented OVA-linked to a phagocytic substrate 102-104-fold more efficiently than soluble Ag. Clonal isolates of two of the M@ cell lines had an identical phenotype and functional properties as the uncloned lines. These results definitively establish that MQ are APC with the capacity of presenting exogenous Ag with MHC class I molecules. Interferon (1FN)-y interleukin-4, granulocyte-macrophage colony stimulating factor and lipopolysaccharide either alone or in combination induced little or no augmentation and in some cases decreased presentation of exogenous OVA with MHC class I. In contrast, all of M a activating factors increased MHC class I expression. Moreover, IFN-y increased the presentation of cytosolic OVA, demonstrating differences between the presentation of cytosolic Ag versus exogenous Ag with MHC class I. Finally, some lines constitutively processed and presented exogenous OVA with MHC class I1 while others only presented after stimulation with IFN-y. These results demonstrate that the pathways involved in the presentation of exogenous Ag with MHC class I and class I1 are independently regulated and that a cloned cell is capable of presenting exogenous Ag through both pathways.

1 Introduction

It is generally accepted that exogenously acquired Ag are taken up by specialized APC, processed and displayed at the cell surface on MHC class11 molecules [l-21 while endogenously synthesized Ag in most cells are presented by MHC class 1 molecules [3-41. This segregation of Ag- presenting pathways is presumably important for the preservation of the immunological identity of healthy cells [5]. However, it has been unclear how a CD8+ Tcell immune response is generated against pathogens that reside in phagolysosomes of MQ [6] and why sometimes exogenous Ag prime CTL responses [7-111 .These phenom- ena may be explained by a novel Ag presenting pathway in a subset of APC capable of processing and presenting exogenous Ag with MHC class I molecules [12, 131.

Although, in most cells, exogenous Ag are excluded from MHC class I presentation pathway [14-161, an APC with

[I 130401 * This work was sponsored by NIH Grants A120248 and

AI31337. Recipient of a fellowship from the Swiss National Science Foundation.

Correspondence: Magdalena Kovacsovics-Bankowski, Division of Lymphocyte Biology, Dana Faber Cancer Institute, 44 Binney Street, Boston MA 0211.5, USA (Fax: 617-6322627)

Key words: Macrophage / Major histocompatibility complex class I I Antigen processing I Cytokines

this activity has been described [12]. These cells are active in vivo, indicating that this Ag presentation pathway is operative under physiological conditions [ 171. The APC that mediate this function are present in the spleen, thymus and peritoneum and copurify with MQ [17,18].These cells are phagocytic and they present Ag bound to particles up to 10 000-fold more efficiently than soluble Ag [ 191. Although these findings suggest that this APC is an M@, other cell types are phagocytic [20] and it is difficult to exclude the possibility that this form of Ag presentation is mediated by some other APC contaminating M a populations purified by conventional techniques. There is also limited evidence that this Ag-presenting pathway is only active in a small subset of cells in purified MQ populations [21]. The studies in this report were initiated to obtain a clonal source of cells with this Ag-presenting activity to definitively establish their identity. A preliminary report of the activity of two such cell lines has been published [ 191. In the present report we generalize these findings to additional cell lines and present evidence indicating that these cells are indeed M a . These cells are then analyzed in parallel with primary McD to further investigate their Ag-presenting capabilities.

In previous studies, APC capable of presenting exogenous Ag with class I were depleted from primary cell populations after treatment with anti-MHC class I1 mAb and comple- ment [12, 171. This raised the possibility that a single cell might be able to present exogenous Ag by both MHC class I and class I1 pathways which could be important to T-T collaboration [ 121. An examination of whether exogenous Ag could be processed by both the MHC class I and class I1

0 VCH Verlagsgesellschaft mbH, D-694.51 Weinheim, 1994 0014-2980/94/1010-2421$10.00 + .2.5/0

2422 M. Kovacsovics-Bankowski and K. L. Rock Eur. J. Immunol. 1994. 24: 2421-2428

pathways in MiP was not possible without a homogenous source of these cells. With the isolation of clones of these APC this analysis became feasible and is reported herein.

MQ are a component of the natural immune system and constitutively express some activities and effector functions. Unstimulated MQ are phagocytic and will ingest and destroy some microorganisms. After a productive interac- tion withT lymphocytes, M@ are stimulated to express new and/or greater levels of effector functions. T cells activate MiP by secretion of several stimulatory cytokines, including IFN-y , IL-4 and granulocyte-macrophage colony stimulating factor (GM-CSF) [21]. Moreover, many of these factors have been documented to affect the MHC classI- and 11-presenting activity of MQ or other cells [22-231. In addition, some microbial products, such as LPS, have MiP-activating activity [24]. It was, therefore, of interest to examine the effect of these activating factors on the Ag presenting activities of MQ that present exogenous Ag with MHC class I.

2 Materials and methods

2.1 Mice and cell lines

Female C57BL/6 mice, 5-8 weeks old, were purchased from the Jackson Laboratories, Bar Harbor, ME. RF33.70 (C57BL/6 anti-OVA-plus KbX BW.CD8.7) has been previously described [25]. MF2.2D9 (C57BL/6 anti- OVA + IAb) was kindly provided by Dr. M. Michalek (Dana Farber Cancer Institute). LB27.4 is a H-2bxd B lymphoblastoid cell line [26] and EL4 is a C57BL/6 T lymphoma. NIH 52 Leuk cells which produce a murine leukemia virus containing raf and myc oncogenes were kindly provided by Dr. U. Rapp (NIH, Bethesda, MD) ~ 7 1 .

