Isolation of functionally active murine follicular dendritic cells

Journal of Immunological Methods 313 (2006) 81 – 95 www.elsevier.com/locate/jim

Research paper

Isolation of functionally active murine follicular dendritic cells☆ Selvakumar Sukumar a , Andras K. Szakal b , John G. Tew a,⁎ b

a Department of Microbiology and Immunology, Virginia Commonwealth University, Richmond, VA, 23298-0678, United States Department of Anatomy and Neurobiology, The Immunology Group, Virginia Commonwealth University, Richmond, VA, 23298, United States

Received 20 October 2005; received in revised form 31 January 2006; accepted 29 March 2006 Available online 16 June 2006

Abstract Biochemical, genetic, and immunological studies of follicular dendritic cells (FDCs) have been hampered by difficulty in obtaining adequate numbers of purified cells in a functional state. To address this obstacle, we enriched FDCs by irradiating mice to destroy most lymphocytes, excised the lymph nodes, and gently digested the nodes with an enzyme cocktail to form single cell suspensions. The FDCs in suspension were selected using the specific mAb FDC-M1 with magnetic cell separation technology. We were able to get nearly a million viable lymph node FDCs per mouse at about 90% purity. When examined under light and transmission electron microscopy, the cytological features were characteristic of FDCs. Furthermore, the cells were able to trap and retain immune complexes and were positive for important phenotypic markers including FDC-M1, CD21/35, CD32, CD40, and CD54. Moreover, the purified FDCs exhibited classical FDC accessory activities including: the ability to co-stimulate B cell proliferation, augment antibody responses induced by mitogens or antigens, maintain B cell viability for weeks, and protect B lymphocytes from anti-FAS induced apoptosis. In short, this combination of methods made it possible to obtain a substantial number of highly enriched functional murine FDCs. © 2006 Elsevier B.V. All rights reserved. Keywords: Follicular dendritic cells; Magnetic bead purification; FDC-M1; FDC phenotype; Isolation procedure; Electron microscopy

1. Introduction Follicular dendritic cells (FDCs) are important immune accessory cells that reside in the follicles of Abbreviations: FDC, follicular dendritic cells; mAb, monoclonal antibody; ICs, Immune complexes; PO, peroxidase; OVA, ovalbumin; HRP, horseradish peroxidase. ☆ This work was supported by the National Institutes of Health Grants AI-17142. ⁎ Corresponding author. Virginia Commonwealth University, Department of Microbiology and Immunology, P.0. Box 980678, Richmond, VA, 23298-0678, United States. Tel.: +1 804 828 9715; fax: +1 804 828 9946. E-mail address: [email protected] (J.G. Tew). 0022-1759/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2006.03.018

secondary lymphoid organs and are functionally active in the light zones of germinal centers (GCs) where they are associated with proliferating B cells. Their unique ability to trap and retain surface bound immune complexes (ICs) together with their restricted follicular location, distinguishes FDCs from all other accessory cells including T cell associated dendritic cells (DCs) (Tew et al., 1982). FDCs bearing specific Ag in ICs are requisite for full development of GCs and are believed to be involved in T-dependent B cell responses: Ig class switching, production of B memory cells, selection of somatically mutated B cells with high affinity antigen receptors (BCR), affinity maturation and augmentation of secondary Ab responses (Tew et al., 2001). In

82

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

addition, ICs trapped on FDCs include large amounts of infectious HIV particles which can persist for prolonged periods of time (Racz et al., 1989; Smith et al., 2001). Retention of HIV particles on FDCs plays an important role in HIV pathogenesis and constitutes a potent reservoir present adjacent to CD4+ve T cells (Burton et al., 1997, 2002; Piris et al., 1987; Fox et al., 1991; Racz et al., 1989; Heath et al., 1995). FDCs have also been implicated in the pathogenesis of transmissible spongiform encephalopathies (TSE) caused by prion proteins. Strong evidence for the accumulation of pathological prion protein on FDCs exists (Fraser and Farquhar, 1987; Kitamoto et al., 1991; McBride et al., 1992). Furthermore, the presence of functional FDCs appears to be important for lodgment, replication and further spread of the infective prion protein (Brown et al., 1999; Mabbott et al., 2000; Manuelidis et al., 2000). Thus, we reason that an appreciation of FDC accessory activities and the regulation of these activities are critical to an understanding of fully functional and mature Ab responses as well as diseases associated with FDCs. Compared with the other immune cells, little genetic and biochemical information is available on FDCs. This lack of information is attributable in large measure to FDCs being: 1 — rare (∼ 1 in 10,000 cells in secondary lymphoid tissues), 2 — very fragile, and 3 — challenging to obtain in a functional state in adequate number. Consequently, very few laboratories work with FDCs. Our knowledge of more typical leukocytes has been largely derived from in vitro studies using purified populations. We reason that efforts to characterize biochemical and genetic properties of FDCs would expand if reasonably simple methods were developed to isolate adequate numbers of functional FDCs in purity. FDCs from humans and mice have been enriched in various states of purity and studied (Lilet-Leclercq et al., 1984; Tsunoda et al., 1990; Wu et al., 1996; Marcoty et al., 1993; Heinen et al., 1993; Stahmer et al., 1991; Parmentier et al., 1991; Sellheyer et al., 1989; Ennas et al., 1989; Cormann et al., 1988; Heinen et al., 1985; Schmitz et al., 1993; Humphrey and Grennan, 1982; Clark et al., 1992). Furthermore, FDC-M1 has been used to positively select for murine FDCs using two cycles of cell sorting (Burton et al., 1993) as well as to deplete murine FDCs from enriched preparations (Burton et al., 1993; Wu et al., 1996; Huber et al., 2005). However, methods routinely used to obtain functional murine FDCs are based on density gradients and yield preparations in the range of 25% to 50% purity (Wu et al., 1996; Burton et al., 1993). Nevertheless, these enriched preparations have been useful in studies which showed that FDCs: 1) cluster with B cells and promote

proliferation of adjacent B cells (Kosco et al., 1992); 2) block apoptosis in B cells (Schwarz et al., 1999) and maintain viable B cells and a functioning immune system in vitro for weeks (Qin et al., 1999); 3) block the ITIM signaling in B cells stimulated by ICs (Aydar et al., 2004, 2003); 4) provide co-stimulatory signals for B cell proliferation stimulated by antigen or mitogen (Burton et al., 1993; Tew et al., 2001; Qin et al., 1998), and 5) promote recall IgG responses (Wu et al., 1996; Fakher et al., 2001). However, 25 to 50% enrichment is not adequate for biochemical, genetic, or immunological studies aimed at determining the molecules critical for accessory functions. Accordingly, we sought to improve purity and yield while maintaining immunological function. Magnetic cell separation, a relatively simple and inexpensive method, has been successfully used to isolate a variety of rare cells including human FDCs (Schmitz et al., 1993). We first used established methods to enrich murine FDCs, which included killing radiosensitive lymph node cells by irradiation (Phipps et al., 1981; Kosco-Vilbois et al., 1993) and an enzyme cocktail to obtain single cell suspension (Schnizlein et al., 1985). Finally, we used FDC-M1, a monoclonal antibody specific for murine FDCs (Kosco et al., 1992; Gray et al., 1991), to label FDCs, which were then positively selected in a magnetic column. Using this combination of techniques with the harvested lymph nodes, we obtained about a million viable lymph node FDCs per mouse. Flow cytometric analysis indicated that about 90% of these cells exhibited the FDC phenotype and that very few contaminating macrophages and lymphocytes were present. We also found that positively selected FDCs maintained their ability to bind and retain surface ICs, protect B cells from Fas/CD95 mediated apoptosis, enhance proliferation of mitogen stimulated B cells, and augment antibody production in mitogen or antigen stimulated B cells. In addition, this methodology was used to positively select FDCs in a published study to determine the influence of IC bearing FDCs on the IgM response, Ig class switching, and production of high affinity IgG (Aydar et al., 2005). In conclusion, the protocol described here enables the isolation of functional FDCs that are suitable for a variety of studies including biochemical and genetic analysis (Sukumar et al., 2006). 2. Materials and methods 2.1. Mice, immunization, and irradiation BALB/c mice, 6 to 8 weeks old, were purchased from the National Cancer Institute. They were housed

