The germinal centers of secondary lymphoid follicles provide a unique environment where high-affinity antibody-forming cells and memory B cells are generated. The major resident cell type in germinal centers responsible for these processes are the follicular dendritic cells (FDCs) that retain the antigen in the form of immune complexes exposed to maturing lymphocytes on the surface of their elaborate cell extensions (Allen and Cyster,2008). However, despite the abundance of functional data, little is known about the molecular regulation of FDC development and maturation. Studies using gene-targeted mice have shown that (lymphotoxin-β receptor signaling being provided) FDC development critically depends on the functional tumor necrosis factor receptor-1 (TNFR1) pathway: if this signaling route is disrupted, FDC clusters fail to develop (Le Hir et al.,1996; Liu and Banchereau,1996; Pasparakis et al.,1996,1997; Matsumoto et al.,1997a,1997b). Recently, it has been demonstrated that lipopolysaccharide (LPS) induces activation of FDCs in vitro (El Shikh et al.,2007). Whether this also occurs in vivo and what is the role of TNFR1 in this situation, however, remained uncertain. Therefore, the aim of our study was to investigate: (i) the possibility of FDC activation by LPS in vivo and (ii) the role of TNFR1 signaling in this process.
Here, we show that LPS injections induce the clear and prominent morphological signs of FDC development/activation in vivo, which is independent of TNFR1 pathway.
MATERIALS AND METHODS
Mice and Lipopolysaccharide Treatment
C57BL/6 mice were obtained from Charles River GmbH (Sulzfeld, Germany), whereas TNFR1−/− mice were kindly provided by Klaus Pfeffer (Düsseldorf, Germany; Pfeffer et al.,1993). Mice of both sexes, between 8 and 12 weeks of age, were used (n = 5 mice per group). This is the usual age range used in our laboratory. The animals were housed and bred under specific pathogen-free conditions in the Central Animal Facility of the University of Lübeck (Germany). Permission to perform these animal experiments was issued by the Ministry of Nature and Environment (V 252-72241.122-1 (24-3/02)).
Mice were injected into the tail vein with 100 μg of LPS in 0.5 ml of phosphate buffered saline (PBS). After 24 hr, the animals were killed by CO2 and splenic tissue samples were collected.
Immunohistochemistry and Antibodies
The splenic tissue was removed, snap-frozen in liquid nitrogen, and stored at −80°C. Cryostat sections (12 μm) were cut, air-dried at room temperature for 2 hr and fixed in methanol–acetone (1:1 (v/v), 10 min at −20°C) followed by fixation in 4% paraformaldehyde (45 min at 4°C). For demonstration of FDCs (Milićević et al.,2011), the sections were incubated with the primary 1:100 diluted rat anti-mouse FDC-M1 monoclonal antibody (BD Pharmingen, Heidelberg, Germany) for 1 hr. Secondary polyclonal rabbit anti-rat IgG (Dako, Glostrup, Denmark; E 0468), which does not cross-react with mouse Ig (the cross-reaction with mouse Ig was less than 0.3%, as determined by ELISA), with an incubation time of 30 min was used. Thereafter, the streptavidin/alkaline phosphatase detection system was used and labeled cells were revealed by the Fast Blue. Omitting of the primary antibody or use of irrelevant mouse antibody, matched isotype, were performed as negative controls.
The widely accepted morphological criteria for cell activation were used in this study to assess the overall activation of FDCs. These morphological criteria are as follows: enlargement of cells, increased thickness, length and number of cell prolongations.
RESULTS AND DISCUSSION
LPS Induces In Vivo Activation of FDCs in Normal Mice
We first wished to verify the distribution and morphological features of FDCs in splenic white pulp of normal intact animals. We noted that clusters of FDCs were localized in lymphoid follicles (Fig. 1a). These cells were large, with prominent cytoplasmic prolongations (Fig. 1a, arrowheads). Sometimes, the elongated processes stretching between the lymphocytes were seen. Additionally, FDC-M1+ cells could also be clearly identified in the marginal zone providing a rim-like structure around the splenic white pulp (Fig. 1a, arrows). These cells showed a weak, finely granular, cytoplasmic FDC-M1+ reaction. Only rarely, such cells (located deeper toward the follicles) were larger in size, with intermediate intensity of FDC-M1+ reaction (Fig. 1a). Generally, the findings regarding the localization and appearance of FDCs in lymphoid follicles were in full agreement with the data in the literature (Allen and Cyster,2008). However, the presence of FDC-M1+ cells in the marginal zone, which are likely to represent the FDC precursors, has not been reported earlier in normal mice.