2.2 Mcf, preparation and activation

BM MiP cell lines were established as previously described [27]. Briefly, BM cells were harvested from femurs and tibias of C57BL/6 mice and mononuclear cells were isolated by centrifugation on a Ficoll-Hypaque gradient. BM mononuclear cells (5 x lo6) were incubated with SN (50%) from NIH 52 Leuk cells. After 24 h at 37 "C, the cells were washed and resuspended (106/ml) in DMEM high glucose supplemented with 10% FCS, L-glutamine and Fungi-Bact solution and cultured in 24-well tissue culture plates. After day 30, the cells were grown on Cytodex beads (Pharmacia LKB Biotechnology, Uppsala, Sweden) and after 6-7 weeks were adapted to grow in stationary plastic tissue culture-flasks from which they were recovered by scraping. After day 60, the cell lines were used for analysis. The BM MQ, cell lines A3.1A7 and C2.3 were further subcloned by limiting dilution.

Peritoneal MQ were obtained by injecting C57BL16 mice i.p. with 1-1.5 ml of thioglycolate broth (Becton Dickin- son, Sunnyvale, CA). Three to four days later, cells were harvested by flushing the peritoneal cavity with 5 ml of RPMI 1640, 10% FCS. After centrifugation, cells were resuspended in culture media and used for different assays. These preparations are >95% MQ [18] and are referred to herein as peritoneal M a .

MiP were activated with 100 U/ml IFN-)I (Lee Biomolecu- lar, SanDiego, CA), 100 U/ml IL-4 (a gift from Drs. R. Tepper and A. Abbas, Harvard Medical School, Bos- ton, MA), 300 U/ml GM-CSF (kindly provided by Genetics Institute, Cambridge, MA) and/or 10 pg/ml LPS. The precise combinations of cytokines and the length of incubation are indicated in individual experimental proto- cols.

2.3 Immunofluorescence

Indirect immunofluorescence was performed as described [28] using mAb-containing culture supernatant followed by FITC-conjugated goat anti-rat Ig, not cross-reactive with mouse Ig (1: 40) (Cappel, Organon Teknika Corporation, West Chester, PA) or FITC-conjugated goat anti-mouse Ig (1:40) (Cappel). Samples were analyzed on a FACScan flow cytometer (Becton Dickinson). mAb were kindly provided by the laboratories of origin and/or obtained from the American Tissue Type Collection: anti-MAC-1 (M1/70) [29], anti-F4/80 [30], anti-FcyRII (2.4.G2) [31], anti-IAbId IEh'k (M5/114) [32], anti-Thy-1 (M5/49) [33] and anti- H2-Kb (Y3) [34].

2.4 Phagocytosis and pinocytosis

Cells (106/ml) adherent to glass cover slips, were incubated at 37 "C for 30 min with 3-pm Latex beads (5 x 106/ml) and then washed. Photomicrographs were taken at a magnifi- cation of x 1000.

Pinocytosis of lucifer yellow was quantified as described previously [35]. Briefly, BM MQ (106/tube) in Eppendorf tubes were incubated with PBS 10% FCS on ice for 30-45 min. PBS 10% FCS was then replaced by 350 pl of 1 mg/ml lucifer yellow and samples were incubated for different period of time either on ice or at 37 "C. Cells were then extensively washed with cold 137 mM NaCl, 3 mM KCl, 7 mM phosphate buffer pH 7.4, lysed in 0.5 ml of 0.1% Triton X-100 for 1 h at 37°C. Pinocytosed lucifer yellow was measured by combining 0.5 ml of lysate with 0.75 ml of 0.1% Triton X-100 and 100 pg/ml BSA, and then reading the fluorescence in a spectrofluorometer. Samples were excited at 430 nm and fluorescence was measured at 540 nm. Lucifer yellow (3.5 mg/ml) was loaded into the cytosol by osmotic lysis of pinosomes as previously described [15]. Samples were then handled as described above.

2.5 Cell culture

APC were incubated with Ag and T-T hybridoma as described [25], except that indomethacin (0.25 p ~ ) was added to the culture media. After 24 h, an aliquot of supernatant (100 pl) was collected, subjected to freeze- thawing and IL-2 content was assayed [36]. Antigenic constructs where OVA was covalently linked to iron oxide beads, were prepared as previously described 1191. OVA was loaded into the cytosol by osmotic lysis of pinosomes as previously described [ 151.