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

in standard shoebox cages and food and water were given ad libitum. Mice were immunized with OVA as described previously (Kosco et al., 1992; Wu et al., 1996). In short, mice were primed by injecting 0.1 ml of 200 μg/ml of alum precipitated OVA with heat killed Bordetella pertussis. Two weeks later mice were boosted by injecting 50 μl of the same immunogen sc in each front leg and hind footpad and memory lymphocytes were isolated two to six weeks later. To isolate FDCs, naïve mice were irradiated with 1000 Rads in a 137Cesium irradiator 24 h prior to FDC isolation to kill most lymphocytes and enrich for FDCs (Kosco-Vilbois et al., 1993). All animals were handled humanely in compliance with Virginia Commonwealth University Institutional Animal Care and Use Committee (IACUC) guidelines. 2.2. Antibodies and reagents The FDC specific monoclonal antibody FDC-M1 was a kind gift from Dr. Marie Kosco-Vilbois, (NovImmune, Switzerland). FDC-M1 was also bought from BD Pharmingen (San Diego, CA). Biotin conjugated anti-rat kappa (clone MRK-1), PE-antiCD45R/B220 (clone RA3-6B2), FITC-anti-CD90.2 (clone 30-H12), PE-anti-CD16/CD32 (clone 2.4G2), FITC-anti-CD21/CD35 (clone 7G6), FITC-anti-mouse ICAM (clone 3E2), hamster anti-mouse CD40 (clone HM40-3), FITC-anti-hamster IgM (clone G188-9), PEstreptavidin, isotype controls FITC-rat IgG2a and PErat IgG2b were purchased from BD Pharmingen (San Diego, CA). PE-anti-mouse F4/80 antigen was purchased from Serotec (Raleigh, NC). PE-goat antimouse IgG and isotype controls, PE-goat IgG, PE-rat IgG2a and FITC-rat IgG2b were purchased from Southern Biotechnology Associates Inc (Birmingham, Alabama). 2.3. Isolation of FDCs Lymph nodes (submental, mandibular, cervical, axillary, brachial, mesenteric, inguinal, and popliteal) from four irradiated naïve mice were dissected out and placed in 35 × 10 mm Petri dishes containing complete Dulbecco's Modified Eagles Medium (DMEM) on ice. The lymph node capsules were gently teased and opened enough for enzymes to enter using sharp sterile disposable 26 G needles mounted on plastic syringes. The opened lymph nodes were then digested using a cocktail consisting of 1 ml of 8 mg/ml Collagenase D (Roche, Indianapolis. IN) and 1 ml of 10 mg/ml DNase (Sigma, St. Louis, MO) plus 1 ml complete DMEM

83

with 10% fetal bovine serum (FBS). Not all lots of collagenase D worked well and an alternative that we used in some experiments consisted of 100 μl Liberase Blendzyme 2 (3 mg/ml, 14 Wünsch units/ml) (Roche Applied Science, Indianapolis, IN), 1 ml of 10 mg/ml DNase and 2 ml of complete DMEM. The lot-to-lot variation of proteases in the crude Collagenase D is high and the use of pure collagenase in Liberase Blendzyme helped to keep this variation to a minimum but in our experience a good lot of collagenase D gives FDCs with the best biological activity. The nodes were digested for 1 h at 37 °C in a humidified incubator and the cells were separated gently with a Pasteur pipette and collected in a 50 ml tube containing ice-cold DMEM enriched with 10% FBS. At this point, most of the lymph node collagen were digested and the tissue disintegrated into fragments with gentle pipetting. If this did not occur, the enzymes had not worked properly and the preparation was discarded. The tissue fragments remaining after pipetting were further digested for 30 min at 37 °C using fresh enzyme cocktail and the cells were again collected by gentle pipetting and added to the cells in the 50 ml tube on ice. At this point only tiny fragments of tissue persisted. 2.4. Labeling and enrichment by MACS The suspended cells from the digested lymph nodes were collected by centrifugation at 300 g for 10 min and incubated with 0.64 μg/ml of purified FDC-M1 (BD Pharmingen, San Diego, CA) per 20 × 106 total viable cells in complete DMEM with 10% FBS for 1 h on ice. In some experiments, 1 ml of FDC-M1 hybridoma supernatant fluid was used and we found that centrifuging the FDC-M1 supernatant fluid at 12,000 g for 10 min reduced debris that could be isolated with FDCs or nonspecifically stuck to the FDCs. The cells were washed once with complete DMEM and incubated for 30 min with 5 μg/ml biotinylated anti-rat kappa antibody (clone MRK-1) in complete DMEM with 10% FBS on ice. Subsequently, cell clumps were removed by straining through a 70 μm strainer and washed with MACS buffer (0.5% bovine serum albumin, 2 mM EDTA in phosphate buffered saline, pH 7.2, sterile filtered and degassed). The cells were then incubated with anti-biotin microbeads (20 μl/400 μl MACS buffer; Miltenyl Biotec, Auburn, CA), per standard isolation (cells from 4 nonimmunized mice) in the dark and on ice for 20 min. Finally, the cell suspension was washed once in MACS buffer and gently placed on a LS column assembled on a VarioMACS separator (Miltenyl Biotec, Auburn, CA).

84

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

The bound cells were washed thoroughly with 15 ml of MACS buffer and gently eluted after removing the column from the magnet. The eluted FDCs were immediately washed and resuspended in complete DMEM with 10% FBS.

tions or left unstained for the better detection of labeling of the FDCs with the monoclonal Ab, FDCM1. Thin sections were studied with a JOEL electron microscope at 50 kV for cytochemical labeling to increase contrast on unstained sections and at 100 kV for detailed morphology.

