Flow cytometric assessment of the reactivity of a panel of monoclonal antibodies (mAb) against two populations of human dendritic cells (DC)

Background The identification of antigens on human DC has been a very difficult and elusive task because of the lack of appropriate reagents. Therefore, we evaluated by flow cytometry a panel of mAb that recognize antigens on human DC, aiming to determine the kinetics of DC antigen expression at 7, 14, 21 and 28 days in (i) Dermal DC like cells (Mo-DC) and (ii) Langerhans cell like DC (Mo-LC). In addition we aimed to identify markers for DC subpopulations. Results It was found at day 7, that mAb BG6, HP-F1, BU10, RFD-1, CMRF-44 recognized <20% of Mo-DC. In contrast, 7H5, ZM3.8, CDlb/c, 55K-2, MMR1.16, MMR190.BB3 and L25 reacted with >50% of Mo-DC. Moreover, 7H5, ZM3.8, CMRF-56, CDlb/c, 55K-2, MMR1.16, MMR190.BB3 and L25 showed increased MFI reactivity against Mo-DC. mAb BG6, BU10 and CMRF-44 recognized <20% Mo-LC while RFD-1 reacted with 21% of Mo-LC. In contrast, HP-F1 showed 87% of Mo-LC positive. Also, 7H5, ZM3.8, RFD-7, MR15-2, CDlb/c, 55K-2, MMR1.16, MMR190.BB3 and L25 reacted with >50% of Mo-LC. The increase in % of positive cells was paralleled by MFI increases. At day 14, fourteen mAb recognized >50% of the Mo-DC, while five recognized 20-50% of Mo-DC. BG6 reacted with 7% of the Mo-DC. Nineteen mAb recognized >48% of Mo-LC while BG6 had negative reactivity. At day 21 and 28, all mAb reacted with >20% of Mo-DC and yielded a significant MFI with Mo-DC. Also nineteen mAb yielded significant MFI with Mo-LC while RFD-7 did not. Conclusions The immunophenotyping assays demonstrated differences between the two DC populations as well as variations in the reactivity of the mAb at diverse time points, suggesting the existence of subpopulations within the Mo-DC and Mo-LC.


Background
Dendritic cells (DC) are a complex group of mainly bone marrow derived cells that play an important role in the afferent branch of the immune response [1]. However, DC represents only a minute subpopulation of the peripheral blood mononuclear cells (PBMC), as well as of bulk cellular populations of the lung, intestine, genitourinary tissue, and lymphoid tissue. DC also has been found in the epidermis, dermis and mucous membranes and constituting about 2% of the total cellular population of the human epidermis [2,3]. The Langerhans cells (LC) are a skin derived-DC, that have the capability to travel to the regional lymphoid organs after take up of antigen and undergo there an activation/maturation step. Thereafter, LC interacts and activates T cells. Because of such significant capability to take up soluble antigens, process and present them to responder cells in the lymphoid tissues in the context of the restricted MHC pathway, LC have been considered one of the most important elements in the afferent arm of the immune response [1][2][3][4].
Recent, successful efforts to generate DC from PBMC derived monocytes or from CD34 blood precursors by utilizing GM-CSF and IL-4, as well as GM-CSF and/or TNF, has enabled us to obtain PBMC derived DC (Mo-DC)) [5][6][7][8]. In addition, an approach has been developed to generate LC from isolated monocytes (Mo-LC) [9].
Despite the successful efforts in the generation of DC from blood precursors, the characterization of surface markers on human DC has been a very difficult and elusive task because the lack of appropriate reagents with high specificity for DC identification [4]. However, some molecules whose genes recently have been cloned and sequenced (e.g. CD83, DEC-205) have been found strongly associated with DC [5,6]. In addition, a panel of monoclonal antibodies (e.g. CMRF-44) that recognize molecules on DC has been raised [7]. There is a growing need for cluster and establishment of a common and comprehensive nomenclature for such DC associated molecules, as well as to clarify and define the lineage(s) of DC and the existence of DC subsets. These developments have prompted the set up of diverse approaches that evaluated the reactivity of a group of mAb against populations of DC [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29]. Therefore and within the scope of this study, we set up a flow cytometry approach and evaluated a panel of 20 mAb against two populations of DC aiming to determine the kinetics of expression of antigens on DC at diverse intervals of time and the likelihood to identify markers for DC subsets [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29].