Eur. J. Immunol. 1994. 24: 2421-2428 Presentation of exogenous Ag with MHC I and I1 by MQ 2423

3 Results

3.1 Immortalization and phenotypic characterization of bone marrow cell lines

To establish definitively the phenotype of the APC capable of presenting exogenous Ag with MHC class I molecules and to provide a homogenous cell population for further studies, we followed the approach of Blasi et al. [27] and immortalized murine BM cells by overexpressing two oncogenes. Briefly, BM mononuclear cells were infected with a retrovirus carrying myc and ruf. Continuous growing cell lines were obtained after 6-8 weeks. The phenotype of the cell lines was then characterized by immunofluorescence analysis and a typical staining is shown for one clone in Fig. 1. The five BM cell lines expressed MAC-1 (M1/70), FcyRII (2.4.G2), F4/80, and MHC class I1 molecules (M5/114). They were negative for surface Ig and Thy-1.2 (M5/49). After 30 min incubation with 3-ym latex beads, most of the cells were phagocytic having ingested three or more beads (Fig. 2A). Under the same conditions,very few B lymphoblastoid cells boundingested latex particles (Fig. 2B). This phenotypic analysis clearly establishes that the immortalized cells are MQ,.

3.2 Presentation of exogenous Ag with MHC class I by BM M a

Preliminary studies indicated that two MQ, cell lines can process and present exogenous Ag (OVA) with MHC class I molecules [19]. To generalize these observations we have isolated and analyzed three additional BM M a cell lines. When OVA was added to the extracellular fluids, all five cell lines stimulated the Kb + OVA257-264-specific T-T hybrid, RF33.70 (Fig. 3). Presentation of soluble OVA required generally from 50 to 400 yg/ml of Ag depending on the particular cell 1ine.We similarly observed that all five cell lines could present OVA-linked to a phagocytic substrate (iron oxide beads) and the presentation of this form of Ag was from 102-104-fold more efficient than of the soluble Ag. In control experiments, we find that several non-MQ, cell lines do not present either soluble or par-

BG slg 2.402 M5/114 M1/70 F4/80

Figure 1. Immunofluorescence analysis of immortalized BM M a cell line, BM A3.1A. Cells were stained by indirect immunofluorescence with the indicated mAb supernatants as described in Sect. 2.3. Staining with the FITC-labeled second-step reagent is indicated as BG.The phenotype of this clone is representative of all five BM MQ.

Figure 2. Phagocytosis by BM M a . Photomicrographs of (A) BM A3.1A MQ and (B) LB27.4 B Lymphoblastoid cell line, incubated with 3-pm Latex beads for 30min at 37°C and then washed. Magnifications is x 1OOO.

-f- OVA-Beads I+ OVAsoluble I

OVA [pg/ml]

Figure 3. Presentation of exogenous Ag with MHC class I molecules by BMMQ. Cultures were prepared using BM C2.3, panel A, BM A3.1A7, panel B, BM B5, panel C, BM B1.2, panel D, BM B4, panel E as APC (5 x 104/well) and RF33.70 (5 x 104/well) and the indicated amount of soluble OVA (open triangles) or OVA-beads (closed triangles) in 96-well plates (200 pl). Cultures were then handled as described in Sect. 2.3.

ticulate OVAwith class I, as expected ([12,19] and datanot shown). These results indicate that the ability to present exogenous Ag with class I is not limited to rare MQ, cell line.

The cell lines described above appear homogenous by flow fluorocytometry but they did not arise at clonal frequency. To verify that the Ag-presentation activity of these cell lines was mediated by an MQ, and not some other contaminating cell, two of the lines were subcloned by limiting dilution. Individual clones of the A3.1A7 and C2.3 cells were isolated and were also found to present soluble and


particulate OVA added to the extracellular fluids (Fig. 3A and 3B).These clonal isolates also expressed uniformly Md, markers (Fig. 1 and data not shown). These analyses allow us to definitively identify a MQ as a cell that can present exogenous Ag with MHC class I. The availability of such cells allowed us to further examine the activities of a single APC clone in studies described below.

3.3 The MHC class I presentation of Ag in the extracellular fluids by M a is slower than for osmotically loaded Ag

To characterize further the presentation pathway of exogenous Ag by MHC class I molecules, we analyzed the kinetics of MHC class I presentation of Ag added to the extracellular fluids and osmotically loaded into the cytosol. The generation of detectable peptide-MHC class I complexes requires 2 h when BM Md, are incubated with OVA-beads (Fig. 4A) or up to 6 h when soluble OVA is used as an Ag (Fig. 4B). In contrast, when OVA is loaded directly into the cytosol of BM M a by osmotic lysis of pinosomes, OVA-H-2b complexes are readily detectable at the cell surface at the earliest time point measured (1 h) (Fig. 4C).

We also examined the kinetics of presentation of the various forms of Ag using PEC as APC. When Ag was

BMM0 60

40

20

0 3 6 122550100

OVA-Beads 40

l ! ! r n 20 0 3 6 12 25 50100

OVA soluble i E B r n :aj-----J 100

0 3 6 122550100 3 6 122550100

400 300

200 300

200 100

100

n 0 3 6 122550100

-~ 1.5 3 6 122550

Hypertonic loaded OVA

Cells x 10-

Figure 4. Kinetics of the generation of surface peptide-MHC class I complexes from Ag in the extracellular fluids and in the cytosol. APC (BMA 3.1A7, A-C or PEC, D-F) were exposed to Ag in serum-free medium under the following conditions: (A and D) OVA-beads (10 kg/ml) continuously in the medium, or (B and E) soluble OVA (2.4 mg/ml) continuously in the medium, or (C and F) OVA (30 mglml) loaded into the cytosol by hypertonic lysis of pinosornes [15]. After 0 h (crosses), 1 h (open circles), 2 h (closed circles), 4 h (open triangles), 6 h (closed triangles) or 24 h (open squares) of incubation at 37”C, cells were fixed with 1% paraformaldehyde and their APC activity was assayed in culture with RF33.70 T cell hybrid. Cultures were prepared and handled as described in Sect. 2.5. In panels B, D and E the open circles, the closed circles and the open triangles are superimposed on the base line. In panel E, the closed triangles are also on the base line.