2.5. Isolation of lymphocytes 2.7. Flow cytometry To isolate lymphocytes, lymph nodes dissected from naïve or OVA immunized mice were squashed between the frosted ends of two sterile glass slides and the cell suspension strained through a 40 μm strainer. These preparations consisted predominantly of T and B cells with few lymph node accessory cells. 2.6. Light and electron microscopy Light microscopy was done after placing 1 × 106 FDCs into a cytobucket, centrifuging the cells onto poly-L-lysine coated glass slides (Cytobuckets and IEC PR-J centrifuge, Damon/IEC Division, Needham Heights, MA), and then visualizing the cells using a BH2 Olympus light microscope equipped with Nomarski optics as described previously (Szakal et al., 1985; Schnizlein et al., 1985). For sections, FDCs were fixed with 1% paraformaldehyde + 0.87–0.9% glutaraldehyde (Electron microscopy grade, Electron Microscopy Sciences, Washington, PA) in 0.1 M cacodylate buffer, postfixed in 1% OsO4 in 0.1 M cacodylate buffer for 90 min, dehydrated in a graded series of ethanols, infiltrated, and embedded in PolyBed 812 epoxy resin. For light microscopy 1 μm-thin sections were cut using a LKB 2128 ultramicrotome and were stained with 1% Toluidine blue solution in 1% sodium borate. Thick sections were visualized using a BH2 Olympus light microscope and the images were captured using an optronics digital camera. To prepare cells for electron microscopy, FDCs were incubated with FDC-M1 ab (0.5 μg/ml) overnight at 4 °C, followed by 5 μg/ml of Horseradish peroxidase (HRP) labeled mouse anti-rat IgG (Zymed, San Francisco, CA) for 1 h on ice. The HRP labeling was further amplified using a tyramide amplification kit (biotin-XX tyramide, Molecular Probes, Carlsbad, CA) following the manufacturer's protocol. The cells were finally incubated with HRP labeled streptavidin for 1 h on ice and peroxidase was developed using diamino benzidine (DAB) (catalog number S15, Biomeda, Foster city, CA). For transmission electron microscopy ∼800 Å thick sections were mounted on slotted grids and stained with lead citrate and uranyl acetate for morphological evalua-

Cells were first incubated with purified mouse IgG or mouse Fc-Block (Pharmingen, CA) for 15 min on ice followed by the appropriate fluorochrome conjugated primary antibody or isotype control for 60 min in dark at 4 °C. Blocking Fc receptors is important because activated FDCs have high levels of Fc receptor that can lead to false positive results. After washing, cells were sequentially incubated with secondary and tertiary reagents for 30 min on ice where necessary. Cells were then washed thoroughly and analyzed by FACScan (Becton-Dickinson). All acquisitions were made using the CellQuest software and analyzed in WinMDI 2.8 (Scripps Institute, CA). For FDC-M1 labeling, FDCs eluted from the MACS column were incubated with avidin followed by biotin to block any biotin that may persist from the isolation procedure. The cells were then labeled with biotinylated FDC-M1 or biotinylated isotype control followed by streptavidin PE and analyzed by flow cytometry. 2.8. Cell cultures To test immune complex binding, 5 × 105 FDCs, peritoneal macrophages or total lymphocytes were incubated with preformed immune complexes (50 ng OVA + 300 ng anti-OVA or 100 ng OVA + 600 ng antiOVA) for 2 h, washed thoroughly and further incubated for 18 h in complete DMEM (DMEM supplemented with 0.02 M HEPES buffer, 0.2 mM MEM non-essential amino acids, 2 mM L-Glutamine and 1 μg/ml gentamicin) + 10% FBS at 37 °C in a humidified incubator with 5% CO2. The cell bound immune complexes were analyzed by flow cytometry using PE-goat anti-mouse IgG (Southern Biotechnology Associates, Birmingham Alabama). In proliferation assays, lymphocytes were cultured with 10 ng/ml LPS in the presence or absence of increasing numbers of FDCs for 3 days and 3Hthymidine (1 μCi/well) was added to the cultures during the last 18 h. The cells were harvested onto a 96 well filter (Unifilter-96 GF/B, PerkinElmer, Boston, MA). Topcount scintillation and luminescence counter

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

85

Table 1 Counts and viability of FDCs isolated by magnetic separation

3. Results

Experiment

Number of viable FDCs per 4 mice

Percentage of viable cells

3.1. Morphology and phenotype of isolated FDC-M1 positive cells

1 2 3 4 5 Average Average number of FDCs per mouse

3.7 × 106 5.7 × 106 4.7 × 106 5.2 × 106 4.5 × 106 4.76 × 106 1.19 × 106

76 90 86 87 78 83.4

(Packard/PerkinElmer, Boston, MA) was used to quantify the 3H-thymidine uptake. 2.9. Antibody assays To determine total antibody production in vitro, 1 × 106 naïve lymphocytes were cultured in 1 ml of complete DMEM + 10% FBS and 10 μg/ml LPS in the presence or absence of 2 × 105 FDCs in 5 ml polypropylene tubes at 37 °C in a humidified incubator with 5% CO2. The media was replaced on day 7 and antibody production was measured on day 14. For analysis of OVA specific IgG production 1 × 106 lymphocytes from OVA immunized mice were stimulated with immune complexes (ICs) consisting of 100 ng OVA and 600 ng anti-OVA in the presence or absence of 2 × 105 FDCs and cultured similarly. Total or anti-OVA specific IgG in the supernatant fluids on day 7 and on day 14 was measured by solid phase ELISA as described previously (Helm et al., 1995; Wu et al., 1996).

We were able to obtain an average 1.19 × 106 FDCs per mouse by positive selection using FDC-M1 with an average viability of 83.4% based on the ability of viable cells to exclude trypan blue (Table 1). We sought to determine whether cells positively selected using FDCM1 Ab have the morphological, phenotypical, and functional characteristics of murine FDCs (Szakal et al., 1985; Schnizlein et al., 1985; Haley et al., 1995; Liu et al., 1997; Tew et al., 1997, 2001). FDCs in reticula/ networks in germinal center light zones trapped and retained large amount of ICs on highly convoluted dendritic processes (Szakal and Hanna, 1968; Hanna and Szakal, 1968). Using surface bound ICs as a means of identifying FDCs, it is possible to identify such cells in vitro (Szakal et al., 1985). However, some FDC-M1 positive cells such as Ag transporting cells lack extensive processes (Haley et al., 1995). When observed through Nomarski optics the positively selected cells appeared rounded and dendritic processes were rarely apparent at the light microscopic level (Fig. 1). Prior to placing the cells on the LS column we could find cells with extensive processes under Nomarski optics but such cells were difficult to find after column passage. A large amount of FDC-M1 positive material was trapped on the LS columns that generally exhibited less forward and side scatter than lymphocytes. This FDC-M1 positive material did not contain DNA and it appears that the extended FDC processes were sheared from the cells at

2.10. Apoptosis assays Fas/CD95 mediated apoptosis in A20 cells (B lymphoma line from the American Type Culture Collection, Manassas, VA) was assessed as described previously (Schwarz et al., 1999). Briefly, 10 ng/ml of anti-Fas (BD Biosciences, Pharmingen, CA) was first added to 5 × 105 A20 cells for 10 min and then incubated with or without FDCs at different ratios at 37 °C. After 6 h, cells were harvested and labeled with FITC-dUTP using TUNEL assay kit (Roche Diagnostics) according to the manufacturer's protocol. A20 cells were also labeled with PE-anti-CD45R/B220 and analyzed by flow cytometry. In wells containing FDCs, the total cell numbers were adjusted to keep the cell to antibody concentration constant. The CD45R/B220 positive cells were gated to calculate the percentage of apoptotic A20 cells and keep FDCs from influencing the analysis.

Fig. 1. Morphological features of FDCs isolated by positive selection using magnetic microbeads. FDCs eluted from the MACS column were examined using a microscope with Nomarski optics (100× magnification).