Results and Discussion
Flow cytometric assessments of monoclonal antibodies (mAb) reactivity against dendritic cells (DC) was undertaken by measurement of the % of reactivity ( Figure  1 (Table 1). In addition, nineteen mAb yielded significant MFI with Mo-LC whilst only 70512 (RFD-7) did not. (Table 1) At day 21 ( Figure 3), seventeen mAb reacted with more than 50% of the Mo-DC, only 70346 (42D1

Figure 1
Percentage of positive cells detected by mAb against Mo-DC and LC-DC after 7 days of culture (CD85f)), 70511 (RFD-1) and 70512 (RFD-7) were below 50%. In contrast, only seven mAb reacted with more than 50% Mo-LC but only 70512 (RFD-7) reacted with less than 20% Mo-LC. At day 28 ( Figure 4), sixteen mAb reacted with more than 50% Mo-DC. The mAb 70346 (42D1 (CD85f)), 70511 (RFD-1), 70512 (RFD-7) and 70808 (DEC-205 (CD205)) were below 50% while 70375 (CMRF-44) reacted with less than 20% of Mo-LC. Seventeen mAb reacted with more than 50% of Mo-LC. In addition, variations in the reactivity of the mAb at di-verse time points were found. Moreover, at day seven it was found that mAb HP-F1 was negative for Mo-DC and positive for Mo-LC, thus discriminating Mo-DC from Mo-LC. Overall, this set of results demonstrates the existence of antigenic differences between Mo-DC and Mo-LC, even though there was only a single difference in the cytokine mixture utilized. Moreover, within the panel analyzed, it was found that at day 7, there is a mAb that showed reactivity only against Mo-DC with high FSC (mAb-TPD153) or against a fraction of Mo-DC (mAb-DC-LAMP, mAb-55K-2). These results point toward the existence of subsets within the Mo-DC and Mo-LC populations.
Of note, the studies on the reactivity of two set of mAb (mannose receptor and immunoglobulin-like transcript molecules) raised very interesting considerations about the role of these antigens in the functionality of DC.
This study showed that DC expressed significant amounts of two lectin-type receptors: the mannose receptor (MMR) detected by the mAbs MR 15-2 (CD206), MMR1.16 (CD206), and MMR190.BB3 (CD206?), and the antigen DEC-205 detected by the mAbs DEC-205  12], have been implicated in the uptake of carbohydrate-conjugated antigens by DC. MMR has been found on the cell surface of macrophages and its carbohydrate recognition domains mediate the endocytosis of (i) glycoconjugates containing mannose, (ii) fucose, (iii) acetylglucosamine, (iv) glucose residues, (v) microorganisms expressing mannose or acetylglucosamine on the surface. Notably, all these terminal sugars are neither common membrane components of mammalian cells nor common components of serum proteins. Therefore, it is feasible to consider that this two lectin-type receptors: the mannose receptor (MMR) and DEC-205 may be involved in the discrimination between self and non-self antigens as well and could contribute to enhance the capability of dendritic cells to generate primary T cell responses against infectious agents and soluble antigens that carry these types of carbohydrates.
In addition, this study provides the first comparison of the distribution of the MMR and DEC-205 on two human blood derived DC, Mo-LC and Mo-DC.
A panel of mAb that recognize several immunoglobulinlike transcript molecules (ILT) was also evaluated in this study. The mAb HP-F1 (CD85i) identified ILT2 [13], ZM3.8 (CD85j) recognized ILT3 [17], 42D1 (CD85f) identified ILT4 [16] and 7H5 (CD85a) recognized ILT5 [15]. One significant feature of the ILTs molecules is their capability to bind MHC class I molecules [13]. Furthermore, the molecules ILT2, ILT3, ILT4, and ILT5 have been considered to play a role as inhibitory receptors because they carry the immunoreceptor tyrosinebased inhibitory motif (ITIM) in the intracytoplasmic tail. Because of (i) the binding capability of ILT2-5 to MHC class I molecules [13,17], (ii) the presence of inhibitory motifs within the cytoplasmic tail of ILT2-5 [13,[15][16][17] and (iii) the identification of ILT2-5, within diverse populations of dendritic cells [10,17], is likely that the ILT molecules identified on DC may not only perform a key role as the receptor for the "missing self [30] but also  Percentage of positive cells detected by mAb against Mo-DC and LC-DC after 28 days of culture could contribute to control the initial steps of activation of the immune response; Thus, this would tune up the fine balance between the activation signals provided by the simultaneous cross-linking of MHC class I or II and co-stimulatory molecules during the interaction between DC with T cells and the inhibitory signals mediated through the whole variety of ILTs molecules carrying ITIM motifs. Further functional studies of the role of ILTs on DC will provide insight toward the understanding of these mechanisms and its role in the regulation of the DC function.

Conclusions
The immunophenotyping assays described in this report enabled us to determine in human DC: (i) the existence of differences between Mo-DC and Mo-LC populations; (ii) the existence of subsets within the Mo-DC and Mo-LC populations; (iii) the kinetics of antigens expression at diverse intervals of time on DC; and (iv) specific markers for subpopulations of DC.
The Mo-DC and Mo-LC populations were harvested and evaluated at 7, 14, 21 and 28 days   [12] cytometry measurement on non-stained or stained only with the second antibody (secondary antibody) were performed and used as control populations. Secondary antibody only stained cells at day 7 served to determine the markers and the quadrant borders. At least 99% of these cells were located in the lower left quadrant (negative). Three independent experiments were performed. Data were analyzed with CellQuest (Becton Dickinson, San Jose Ca). Statistical analysis (mean, standard deviation and graphics) was performed with Microsoft Excel.