added to the extracellular fluids, the detection of OVA + Kb required at least 6 h for OVA-beads (Fig. 4D) and an overnight incubation for soluble OVA (Fig. 4E). When Ag was loaded into the cytosol, at least 2 h was required before detecting OVA-MHC class I complexes on the cell surface (Fig. 4F). Although, normal M a process and present all forms of the Ag to RF33.70 more slowly than BM M@, we observed, for both types of cells, the same rank order for the kinetics of processing and presentation of the different forms of Ag.

We also analyzed the amount of Ag that M a acquired when incubated with the various forms of OVA. To estimate the amount of OVA taken up from isotonic and hypertonic buffers, we measured the rate of uptake of the fluid phase marker lucifer yellow [35]. This allowed us to determine the amount of extracellular fluid internalized by the Md, and, based on the concentration of OVA in the buffers, to calculate the amount of OVA that would be acquired. Since OVA might not have behaved exactly as the fluid phase marker, the values obtained should be considered to represent only approximate values. To measure the amount of Ag acquired when cells were incubated with OVA-beads, we quantified by microscopy the number of beads containing a known amount of Ag that were internalized.

Under isotonic conditions, the BM MQ took up approximately 3.5 ng of OVA/min/106 cells, while the hypertonic loading of cells delivered approximately 180 ng of 0VA/1O6 cells. Thus, after 1 h, BM Md, accumulated the same amount of OVA as the cells which were hypertonically loaded for 10 min. In contrast, when conjugated to beads, the Ag was acquired 30-fold faster than soluble Ag (100 ng of OVA/min/106 cells).

3.4 Regulation of MHC class I and class I1 expression by lymphokines and bacterial lipopolysaccharides

Several activities of Md, are regulated by cytokines and bacterial products. We were interested in determining how these Ma-activating factors influenced the function of the

IFNy IL-4 LPS GM-CSF

- - a IAb 0

loo 10’ lo2 l o 3 1 0 ~ 1 0 ~ l o 2 1 0 3 100 10’ 102 1 0 3 loo 10’ lo2 lo3

Fluorescence Intensity -L

Figure 5. Effect of cytokines and LPS on MHC class I and I1 molecules expression on EM MQ. Panel A-F, BM C2.3 were incubated for 72 h with either IFN-y, IL-4 or LPS, Panel G-H, EM A3.1A7 were incubated for 36 h with GM-CSF as described in Sect. 2.2. Cells were then stained with Y3 (anti-H-2Kh), Panel A, C, E and G, or MY114 (anti-IA”), Panel B, D, F and H, followed by a second-step reagent. A total of 5 x 103-10 x lo3 cells were analyzed on a FACScan flow cytometer. Cells stained with the second-step reagent only are used as the background and are display in black. Staining of untreated cells is indicated by a thin broken line.

Eur. J. Immunol. 1994. 24: 2421-2428 Presentation of exogenous Ag with MHC I and I1 by M@ 2425

subset of M a which are capable of presenting exogenous Ag with MHC class I. In initial studies we examined the effect of IFN-y, IL-4, GM-CSF and LPS on MHC class I and class I1 expression. MHC class I expression was up-regulated after stimulation with each of the four factors (Fig. 5). IFN-y and LPS treatment induced the greatest up-regulation of MHC class I expression (Fig. 5A and E), while the effects of IL-4 and GM-CSF were less and comparable to one another (Fig. 5C and G). MHC class I1 expression was increased by IFN-y, LPS and GM-CSF (Fig. 5B, F and H) but not by IL-4 (Fig. 5D).The failure of IL-4 to up-regulate MHC class I1 expression may be a specific characteristic of the BM M a as IL-4 is a potent inducer of Ia molecules on other M a [37]. IFN-y and LPS induced the largest increase in MHC class I1 expression. MAC-1 expression, used as a control, was only slightly up-regulated by treatment with all of the factors except GM-CSF which did not affect its expression (data not shown). These results indicate that the BM M a are responsive to all four factors.

Table 1. Ability of different activating factors to modulate the capacity of BM M@ and PEC to present exogenous. Ag with MHC class I moleculesa)