86

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fig. 2. Nuclear features of isolated FDCs. One-micron thick sections of FDCs were stained with methylene blue and visualized using a light microscope (100× magnification).

this step in the isolation procedure. Microscopy of 1 μmthin sections of these FDC-M1 positive cells revealed the characteristic nuclear features of FDCs in vivo (Szakal and Hanna, 1968; Hanna and Szakal, 1968) (Fig. 2). These features included relatively large euchromatic nuclei with marginated chromatin and a well-developed nucleolus. Occasionally the nucleus looked cleaved. The

Fig. 3. A transmission electron micrograph illustrating two isolated cells, a small lymphocyte (L) and the cell body of a FDC identified by the electron dense (black) label on its irregular plasma membrane. The labeling represents the HRP labeling using the monoclonal Ab, FDCM1 bound to epitopes on the plasma membrane of the FDC. Note that the labeling is specific for the FDC and it does not bind to the attached lymphocyte. Arrows represent areas heavily labeled with FDC-Ml. The nucleus appears more or less round and relatively euchromatic. The cytoplasm contains numerous small vesicles reflecting endocytic activity in solution. The section is unstained reflecting contrast due only to osmication. The bar represents l μm.

cytoplasmic outline of these cells varied from oval to irregular with the cytoplasmic edges being ruffled and mostly free of obvious dendritic processes. The cytoplasm of these cells contained few dark granules when stained with methylene blue (Fig. 2). At the ultrastructural level, FDCs can be identified using FDC-M1 and the plasma membrane of these isolated cells exhibited electron dense FDC-M1peroxidase labeling (Fig. 3). The cells had the typical oval or irregular euchromatic nuclei (Figs. 3 and 4), but also numerous endocytic vesicles in their cytoplasm apparently as a consequence of taking up media (Fig. 3). On some sections of the cell-pellet, FDCs were grouped (Fig. 4) and showed bipolarity in that there were numerous small dendritic processes extending from one side of the cytoplasm while none from the opposite side (Fig. 4). Also, larger dendritic processes labeled with FDC-M1-peroxidase could be found near cell bodies of FDCs (Fig. 5). Other cell types seen, such as an occasional small lymphocyte, were not labeled with FDC-M1 and thus serve as specificity

Fig. 4. Transmission electron micrograph illustrating a group of isolated FDCs stained with lead citrate and uranyl acetate to show various morphological features i.e., irregular euchromatic nuclei and dendritic cell processes typical of FDCs. Uranyl acetate diminishes the peroxidase (PO) labeling with FDC-M1-PO and it is not discernable at this magnification. The FDCs shown represent the great majority of the cells in the preparation. The bar represents 2 μm.

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

87

Very few contaminating macrophages (0.04%) (Fig. 8B), T cells (3.3%) or B cells (2.1%) (Fig. 8E) were detectible. Freshly isolated peritoneal macrophages (Fig. 8C), lymph node B and T cells (Fig. 8F) showed normal labeling with these markers under the same conditions. Thus, by flow cytometric analysis about 90% of the cells isolated by this technique exhibited the FDC phenotype. 3.2. Ability to bind immune complexes The ability to trap and retain ICs on the cell surface for long periods of time via CD32 (FcγRII) and CD21/ CD35 (CR1/CR2) is a “cardinal feature” that distinguishes FDCs from other accessory cells (Tew et al., 1982; Tsiagbe et al., 1992; Qin et al., 2000). This prompted questions about how well the isolated FDCs Fig. 5. A transmission electron micrograph showing a convoluted process of a FDC and a FDC cell body (upper left) both identified by the FDC-M1 label on their plasma membranes as belonging to FDCs. Note the arrows pointing to heavier labelings on the plasma membrane. (P) FDC process; (C) transverse section through collagen fibres, which appear to be labeled. The section is unstained reflecting contrast due only to osmication. The bar represents 2 μm.

control for FDC-M1 labeling in the cell isolate (see Fig. 3). FDC-M1 may also label tingible body macrophages (TBM) (Dr. Kosco-Vilbois personal communication), but macrophages were found infrequently in the FDC isolates. FDC-M1 positive macrophages did not contain typical tingible bodies, as apparently during isolation these bodies were degraded (Fig. 6A and B). However, no other type of macrophage is known to react with the monoclonal Ab, FDC-M1. Particles compatible in size with the previously described immune complex coated bodies (iccosomes) (Szakal et al., 1988; Wu et al., 1996) were apparent at the ultrastructural level but we could not be certain that they were iccosomes as ICs were not detectable with the methods used. Phenotypically, as assessed by flow cytometry, about 90% of the cells that emerged from the LS column were positive for FDC-M1 (Kosco et al., 1992) (Fig. 7A). In addition, a similar percentage of cells was positive for other FDC associated molecules including: CD21/CD35 — 83%, CD16/CD32 — 92%, and CD40 — 90% (Fig. 7B). Most cells also expressed ICAM-1 (72%) but a subset of FDCs appeared to be ICAM-1-low to negative (Fig. 7B). The expected contaminants were B and T cells which can adhere to FDCs. To determine the frequency of these cells, freshly isolated FDCs were labeled with anti-CD45R/B220, anti-CD90.2/thy-1.2, or anti-F4/80.

Fig. 6. Transmission electron micrograph showing a macrophage found in the population of isolated cells. (A). While peritoneal macrophages did not label with FDC-Ml this macrophage did. See arrows pointing at areas of heavier labeling. (B). Further magnification of the boxed area illustrating plasma membrane labeling, with FDCM1-PO (arrow) and the electron dense group of microbeads (with iron) (arrowhead indicates an individual microbead) used during the separation of FDC-Ml-microbead isolation on the magnetic column. This cell was identified on the account of its labeling with FDC-M1 as a tingible body macrophage (TBM). The preparation contained very few of these TBMs. The section is unstained reflecting contrast due only to osmication. The bar represents 1 μm.

88

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fig. 7. Phenotypic features of FDCs isolated by magnetic separation. (A). Positively selected FDCs were labeled with biotinylated FDC-M1 or biotinylated isotype control (dotted line) followed by streptavidin PE. (B). Thick lines represent FDCs labeled with anti-CD21/CD35, anti-CD32/ CD16, anti-CD40 or anti-CD54 (ICAM-1) and thin lines represent corresponding isotype matched control antibodies. The percentage of cells positive is indicated above the marker line. The results are representative of at least three experiments.

would bind and retain ICs. The ability to trap and retain ICs was tested by incubating FDCs with OVA and murine anti-OVA ICs for 2 h at 4 °C, washing thoroughly, and then further culturing for additional 18 h at 37 °C. For comparative purposes, peritoneal macrophages and lymphocytes were also incubated with the same IC preparation and culture conditions. The majority of the isolated FDCs bore sufficient IC to be labeled with anti-mouse IgG without the addition of fresh ICs (Fig. 9A). As expected, incubating with increasing amounts of ICs resulted in increased labeling with over 90% of the isolated cells exhibiting a marked increase in surface IgG labeling. The data suggest that isolated FDCs retained some ICs trapped in vivo and that they were able to trap and retain additional ICs. In marked contrast, no significant increase in IgG labeling was observed when macrophages or lymphocytes were incubated over the same period of time with the same ICs at the highest concentration (Fig. 9B and C). Macrophages and lymphocytes express Fc receptors and can bind ICs, but we reasoned that the 18 h incubation

period allowed sufficient time for ICs trapped by macrophages and lymphocytes to be endocytosed and cleared from the cell surface while FDCs retained ICs on their surfaces. 3.3. Anti-apoptotic activity FDCs are remarkable in their ability to block apoptotic signals delivered to normal or malignant B cells by agonistic anti-FAS/CD95 or chemotherapeutic agents (Schwarz et al., 1999). To determine if these FDCs retain anti-apoptotic activity, A20 cells were pretreated with anti-FAS and incubated for 6 h in the absence or presence of varying numbers of FDCs. The positively selected FDCs were able to dramatically reduce the percentage of apoptotic A20 cells in ratios as low as 1 FDC to 16 A20 cells (Fig. 10) and this compares favorably with the less pure FDCs obtained from Percoll gradients where ratios of 1/8 worked well but 1/16 was near the limit of FDC activity (Schwarz et al., 1999).