Next, we wished to investigate the effects of LPS on morphology of FDCs in vivo. We found that LPS application induced the prominent signs of FDC activation in the splenic follicles of normal animals: the number of these cells tremendously increased, they were markedly enlarged in size, with abundant cytoplasm, very prominent cell prolongations, and increased intensity of FDC-M1+ reaction (Fig. 1b). The rim of FDC-M1+ cells in the marginal zone also became more prominent. These cells were increased in number and size showing stronger FDC-M1+ reaction (Fig. 1b, arrows). Sometimes, the distinction between marginal zone and follicular FDCs could not be made, as these two populations appeared intermingled due to the increased numbers (Fig. 1b). This experiment showed that in normal mice after LPS injection, FDCs in vivo display clear morphological signs of activation in the lymphoid follicles as well as in the marginal zone, that is, enlargement of cells, increased thickness, length and number of cell prolongations. This finding is in line with the results of in vitro studies, which demonstrated that LPS induces the activation of FDCs (El Shikh et al.,2007). Furthermore, very recently, similar in vivo differentiation and activation of FDCs was reported after LPS in mouse lymph nodes (Garin et al.,2010). However, the changes of FDC-M1+ cells in the marginal zone have not been described earlier and deserve further attention providing a possible model to study the transition of putative precursor cells to full-fledged FDCs.
LPS-Induced In Vivo Development and Activation of FDCs is TNFR1-Independent
To study the role of TNFR1 in LPS-induced activation of FDCs, we first investigated the presence of FDC-M1+ cells in the splenic white pulp of intact mice deficient in this receptor molecule. In TNFR1−/− mice, the FDCs were localized only in the marginal zone and provided a broad rim surrounding the white pulp (Fig. 1c). This rim of FDC-M1+ cells was markedly broader in comparison with normal animals. As the lymphoid follicles were absent in TNFR1−/− animals, the clusters of FDCs, similar to those in normal mice, could not be seen (Fig. 1c). In TNFR1−/− mice, the FDCs appeared morphologically underdeveloped: small in size, most often with a dot-like appearance and very sparse, inconspicuous cellular prolongations (Fig. 1c, arrows). Only extremely rarely, some larger FDC-M1+ cells resembling well-developed FDCs (larger in size, with prominent processes and strong FDC-M1+ reaction) could be seen interspersed with other FDC-M1+ cells in the marginal zone (Fig. 1c, arrowheads). Earlier, it has been claimed that TNF/TNFR1 signaling is of critical importance for the development of FDCs. Hence, the animals deficient in these molecules were always considered to be devoid of FDCs (Le Hir et al.,1996; Liu and Banchereau,1996; Pasparakis et al.,1996,1997; Matsumoto et al.,1997a,1997b). However, Pasparakis et al. (2000) and Victoratos et al. (2006) opposed the prevailing opinion that FDCs are completely lacking in the spleen of TNFR1-deficient animals and reported on the presence of FDC-M1+ putative FDC precursors in the splenic marginal zone. Here, we confirm their findings. This shows that FDCs are not completely absent in mice devoid of TNFR1 signaling but appear underdeveloped in the marginal zone.
To investigate whether these FDC-M1+ cells from the marginal zone can give rise to full-fledged FDCs, we treated TNFR1−/− mice with LPS injections. LPS application markedly affected the FDCs in TNFR1−/− animals and they displayed all morphological signs of activation. The number and size of these cells unquestionably increased (Fig. 1d). The cells spread throughout the splenic white pulp. They morphologically fully corresponded to FDCs observed in LPS-treated normal animals being very large with prominent cell extensions (Fig. 1d). The FDC-M1+ cells in the marginal zone also became enlarged (Fig. 1d, arrows). However, their frequency appeared somewhat decreased in comparison with untreated TNFR1−/− mice. Negative controls of LPS-treated normal and LPS-treated TNFR1−/− mice did not reveal any FDC-M1+ cells (Fig. 1e,f). This is the first report, based on morphological criteria, that the development and activation of FDCs may be achieved by LPS stimulation in the absence of TNFR1. This finding strongly suggests that TNFR1 signaling is not required for development and activation of FDCs, because LPS is sufficient to induce these processes in the absence of this receptor molecule. One is tempted to propose that this is due to low availability of LPS from the intestinal tract of TNFR1−/− mice. Such opinion is confirmed by the fact that parenteral application of LPS may overcome this defect and induce the development and activation of FDCs in TNFR1−/− mice, as demonstrated in this work. The germ-free animals, which are devoid of intestinal bacteria and suffer from impaired intestinal absorption (Tennant et al.,1970), also show the absence of FDC networks (Fossum and Vaaland,1983). Unfortunately, the data on the intestinal absorption in TNFR1−/− animals are not available, and further studies are necessary to elucidate this issue.
According to data in the literature and our findings presented herein, three-stage development of FDCs seems plausible: (1) lymphotoxin-β receptor signaling induces the development of FDC precursors in the marginal zone (as seen in TNFR1−/− mice, Pasparakis et al.,2000), (2) toll-like receptor-4 (TLR4) signaling induces the development of full-fledged FDCs in the splenic B-cell zone (as seen in LPS-treated TNFR1−/− mice), and (3) TNFR1 signaling elicits the clustering of FDCs accompanied by lymphoid follicle formation (as seen in normal mice, Victoratos et al.,2006).
Finally, our study shows that LPS, as a prototypic antigen of innate immunity, very promptly (within 24 hr) induces in vivo activation of FDCs, which are positioned at the very heart of adaptive immunity. This shows that, at least in this case, the activation of both systems is simultaneously initiated and functionally synchronized confirming the academic nature of this division.
In conclusion, we demonstrate that: (i) LPS induces in vivo activation of FDCs and (ii) LPS-induced FDC development and activation are TNFR1 independent.