Cells Activating Shift minimal factor(s) stimulatory concentration

Soluble Ag Phagocytic Ag

BM PEC BM PEC BM PEC BM PEC BM PEC BM BM PEC

IFN-y IFN-y IL-4 IL-4

GM-CSF GM-CSF

LPS LPS

IFN-y + IL-4 IFN-y + IL-4 IFN-y + LPS IL-4 + LPS IL-4 + LPS

0 - 2 0 + 4 0 0 0 + 2 0 0 0 ND

-4 - 8 0 0 0 - 4 0 + 90

- 4 0 0 0 0 ND

3.5 Effect of activating factors on MHC class I Ag-presenting capacity of M a

To determine whether the ability to present exogenous Ag with MHC class I molecules might be a pathway whose activity is regulated in MQ, we examined the effect of the Ma-activating factors on presentation by the cloned BM M a . Representative data from this analysis are shown in Fig. 6 and summarized inTable 1. None of the cytokines significantly enhanced the ability of the BM M a to present exogenous Ag with MHC class I.The concentrations of Ag needed to detect OVA + Kb presentation (minimal stimulatory concentration) for the experimental and control groups varied f two fold on average were unchanged. IFN-y treatment slightly decreased the maximal stimulation of the RF33.70 cells (Fig. 6A). IL-4 did not affect the minimal concentration of Ag (Fig. 6B and Table l ) , however, in some experiments, there was a decrease in the maximal response. The significance and the mechanistic basis for the decrease in the maximal response induced by some cytokines is unclear. Treatment with LPS both increased the minimal concentration of Ag that was needed to detect a response and decreased the slope of the dose-response curve resulting in an inhibition in the maximum response observed (Fig. 6A). The amount of

200 I I A n n . .""

300

200

5 0 100

0 0 u

0.01 0.1 1 10 0.1 1 10

OVA pg/ml

Figure 6. Effect of activating factors on MHC class I Ag presenting capacity of BM MQ. BM A3.1A7 were incubated for 72 h with LFN-y (closed squares) or LPS (closed triangles) (Panel A) or IL-4 (closed squares) (PanelB) or in medium alone (open circles) (A-B) as described in Sect. 2.2. Cells (5 X 104/well) were subsequently cultured with RF33.70 cells (5 x 104/well) and the indicated amount of OVA-beads in 96-well plates (200 pl). Cul- tures were then handled as described in Sect. 2.5.

a) APC were incubated with or without IFN-y, IL-4, LPS, GM-CSF alone or in combination as described in Sect. 2.2. APC (5 X 1O"/well) were then cultured with RF33.70 T-T hybrid (5 X 104/well) with a titration of OVA from 2.5 mg/ml to 60 ng/ml. Microcultures were then prepared and handled as described in Fig. 3.The data are expressed as an average shift in the minimal stimulatory concentration and as the average percentage of the maximal values of T-T hybridoma's IL-2 response induced by the different stimulatory factors. Increase and inhibition by those factors were indicated by the symbols (+) and (-), respectively; e.g. (+4) indicates that four times less OVA was required in the presence of the activating factor to elicit a detectable response relative to untreated APC, whereas (- 8) indicates that eight times more OVA was required. In both analysis, no effect of activating factors was represented by the symbol (0).

peptide required to detect a response was unchanged for MQ incubated with IL-4 or LPS. In some experiments, IFN-y-treated M a needed an increased amount of peptide to detect a response and shown a decrease in the maximum response. Since, all three factors increased MHC class I expression (Fig. 5A, C, E), the decrease in peptide presentation by IFN-y-treated M a could be due to an increase secretion of proteases; however, this hypothesis was not tested. Given these results and since activation of some MQ effector functions can require two signals, we also tested the effects of different combinations of the various activating factors. Again, no increase in presentation was observed and some modest inhibition was observed with several of the combinations of factors. In general, similar effects were observed when OVA was added in soluble or particulate form (Table 1).

A parallel analysis was performed with PEC and overall similar results were obtained (Table 1). One difference that was noted was that presentation by PEC was not inhibited by LPS. Another difference was that IFN-y, and to some extent IL-4, modestly enhanced the presentation of OVA particles. Interestingly, this effect was not observed for the presentation of soluble OVA. The difference between soluble and particulate OVA may indicate that the two forms of Ag are presented by somewhat different mechanisms or that IFN-y might be stimulating increased uptake of particles but not of soluble protein.


The results from the analysis of MHC expression and of class I1 Ag presentation (described below) indicate that the MQ, are responsive to the factors tested in these experiments. Therefore, these results indicate that MQ, activation does not enhance the exogenous Ag-MHC class I-presentation pathway, at least under the conditions of our experiments. Also, the similar results with PEC suggest that these treatments do not induce de n o w this pathway of presentation in significant numbers of previously inactive cells in the heterogeneous cell population.

The lack of an increase in Ag presentation in the above experiments was surprising given the up-regulation of MHC class I expression by the Ma-activating factors, since this result implies some induction of the class I MHC pathway. Also, IFN-y is a potent inducer of the MHC class I pathway of Ag presentation for endogenous Ag. To verify that the “endogenous-pathway’’ of Ag presentation was actually affected in these cells, we tested the effect of IFN-y on the presentation of OVA that was osmotically loaded into cells. As shown in Fig. 7, IFN-y enhanced this form of presentation by both BM M@ and PEC.

1 o o u 50 ;u 0

0

1.5 3 6 12 25 50 1.5 3 6 12 25 50

Cells x 10-

Figure 7. Effect of IFN-y on the endogenous MHC class I presentation pathway. BM A3.1A7 (Panel A) or PEC (Panel B) were treated for 72 h with either IFN-y (circles) or medium alone (triangles). OVA (20 mg/ml, Panel A or 30 mglml, Panel B) was loaded into the cytosol by hypertonic lysis of pinosomes [15]. After 0 min (open symbols) or 60 min (closed symbols) of incubation at 37°C cells were fixed with 1% paraformaldehyde and their APC activity was assayed in culture with RF33.70 T cell hybrid. Cultures were prepared and handled as described in Sect. 2.5.