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

89

Fig. 8. FDCs isolated by the magnetic cell separation technique contain few contaminating T and B cells or macrophages. FDCs isolated by positive selection using FDC-M1 (A, B, D, E) or peritoneal macrophages (C) or lymphocytes (F) were incubated with PE labeled anti-F4/80 antibody (B, C), PE labeled anti-CD45R/B220 and FITC labeled anti-CD90.2 (E, F) or with appropriate isotype matched control antibodies (A, D) and were analyzed by flow cytometry. The percentages at the corners of the density plots correspond to each quadrant. Results are representative of at least three separate experiments.

3.4. Augmentation of B cell proliferation FDCs deliver a co-stimulatory signal(s) that dramatically enhances proliferative responses of mitogen stimulated B cells (Kosco et al., 1992; Burton et al., 1993; Tew et al., 2001). To determine if positively selected FDCs retain this accessory activity, we cultured LPS stimulated lymphocytes in the presence of increasing numbers of FDCs. In the absence of FDCs, B cell proliferation was barely detectable at the suboptimal dose of LPS used (175 CPM). As shown in Fig. 11, an FDC dose-dependent increase in B cell proliferation was apparent as evidenced by a 120-fold increase in 3H-thymidine uptake with one FDC per 20 B cells. FDCs do not proliferate in response to mitogens including LPS and did not contribute to the 3H-thymidine uptake (data not shown). 3.5. Augmentation of IgG responses in vitro FDCs bearing CD21L can engage B cells via the co-receptor complex (CD21/CD19/CD81) and mark-

edly enhance IgG responses to mitogens and the combination of specific Ag and CD21L dramatically enhances recall responses to specific Ags in vitro (Qin et al., 1998). To determine if positively selected murine FDCs retain this accessory activity, we cultured LPS stimulated B cells with or without FDCs and total IgG production was measured by ELISA after 14 days in culture. FDCs markedly enhanced the IgG response of LPS stimulated B cells (Fig. 12A) as well as the OVA specific IgG response when B cells from OVA immune mice were stimulated with OVA-anti-OVA ICs in the presence of FDCs (Fig. 7B). The cells persisting in these cultures were examined under a microscope and it was apparent that very few if any lymphocytes persisted after two weeks in cultures lacking FDCs while lymphocytes were abundant in cultures with FDCs (data not shown). This ability to maintain lymphocyte viability is consistent with the previous data with Percoll gradient based isolation of FDCs supporting the concept that FDCs maintain viable lymphocytes and a functional immune system for weeks (Qin et al., 1999; Tew et al., 1997).

90

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fig. 9. Positively selected FDCs bound and retained immune complexes on their surfaces. (A) FDCs incubated under conditions indicated in the legend were labeled for surface IgG. Peritoneal macrophages (B) or lymphocytes (C) were incubated with media (dotted line) or 100 ng OVA + 600 ng anti-OVA (thick line) and similarly labeled for mouse IgG. Results are representative of at least three separate experiments.

3.6. Maintenance of viable FDCs in long-term cultures

Fig. 10. Positively selected FDCs decrease anti-FAS/CD95 induced apoptosis. A20 cells treated with anti-FAS/CD95 were cultured for 6 h to induce apoptosis in the presence or absence of FDCs at a ratio of 1 FDC to either 8 or 16 A20 cells. The cells were then fixed and labeled for DNA strand breaks using a TUNEL kit analyzing cells positive for CD45R/B220 with anti-B220 antibody. The percentage of apoptotic A20 cells is plotted. ⁎ denotes a p value <0.01. Results are representative of at least three independent experiments.

FDCs are long-lived and enriched human FDCs can be maintained for weeks to months in cultures (Tsunoda et al., 1990). Murine FDCs appear to be “turned off” in vivo in the absence of B cell contact and maintenance signals from B cell membrane associated LT-β (Mackay and Browning, 1998; Mackay et al., 1997). To determine if FDC-M1 selected FDCs survive in culture and maintain their surface phenotype, highly enriched preparations of FDCs were cultured in 6-well culture plates (35 mm diameter) to minimize contact with the small number of remaining B cells for two weeks. After one or two weeks in culture, over 80% of the FDCs that were viable at the beginning of the culture was still alive and labeled for surface expression of CD21/CD35, CD32, CD40, ICAM-1 and VCAM by flow cytometry (Fig. 13). A marked decrease in the surface expression

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fig. 11. Positively selected FDCs enhanced LPS mediated proliferation of B cells. B cells stimulated with LPS were cultured with increasing numbers of FDCs with ratios from 1:160 to 1:20 and proliferation was measured by 3H-thymidine uptake, which is plotted as counts per minute (CPM). Regression analysis on the dose response yielded an r2 of 0.8. Error bars indicate standard error of the mean (SEM). Results are representative of at least three independent experiments.

of CD21, CD32 and CD40 was apparent on day 7 and CD21 and CD40 decreased further by day 14, but significant amounts of all markers persisted through day 14 (Fig. 13). In addition, FDCs maintained most of their ICAM-1 and VCAM levels through the 14-day culture period (Fig. 13).

91

dendrites on FDCs (Szakal et al., 1985; Schnizlein et al., 1985). However, these extensive fragile processes appear to have been sheared from the cells passing through the LS column (see Fig. 1). Nevertheless, electron microscopy confirmed that some dendritic processes persist after isolation (Figs. 4 and 5) but these processes were not apparent under the light microscope. We now rely on flow cytometric data showing ∼ 90% of the viable cells with surface IgG or FDC-M1 for quality control. The biggest technical challenge is the enzyme digestion step, which involves dislodging FDCs from the stromal tissues with minimal damage, and that takes practice. As described in the methods section under “Isolation of FDCs” a variety of collagenases may be used but in our experience a lot of collagenase D that digests the tissues quickly and almost completely gives FDCs the best biological activity. Nevertheless, not all lots of collagenase D digest well and we purchase large amounts when we find a lot that works well. All other steps in the protocol are routine for cellular immunology laboratories and it is not difficult to obtain enough FDCs for in vitro studies. If increased

4. Discussion Here we describe methodology for isolating murine FDCs using magnetic cell separation technology in combination with irradiation and enzymatic digestion. A substantial number of highly enriched FDCs (∼ 90% pure population), useful for a variety of purposes, can be obtained using this method. The phenotype of these cells is consistent with FDCs in that they express FDCM1, CD16/CD32, CD21/CD35, CD40, and are surface IgG reactive but lack B220, Thy-1, and F4/80. FDCs are known to trap and retain ICs and overnight incubation with ICs followed by labeling with anti-IgG indicated that the FDC-M1 positive cells trapped and retained ICs and maintain this distinctive FDC activity. In contrast, ICs trapped by Fc receptors on macrophages or B cells were largely cleared from their surfaces after 18 h of culture. These results together with the electron microscopy, which confirmed that the vast majority of cells exhibited cytological features consistent with FDCs, provide assurance that the FDC preparations are in the 90% purity range. We find the flow cytometric methods to be helpful in routine validation of the isolation procedure. In previous studies we used cells from BSA or Percoll gradients and our quality control included examining the cells for processes using a microscope with Nomarski optics to view the extensive

Fig. 12. FDCs enhanced in vitro production of IgG antibody. (A) Lymphocytes stimulated with LPS were cultured in the presence or absence of FDCs for 14 days. The culture media was replaced on day 7 and the amount of IgG in the supernatants on day 14 was measured by ELISA. (B) OVA sensitized lymphocytes were stimulated in vitro with immune complexes (100 ng OVA and 600 ng anti-OVA) and cultured in the presence or absence of FDCs. The culture media was replaced on day 7 and supernatants were assayed for OVA specific IgG on day 14 of culture. Results are representative of at least three different experiments done in triplicates. Error bars indicate standard deviation of the mean. ⁎ denotes a p value of <0.01.

92

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fig. 13. Expression of surface molecules associated with FDC function decreased during in vitro culture of FDCs. FDCs cultured in vitro for 7 or 14 days in the absence of T or B cell contact were labeled with CD21/CD35 (Panel A), CD32 (Panel B), CD40 (Panel C), ICAM (Panel D) and VCAM (Panel E) and analyzed by flow cytometry. The legend indicates culture conditions for all histograms.

FDC yield is desirable, we have observed that the number of FDCs per mouse doubled when mice were immunized with sheep red blood cells 10–14 days prior to FDC isolation (Unpublished observation). Although not studied here, cyclophosphamide at 300 mg/kg is also effective in eliminating most murine lymphocytes without seriously damaging FDCs and should work in place of irradiation (Phipps et al., 1984, 1981; Schwarz et al., 1999). B cell factors such as lymphotoxin alpha, lymphotoxin beta and tumor necrosis factor have been shown to be important for the initial development and maintenance of FDC-networks and phenotype in vivo (Le Hir et al., 1995; Endres et al., 1999; Mackay and Browning, 1998). Hence, the time between irradiation

(B cell depletion) and FDC isolation was kept to a minimum (24 h) where our goal was to obtain FDCs with optimal function. Nevertheless, we were able to maintain viable FDCs in long-term cultures with detectable levels of phenotypic surface markers with very few B cells present. We also used these methods to isolate splenic FDCs and the yield of cells was good. However, the spleen has more contaminating cells and in our experience lymph node FDCs have been more reliable in functional studies and we focused on lymph nodes for this reason. FDCs appear to be important in the initial development and maintenance of secondary lymphoid tissue architecture. Specifically, membrane bound molecules on

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

FDCs like TNFR1 and LTβR (Endres et al., 1999; Le Hir et al., 1995; Matsumoto et al., 1996) and follicular homing chemokines like CXCL13 (Cyster et al., 2000; Estes et al., 2004) secreted by FDCs are important in regulating lymphoid structure. Highly enriched FDCs obtained by the technique described here should be helpful in further characterizing these functions in in vitro. Recent evidence suggests that FDCs are able to modulate gene expression resulting in FDCs with a resting phenotype in primary follicles or an activated phenotype in secondary follicles (Aydar et al., 2003; Balogh et al., 2001). Highly enriched preparations of FDCs obtained by positive selection should be helpful in future genetic studies. Preliminary experiments suggest that these purified FDCs are a good source of mRNA for gene expression studies and changes in gene expression for CD23a and CD23b, CD32, CD54, CD106, and CD320 appear to take place when FDCs are stimulated with anti-CD40 plus cytokines (Sukumar et al., 2006) or encounter ICs (El Shikh et al., manuscript in preparation). FDCs with appropriate ICs have remarkable accessory activity when interacting with B cells (Kosco et al., 1992; Burton et al., 1993; Lindhout et al., 1995; Wu et al., 1996; Schwarz et al., 1999; Qin et al., 1999; Tew et al., 2001; Fakher et al., 2001; Aydar et al., 2004) and these purified FDCs exhibited all of the known activities. Controls in the previous work included removing the FDC-M1 positive cells and documenting that the remaining cells consisting of some lymphocytes, fibroblasts, macrophages and other radio resistant cells did not have the FDC accessory activity (Burton et al., 1993; Wu et al., 1996; Schwarz et al., 1999; Qin et al., 1999, 2000; Payet-Jamroz et al., 2001). Some information on FDC-B lymphocyte interactions at the molecular level is summarized in a recent review (Tew et al., 2001) but much more information is needed. The lack of information on FDCs is attributable in large measure to the difficulty in isolating functionally active cells to work with in vitro models. FDCs isolated using the methods described here exhibited accessory activities including: promotion of class switching, blocking apoptosis, enhancing B cell proliferation and antibody production including selecting cells for high affinity Ab production (Figs. 9–12) (Aydar et al., 2005). Hence, we believe these methods will be well suited for future research aimed at delineating molecules on FDCs that are critical for accessory activities. Acknowledgments We like to thank Dr. Marie Kosco-Vilbois for FDCM1 hybridoma, Miltenyi biotec for the initial samples of

93

various microbeads and Roche Diagnostics for samples of liberase enzymes. References Aydar, Y., Balogh, P., Tew, J.G., Szakal, A.K., 2003. Altered regulation of FcgammarII on aged follicular dendritic cells correlates with immunoreceptor tyrosine-based inhibition motif signaling in B cells and reduced germinal center formation. J. Immunol. 171 (11), 5975. Aydar, Y., Wu, J., Song, J., Szakal, A.K., Tew, J.G., 2004. FcgammarII expression on follicular dendritic cells and immunoreceptor tyrosine-based inhibition motif signaling in B cells. Eur. J. Immunol. 34 (1), 98. Aydar, Y., Sukumar, S., Szakal, A.K., Tew, J.G., 2005. The influence of immune complex-bearing follicular dendritic cells on the IgM response, Ig class switching, and production of high affinity IgG. J. Immunol. 174 (9), 5358. Balogh, P., Aydar, Y., Tew, J.G., Szakal, A.K., 2001. Ontogeny of the follicular dendritic cell phenotype and function in the postnatal murine spleen. Cell. Immunol. 214 (1), 45. Brown, K.L., Stewart, K., Ritchie, D.L., Mabbott, N.A., Williams, A., Fraser, H., Morrison, W.I., Bruce, M.E., 1999. Scrapie replication in lymphoid tissues depends on prion protein-expressing follicular dendritic cells. Nat. Med. 5 (11), 1308. Burton, G.F., Conrad, D.H., Szakal, A.K., Tew, J.G., 1993. Follicular dendritic cells and B cell costimulation. J. Immunol. 150 (1), 31. Burton, G.F., Masuda, A., Heath, S.L., Smith, B.A., Tew, J.G., Szakal, A.K., 1997. Follicular dendritic cells (Fdc) in retroviral infection: host/pathogen perspectives. Immunol. Rev. 156, 185. Burton, G.F., Keele, B.F., Estes, J.D., Thacker, T.C., Gartner, S., 2002. Follicular dendritic cell contributions to HIV pathogenesis. Semin. Immunol. 14 (4), 275. Clark, E.A., Grabstein, K.H., Shu, G.L., 1992. Cultured human follicular dendritic cells. Growth characteristics and interactions with B lymphocytes. J. Immunol. 148 (11), 3327. Cormann, N., Heinen, E., Kinet-Denoel, C., Tsunoda, R., Simar, L.J., 1988. Isolation of follicular dendritic cells from human tonsils and adenoids. In vitro culture. Adv. Exp. Med. Biol. 237, 171. Cyster, J.G., Ansel, K.M., Reif, K., Ekland, E.H., Hyman, P.L., Tang, H.L., Luther, S.A., Ngo, V.N., 2000. Follicular stromal cells and lymphocyte homing to follicles. Immunol. Rev. 176, 181. Endres, R., Alimzhanov, M.B., Plitz, T., Futterer, A., Kosco-Vilbois, M.H., Nedospasov, S.A., Rajewsky, K., Pfeffer, K., 1999. Mature follicular dendritic cell networks depend on expression of lymphotoxin beta á receptor by radioresistant stromal cells and of lymphotoxin beta á and tumor necrosis factor by B cells. J. Exp. Med. 189 (1), 159. Ennas, M.G., Chilosi, M., Scarpa, A., Lantini, M.S., Cadeddu, G., Fiore-Donati, L., 1989. Isolation of multicellular complexes of follicular dendritic cells and lymphocytes: immunophenotypical characterization, electron microscopy and culture studies. Cell Tissue Res. 257 (1), 9. Estes, J.D., Thacker, T.C., Hampton, D.L., Kell, S.A., Keele, B.F., Palenske, E.A., Druey, K.M., Burton, G.F., 2004. Follicular dendritic cell regulation of CXCR4-mediated germinal center CD4 T cell migration. J. Immunol. 173 (10), 6169. Fakher, M., Wu, J., Qin, D., Szakal, A., Tew, J., 2001. Follicular dendritic cell accessory activity crosses MHC and species barriers. Eur. J. Immunol. 31 (1), 176.