... m

I 80 ’ 60 X

40 E g- 20

0 loo 1 0 ’ lo2 lo3

300

200

100

0 1 0 4 l o o 10’ l o 2 lo3 l o 4

3.6 Presentation of exogenous Ag with MHC class I1 molecules by BM MQ,

Heterogenous populations of MQ, are capable of processing and presenting exogenous Ag with MHC class I1 molecules. The availability of clones of MQ, that could present exogenous Ag with class I allowed us to determine whether a single cell could simultaneously process and present an exogenous Ag with MHC class I and class 11 molecules. For this purpose we incubated the BM MQ, with OVA added to the extracellular fluids and assayed the presentation of OVA258-276 + IAb to the MF2.2D9 T cell hybridoma. The BM C2.3 MQ, clone and the BM B1.2 and BM B5 MQ, cell lines constitutively processed and presented exogenous OVA with MHC class I1 molecules (Fig. 8A). The activity of these APC in this assay fluctuated from experiment to experiment but presentation was generally detected. In contrast, the BM A3.1A7 and BM B4 failed to present OVA to the MF2.2D9 hybridoma (Fig. 8B). These latter two APC cell lines expressed MHC class I1 molecules (Fig. 5 and data not shown), although surprisingly, they also did not present the processing independent peptide OVA258-276 (data not shown).

To compare the regulation of the MHC class I and class I1 presentation pathway, we studied the effect of MQ,- activating factors on the capacity of BM MQ, to present Ag with MHC 11. Representative data from these experiments are shown in Fig. 8. IFN-y and IL-4 increased the presentation by those BM MQ, that constitutively presented exogenous Ag with class I1 (Fig. 8A). In contrast, IFN-y but not IL-4 induced presentation by BM MQ, that were unable to present Ag under basal conditions. LPS totally inhibited the capacity of BM M@ to present exogenous Ag with MHC class I1 molecules (Fig. 8A). This inhibitory effect of LPS contrasts with its unduction of MHC class I1 expression (Fig. 5F) and its augmentation of Ag presentation in other systems [23]. This inhibition of presentation could be of interest since LPS is a component of many bacterial pathogens but this issue requires further study. GM-CSF had no effect on class I1 presentation. We also tested the effects of combinations of the various activating factors (Data not shown). IFNy with IL-4 or with LPS increased the Ag presenting capacity of both types of BM MQ, cell lines. IL-4 with LPS decreased the ability of the constitutively active MQ, to stimulate the MHC class I1 restricted T-T hybrid and had no effect on the other type of BM MQ, lines. Taken together, these results show a general stimulatory effect of IFN-y either alone or with other factors and an inhibitory effect of LPS. When these results are compared to those in Table 1, it is clear that the presentation of exogenous Ag by the class I and class I1 pathways is independently regulated.

OVA [ ~ g l m l ] 4 Discussion

Figure 8. Effect of activating factors on MHC class I1 Ag presenting capacity of BM M@, BM C2.3 (Panel A) or BM A3.1A7 (Panel B) were treated for 72 h with either IFN-y (closed squares) or IL-4 (closed circles) or LPS (closed triangles) or medium alone (open circles) and then cultures were prepared as described in Fig. 6 with the exception that MF2.2D9 (anti-OVA-IAb) T cells were added instead of RF33.70 T cells. In Panel B, the open circles, the closed circles and the closed triangles are superimposed on the x axis. In Panel A the amount of Ag with the IFN-y-treated cells was not limiting; without Ag these cells gave 852 cpm.

In this report we describe the isolation and characterization of cloned APC which are capable of presenting exogenous Ag with MHC class I molecules. These cells express MQ, differentiation Ag (e.g. MAC-I and F4/80) and lack B and T cell markers. The expression of FcR and the cells’ ability to avidly phagocytosis particles distinguishes them from dendritic cells. These results definitively establish that a M a is a cell which can mediate this unique Ag-presenting

Eur. J. Immunol. 1994. 24: 2421-2428 Presentation of exogenous Ag with MHC I and I1 by MQ, 2427

activity. These results are consistent with and confirm previous analyses with fractionated heterogeneous cell populations and rule out the possibility that some other contaminating cell type was responsible for the activities that were observed in earlier studies [12, 13, 171.

Five of five M@ cell lines that were immortalized from BM with rnyc and ruf were active in our system. Each of these cells arose in independent cultures and represent independent transformation events. These results generalize our findings and indicate that the isolation of such APC is not a very rare event. Previous estimates suggested that there was only a low frequency of APC within primary peritoneal macrophages populations that were capable of presenting exogenous Ag with class I [17]. If these estimates are accurate, then the present results might indicate that APC with this activity are more frequent among BM M@, a possibility which would be consistent with other results [38]. Alternatively, it is possible that rnyc and/or ruf induce this pathway of presentation. However, if this is the case it would be APC-specific since B cells that are transformed with these same oncogenes do not present exogenous Ag with class I (data not shown).