94

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95

Fox, C.H., Tenner-Racz, K., Racz, P., Firpo, A., Pizzo, P.A., Fauci, A.S., 1991. Lymphoid germinal centers are reservoirs of human immunodeficiency virus type 1 RNA. J. Infect. Dis. 164 (6), 1051. Fraser, H., Farquhar, C.F., 1987. Ionising radiation has no influence on scrapie incubation period in mice. Vet. Microbiol. 13 (3), 211. Gray, D., Kosco, M., Stockinger, B., 1991. Novel pathways of antigen presentation for the maintenance of memory. Int. Immunol. 3 (2), 141. Haley, S.T., Tew, J.G., Szakal, A.K., 1995. The monoclonal antibody FDC-M1 recognizes possible follicular dendritic cell precursors in the blood and bone marrow. Adv. Exp. Med. Biol. 378, 289. Hanna Jr., M.G., Szakal, A.K., 1968. Localization of 125I-labeled antigen in germinal centers of mouse spleen: histologic and ultrastructural autoradiographic studies of the secondary immune reaction. J. Immunol. 101 (5), 949. Heath, S.L., Tew, J.G., Tew, J.G., Szakal, A.K., Burton, G.F., 1995. Follicular dendritic cells and human immunodeficiency virus infectivity. Nature 377 (6551), 740. Heinen, E., Kinet-Denoel, C., Radoux, D., Simar, L., 1985. Mouse lymph node follicular dendritic cells: quantitative analysis and isolation. Adv. Exp. Med. Biol. 186, 171. Heinen, E., Tsunoda, R., Marcoty, C., Antoine, N., Bosseloir, A., Cormann, N., Simar, L., 1993. Follicular dendritic cells: isolation procedures, short and long term cultures. Adv. Exp. Med. Biol. 329, 333. Helm, S.L., Burton, G.F., Szakal, A.K., Tew, J.G., 1995. Follicular dendritic cells and the maintenance of IgE responses. Eur. J. Immunol. 25 (8), 2362. Huber, C., Thielen, C., Seeger, H., Schwarz, P., Montrasio, F., Wilson, M.R., Heinen, E., Fu, Y.X., Miele, G., Aguzzi, A., 2005. Lymphotoxin-beta receptor-dependent genes in lymph node and follicular dendritic cell transcriptomes. J. Immunol. 174 (9), 5526. Humphrey, J.H., Grennan, D., 1982. Isolation and properties of spleen follicular dendritic cells. Adv. Exp. Med. Biol. 149, 823. Kitamoto, T., Muramoto, T., Mohri, S., Doh-Ura, K., Tateishi, J., 1991. Abnormal isoform of prion protein accumulates in follicular dendritic cells in mice with Creutzfeldt–Jakob disease. J. Virol. 65 (11), 6292. Kosco, M.H., Pflugfelder, E., Gray, D., 1992. Follicular dendritic cell-dependent adhesion and proliferation of B cells in vitro. J. Immunol. 148 (8), 2331. Kosco-Vilbois, M.H., Gray, D., Scheidegger, D., Julius, M., 1993. Follicular dendritic cells help resting B cells to become effective antigen-presenting cells: induction of B7/BB1 and upregulation of major histocompatibility complex class II molecules. J. Exp. Med. 178 (6), 2055. Le Hir, M., Bluethmann, H., Kosco-Vilbois, M.H., Muller, M., di Padova, F., Moore, M., Ryffel, B., Eugster, H.P., 1995. Tumor necrosis factor receptor-1 signaling is required for differentiation of follicular dendritic cells, germinal center formation, and full antibody responses. J. Inflam. 47 (1–2), 76. Lilet-Leclercq, C., Radoux, D., Heinen, E., Kinet-Denoel, C., Defraigne, J.O., Houben-Defresne, M.P., Simar, L.J., 1984. Isolation of follicular dendritic cells from human tonsils and adenoids: I. Procedure and morphological characterization. J. Immunol. Methods 66 (2), 235. Lindhout, E., Lakeman, A., de Groot, C., 1995. Follicular dendritic cells inhibit apoptosis in human B lymphocytes by a rapid and irreversible blockade of preexisting endonuclease. J. Exp. Med. 181 (6), 1985.