The availability of cloned cells that can process exogenous Ag with MHC class I molecules provides a useful tool for many studies. As in the present report, they allow an analysis of the phenotype and functional properties of a single clone. These homogeneous cells should be very useful for future cell biological and biochemical studies. An important potential caveat is that immortalized clones may differ in some ways from untransformed cells. For this reason, in the present report we have compared the activity of these clones to normal primary M@. In most aspects we obtained similar results with both sources of cells.

Previous studies with primary populations of M@ indicated that the presentation of soluble protein with class I required high concentrations of the Ag. It was difficult to assess whether high concentration of Ag were needed because this pathway of presentation was inefficient or whether this resulted from the low frequency of active APC. While we cannot answer this question for primary APC, the five M@ cell lines we have analyzed present exogenous Ag at significantly lower concentrations when compared to PEC. In fact, these concentrations of Ag are not that much less than is typically required for MHC class 11-restricted presentation in many systems. We also confirm here with multiple cell lines that particulate forms of the Ag are presented with remarkable efficiency.

In this report we also examined the kinetics for the processing and presentation of exogenous Ag when it was added into the extracellular fluids versus when it was loaded into the cytosol. In the past such a comparison was not interpretable because different APC populations would have been compared. Ag added to the extracellular fluids would be presented by the few specialized APC while the cytosolic-loading procedure introduced Ag into all MHC class I-positive cells. With the availability of a homogeneous source of active APC we performed this kinetic analysis. OVA loaded into the cytosol of BM M@ is very rapidly processed ans presented, as we and others have observed for other APC [12,39]. In contrast, the processing and presentation with MHC class I of soluble or particulate

forms of OVA by the BM MQ, required extended periods of incubation. Similar results were obtained with primary PEC. These kinetic differences might be due to differences in the amount or rate of Ag delivered to the class I- processing pathway and further study will be needed to resolve the mechanisms responsible for these differences.

We also report an analysis of the effects of cytokines and LPS on the presentation of exogenous Ag with class I.The rationale for this analysis was several-fold. First, there was considerable precedence that the classical Ag-presenting pathways can be regulated by such factors. Second, it was of interest to determine whether the capability of presenting exogenous Ag with class I was an inducible function and whether this might explain the apparent heterogeneity in the ability of primary MQ to perform this function. Third, understanding the regulation of this pathway could have practical significance since this pathway can be exploited for priming CTL in vivo [ 191 .The picture that emerged from this analysis is that this pathway of presentation is not inducible, at least with the three cytokines and one microbial M@ activator analyzed under the conditions of our experiments. However, as transformed M@ cell lines may not reflect the behavior of fresh M@,we performed the same studies with freshly isolated PEC. The results with PEC were overall similar with one exception. An increased presentation of particular OVA was observed with IFN-y or IFN-y + IL-Ctreated PEC. Perhaps the most interesting aspect of these results is that they point out a clear difference between the pathway of presentation for exogenous Ag and endogenous Ag. The various Mat-activating factors all induced significant increases in MHC class I expression. For IFN-y, we also determined that it aug- mented the presentation of the same Ag (OVA) loaded into the cytosol. IFN-y is known to induce several components of the classical MHC class I pathway, including transcrip- tion of heavy and light chains, TAP 1/TAP 2, and protea- some subunits [40-421. The failure of the exogenous-class I presenting pathway to be affected by the cytokines may indicate that it is not dependent on the same underlying molecular mechanisms.

Our previous studies had indicated that APC were capable presenting exogenous Ag with MHC class I also expressed MHC class I1 molecules. The isolation of clones of such APC allowed us to examine the activity of the MHC class I1 pathway in these cells. The present results indicate that a cloned APC can present exogenous Ag via both the class I and class I1 pathways. This result is of interest because it provides a potential cellular mechanism for allowing peri- pheral Ag to be re-presented in the central lymphoid organs in the context of MHC class1 [6-8, 43, 441 and for promoting interactions between Ag-specific T helper cells and CTL.

Our analysis of BM M@ also indicates that the presentation of exogenous Ag with MHC class I and class I1 are independently regulated. Thus, some clones constitutively present exogenous Ag with MHC class I but not with MHC class 11. In addition, the presentation of exogenous Ag with MHC class I1 is induced by treatment with cytokines, while in the same cells presentation of the Ag with MHC class I is unaffected.TheT-T hybridomas used to assess presentation in this analysis even recognize the same region of OVA (residues 257-264 for class I and residues 258-276 for

2428 M. Kovacsovics-Bankowski and K. L. Rock Eur. J. Irnmunol. 1994. 24: 2421-2428

class 11). These results point to a marked difference between these two pathways of presentation for exogenous Ag.

We would like to thank Dr. C. Cunningham for helping with the photomicrographs and Dr. A . Goldberg for making his spectro- fluororneter available to us. We thank Dr. B. Benacerraf for helpful discussions.

Received April 15,1994; in revised form July5,1994; accepted July 7, 1994.