Liu, Don, et al., 1997. Immunohistochemical Phenotype of Murine Follicular Dendritic Cells (FDCs). Dendritic Cells: Proceedings of the 7th Annual Meeting of Japanese Dendritic Cells Society, vol. 7, p. 47. Ref Type: Journal (Full). Mabbott, N.A., Williams, A., Farquhar, C.F., Pasparakis, M., Kollias, G., Bruce, M.E., 2000. Tumor necrosis factor alpha-deficient, but not interleukin-6-deficient, mice resist peripheral infection with scrapie. J. Virol. 74 (7), 3338. Mackay, F., Browning, J.L., 1998. Turning off follicular dendritic cells. Nature 395 (6697), 26. Mackay, F., Majeau, G.R., Lawton, P., Hochman, P.S., Browning, J.L., 1997. Lymphotoxin but not tumor necrosis factor functions to maintain splenic architecture and humoral responsiveness in adult mice. Eur. J. Immunol. 27 (8), 2033. Manuelidis, L., Zaitsev, I., Koni, P., Lu, Z.Y., Flavell, R.A., Fritch, W., 2000. Follicular dendritic cells and dissemination of Creutzfeldt– Jakob disease. J. Virol. 74 (18), 8614. Marcoty, C., Heinen, E., Antoine, N., Tsunoda, R., Simar, L.J., 1993. Rapid and selective isolation of follicular dendritic cells by low speed centrifugations on discontinuous BSA gradients. Adv. Exp. Med. Biol. 329, 425. Matsumoto, M., Mariathasan, S., Nahm, M.H., Baranyay, F., Peschon, J.J., Chaplin, D.D., 1996. Role of lymphotoxin and the type I TNF receptor in the formation of germinal centers. Science 271 (5253), 1289. McBride, P.A., Eikelenboom, P., Kraal, G., Fraser, H., Bruce, M.E., 1992. PrP protein is associated with follicular dendritic cells of spleens and lymph nodes in uninfected and scrapie-infected mice. J. Pathol. 168 (4), 413. Parmentier, H.K., van der Linden, J.A., Krijnen, J., van Wichen, D.F., Rademakers, L.H., Bloem, A.C., Schuurman, H.J., 1991. Human follicular dendritic cells: isolation and characteristics in situ and in suspension. Scand. J. Immunol. 33 (4), 441. Payet-Jamroz, M., Helm, S.L., Wu, J., Kilmon, M., Fakher, M., Basalp, A., Tew, J.G., Szakal, A.K., Noben-Trauth, N., Conrad, D. H., 2001. Suppression of IgE responses in CD23-transgenic animals is due to expression of CD23 on nonlymphoid cells. J. Immunol. 166 (8), 4863. Phipps, R.P., Mandel, T.E., Tew, J.G., 1981. Effect of immunosuppressive agents on antigen retained in lymphoid follicles and collagenous tissue of immune mice. Cell. Immunol. 57 (2), 505. Phipps, R.P., Mandel, T.E., Schnizlein, C.T., Tew, J.G., 1984. Anamnestic responses induced by antigen persisting on follicular dendritic cells from cyclophosphamide-treated mice. Immunology 51 (2), 387. Piris, M.A., Rivas, C., Morente, M., Rubio, C., Martin, C., Olivia, H., 1987. Persistent and generalized lymphadenopathy: a lesion of follicular dendritic cells? An immunohistologic and ultrastructural study. Am. J. Clin. Pathol. 87 (6), 716. Qin, D., Wu, J., Carroll, M.C., Burton, G.F., Szakal, A.K., Tew, J.G., 1998. Evidence for an important interaction between a complement-derived CD21 ligand on follicular dendritic cells and CD21 on B cells in the initiation of IgG responses. J. Immunol. 161 (9), 4549. Qin, D., Wu, J., Burton, G.F., Szakal, A.K., Tew, J.G., 1999. Follicular dendritic cells mediated maintenance of primary lymphocyte cultures for long-term analysis of a functional in vitro immune system. J. Immunol. Methods 226 (1–2), 19. Qin, D., Wu, J., Vora, K.A., Ravetch, J.V., Szakal, A.K., Manser, T., Tew, J.G., 2000. Fc gamma receptor IIB on follicular dendritic cells regulates the B cell recall response. J. Immunol. 164 (12), 6268.

S. Sukumar et al. / Journal of Immunological Methods 313 (2006) 81–95 Racz, P., Tenner-Racz, K., Schmidt, H., 1989. Follicular dendritic cells in HIV-induced lymphadenopathy and AIDS. APMIS Suppl 8, 16. Schmitz, J., Petrasch, S., van Lunzen, J., Racz, P., Kleine, H.D., Hufert, F., Kern, P., Schmitz, H., Tenner-Racz, K., 1993. Optimizing follicular dendritic cell isolation by discontinuous gradient centrifugation and use of the magnetic cell sorter (MACS). J. Immunol. Methods 159 (1–2), 189. Schnizlein, C.T., Kosco, M.H., Szakal, A.K., Tew, J.G., 1985. Follicular dendritic cells in suspension: identification, enrichment, and initial characterization indicating immune complex trapping and lack of adherence and phagocytic activity. J. Immunol. 134 (3), 1360. Schwarz, Y.X., Yang, M., Qin, D., Wu, J., Jarvis, W.D., Grant, S., Burton, G.F., Szakal, A.K., Tew, J.G., 1999. Follicular dendritic cells protect malignant B cells from apoptosis induced by anti-Fas and antineoplastic agents. J. Immunol. 163 (12), 6442. Sellheyer, K., Schwarting, R., Stein, H., 1989. Isolation and antigenic profile of follicular dendritic cells. Clin. Exp. Immunol. 78 (3), 431. Smith, B.A., Gartner, S., Liu, Y., Perelson, A.S., Stilianakis, N.I., Keele, B.F., Kerkering, T.M., Ferreira-Gonzalez, A., Szakal, A.K., Tew, J.G., Burton, G.F., 2001. Persistence of infectious HIV on follicular dendritic cells. J. Immunol. 166 (1), 690. Stahmer, I., Zimmer, J.P., Ernst, M., Fenner, T., Finnern, R., Schmitz, H., Flad, H.D., Gerdes, J., 1991. Isolation of normal human follicular dendritic cells and CD4-independent in vitro infection by human immunodeficiency virus (HIV-1). Eur. J. Immunol. 21 (8), 1873. Sukumar, S., Conrad, D.H., Szakal, A.K., Tew, J.G., 2006. Differential T cell-mediated regulation of CD23 (Fc epsilonRII) in B cells and follicular dendritic cells. J. Immunol. 176 (8), 4811.

95

Szakal, A.K., Hanna Jr., M.G., 1968. The ultrastructure of antigen localization and viruslike particles in mouse spleen germinal centers. Exp. Mol. Pathol. 8 (1), 75. Szakal, A.K., Gieringer, R.L., Kosco, M.H., Tew, J.G., 1985. Isolated follicular dendritic cells: cytochemical antigen localization, Nomarski, SEM, and TEM morphology. J. Immunol. 134 (3), 1349. Szakal, A.K., Kosco, M.H., Tew, J.G., 1988. A novel in vivo follicular dendritic cell-dependent iccosome-mediated mechanism for delivery of antigen to antigen-processing cells. J. Immunol. 140 (2), 341. Tew, J.G., Thorbecke, G.J., Steinman, R.M., 1982. Dendritic cells in the immune response: characteristics and recommended nomenclature (A report from the Reticuloendothelial Society Committee on Nomenclature). J. Reticuloendothel. Soc. 31 (5), 371. Tew, J.G., Wu, J., Qin, D., Helm, S., Burton, G.F., Szakal, A.K., 1997. Follicular dendritic cells and presentation of antigen and costimulatory signals to B cells. Immunol. Rev. 156, 39. Tew, J.G., Wu, J., Fakher, M., Szakal, A.K., Qin, D., 2001. Follicular dendritic cells: beyond the necessity of T-cell help. Trends Immunol. 22 (7), 361. Tsiagbe, V.K., Linton, P.J., Thorbecke, G.J., 1992. The path of memory B-cell development. Immunol. Rev. 126, 113. Tsunoda, R., Nakayama, M., Onozaki, K., Heinen, E., Cormann, N., Kinet-Denoel, C., Kojima, M., 1990. Isolation and long-term cultivation of human tonsil follicular dendritic cells. Virchows Arch., B Cell Pathol. Incl. Mol. Pathol. 59 (2), 95. Wu, J., Qin, D., Burton, G.F., Szakal, A.K., Tew, J.G., 1996. Follicular dendritic cell-derived antigen and accessory activity in initiation of memory IgG responses in vitro. J. Immunol. 157 (8), 3404.