5 References

1 Allen, P. M., Babbitt, B. l? and Unanue, E. R., Irnrnunol. Rev.

2 Long, E. O., New Biologist 1992. 4: 274. 3 Townsend, A. and Bodmer, H., Ann. Rev. Irnrnunol. 1989. 7:

4 Yewdell, J. W. and Bennink, J. R., Adv. Irnrnunol. 1992.52: 1. 5 Bevan, M. J., Nature 1987. 325: 192. 6 Kaufmann, S. H. E., Irnrnunol. Today 1988. 9: 168. 7 Carbone, F. R. and Bevan, M. J., J. Exp. Med. 1990.171; 377. 8 Bevan, M. J., J. Exp. Med. 1976. 143: 1283. 9 Debrick, J. E., Campbell, P. A. and Staerz, U. D., J. Irnrnunol.

1991. 147: 2846. 10 Collins, D. S., Findlay, K. andHarding, C.V., J. Irnrnunol. 1992.

148: 3336. 11 Zhou, F., Rouse, B.T. and Huang, L., J. Irnrnunol. 1992. 149:

1599. 12 Rock, K. L., Gamble, S. and Rothstein, L., Science 1990.249:

918. 13 Pfeifer, J. D.,Wick, M. J., Roberts, R. L., Findlay, K., Normak,

S. J. and Harding, C.V., Nature 1993. 361: 359. 14 Braciale,T. J., Morrison, L. A., Sweetser, M.T., Sambrook, J.,

Gething, M.-J. and Braciale,V. L., Irnrnunol. Rev. 1987.98: 95. 15 Moore, M. W., Carbone, F. R. and Bevan, M. J., Cell 1988.54:

777. 16 Yewdell, J.W., Bennink, J. R. and Hosaka,Y., Science 1988.239:

637. 17 Grant, E. l? and Rock, K. L., J. Irnmunol. 1992. 148: 13. 18 Rock, K. L., Rothstein, L., Gamble, S. and Fleischacker, C., J.

1987. 98: 171.

601.

Irnrnunol. 1993. 150: 438.

19 Kovacsovics-Bankowski, M., Clark, K., Benacerraf, B. and Rock, K. L., Proc. Natl. Acad. Sci. USA 1993. 90: 4942.

20 Brown, E. J., Curr. Opin. Irnrnunol. 1991. 3: 76. 21 Nacy, C. A. and Meltzer, M. S., Curr. Opin. Irnrnunol. 1991.3:

22 Korber, B., Mermod, N., Hood, L. and Strynowski, I., Science

23 Zlotnik, A., Fischer, M., Roehm, N. and Zipori, D., J.

24 Marshall, N. E. and Ziegler, H. K., J. Irnrnunol. 1991. 147:

25 Rock, K. L., Rothstein, L. and Gamble, S., J. Irnrnunol. 1990.

26 Kappler, J. ,White, J. ,Wegmann, D., Mustain, F. and Marrack,

27 Blasi, E., Mathieson, B. J., Varesio, L., Cleveland, J. L.,

28 Rock, K. L. ,Yeh, E. T. H., Gramm, C. F., Haber, S. I., Reiser,

29 Springer,T., Galfre, G., Secher, D. S. and Milstein, C., Eur. J.

30 Austyn, J. M. and Gordon, S., Eur. J. Irnrnunol. 1981.11: 805. 31 Unkeless, J. C., J. Exp. Med. 1979. 150: 580. 32 Bhattacharya, A,, Dorf, M. E. and Springer,T. A., J. Irnrnunol.

33 Davignon, D., Martz, E., Reynolds, T., Kurzinger, K. and

34 Jones, B. and Janeway, C. A. Jr., Nature 1981. 292: 547. 35 Swanson, J. A., J. Cell Sci. 1989. 94: 135. 36 Watson, J., J. Exp. Med. 1979. 150: 1510. 37 Stuart, M. P., Zlotnik, A. and Woodward, J. G., J. Irnrnunol.

1988. 140: 1542. 38 Denkers, E. Y., Grazzinelli, R. T., Hieny, S., Caspar, l? and

Sher, A., J. Irnrnunol. 1993. 150: 517. 39 Hosken, N. A., Bevan, M. J. and Carbone, F. R., J. Irnrnunol.

1989. 142: 1079. 40 Yang,Y. ,Waters, J. B., Fruh, K. and Peterson, l? A., Proc. Natf.

Acad. Sci. USA 1992. 849: 4928. 41 Wong, G. H. W., Clark-Lewis, I., Mckimm-Breschkin, J. L.,

Harris, A. W. and Schrader, J. W., J. Irnrnunol. 1983.131: 788. 42 Epperson, D. E., Arnold, D., Spies,T., Creswell, P., Pober, J. S.

and Johnson, D. R., J. Irnrnunol. 1992. 149: 3297. 43 Jin,Y. J., Shih,W.-K. andBerkower, I., J. Exp. Med. 1988.168;

293. 44 Staerz, U. D., Karasuyama, H. and Garner, A. M., Nature

1987. 329: 449.

330.

1988. 239: 1302.

Imrnunol. 1987. 138: 4275.

2333.

145: 804.

P., Proc. Natl. Acad. Sci. USA 1982. 79: 3604.

Bochert, l? A. and Rapp U. R., Nature 1985. 318: 667.

H. and Benacerraf, B., J. Exp. Med. 1986. 163: 315.

Irnrnunol. 1979. 9: 301.

1981. 127: 2488.

Springer, T. A., Proc. Natl. Acad. Sci. USA 1981. 78: 4535.

presentation of exogenous antigens by macrophages: analysis of major histocompatibility complex...

Documents