Increased susceptibility to oral Trichuris muris infection in the specific absence of CXCR5+ CD11c+ cells

Summary Trichuris muris is a natural mouse helminth pathogen which establishes infection specifically in the caecum and proximal colon. The rapid expulsion of T. muris in resistant mouse strains is associated with the induction of a protective T helper cell type 2 (Th2)‐polarized immune response. Susceptible mouse strains, in contrast, mount an inappropriate Th1 response to T. muris infection. Expression of the chemokine CXCL13 by stromal follicular dendritic cells attracts CXCR5‐expressing cells towards the B‐cell follicles. Previous studies using a complex in vivo depletion model have suggested that CXCR5‐expressing conventional dendritic cells (cDC) help regulate the induction of Th2‐polarized responses. Here, transgenic mice with CXCR5 deficiency specifically restricted to CD11c+ cells were used to determine whether the specific absence CXCR5 on CD11c+ cells such as cDC would influence susceptibility to oral T. muris infection by affecting the Th1/Th2 balance. We show that in contrast to control mice, those which lacked CXCR5 expression on CD11c+ cells failed to clear T. muris infection and developed cytokine and antibody responses that suggested a disturbed Th1/Th2 balance with enhanced IFN‐γ expression. These data suggest an important role of CXCR5‐expressing CD11c+ cells such as cDC in immunity to oral T. muris infection.


| INTRODUC TI ON
Mononuclear phagocytes (MNP) arise from precursors in the bone marrow and comprise a heterogeneous population of monocytes, conventional dendritic cells (cDC) and tissue macrophages. The intestinal mucosa is populated by distinct MNP populations including MNP expressing the fractalkine receptor CX3CR1 and subsets of cDC marked by differential expression of the integrins CD11b and CD103. 1 The cDC are specialized antigen-presenting cells, and antigen presentation by cDC to uninfected cognate CD4 + T cells may induce T helper cell type 1 (Th1), Th2 or Th17 responses dependent on their subset. [2][3][4][5][6] The development of the intestinal cDC1 subclass of CD103 + CD11b − cDC is dependent on the transcription factors IRF8, BATF3 and ID2. [7][8][9][10] In contrast, the lack of IRF4 or Notch2 results in a loss of cDC2 subclass CD103 + CD11b + cDC and reduced numbers of CD103 − CD11b + cDC in the intestine-draining mesenteric lymph nodes (MLN). [3][4][5][6] Trichuris muris is a natural nematode parasite of mice whose larvae hatch in the caecum and proximal colon and invade the epithelium.
Resistance to high-level infection with T. muris varies considerably between different conventional mouse strains. In resistant mouse strains, the rapid expulsion of T. muris before the adult worms reach fecundity is associated with the induction a protective Th2-polarized immune response characterized by the production of the cytokines interleukin (IL)-4, IL-5, IL-9 and IL-13. [11][12][13][14] In contrast, susceptible mouse strains mount an inappropriate Th1-polarized response to T. muris infection that is associated with high levels of IFNγ and IL-12, and results in susceptibility and persistent infection. 15,16 While the development of Th1 immunity is well understood and regulated by cDC-derived production of the cytokine interleukin (IL)-12, the factors that regulate the development of Th2 immunity are less clear. Expression of the chemokine CXCL13 by stromal follicular dendritic cells (FDC) and follicular stromal cells mediates the attraction of CXCR5-expressing cells, including cDC, towards and into the B-cell follicles. [17][18][19][20] A requirement for CXCR5-expressing cDC has been suggested for the efficient development of Th2 responses to the intestinal parasite Heligmosomoides polygyrus. 21 This evidence, however, was derived from the use of a complex irradiation chimeric mouse model. Briefly, C57BL/6 wild-type mice were first lethally γ-irradiated and reconstituted with an 80:20 mixture of bone marrow from CD11c-DTR mice (in which CD11c + cells can be transiently ablated by diphtheria toxin treatment 22 ) and Cxcr5 −/− mice. After reconstitution, purified uninfected CD4 + T cells were then transferred into these chimeric mice before they were infected with H. polygyrus. Data from the use of these "DC-Cxcr5 −/− " chimeric mice suggested that CXCR5-expressing cDC helps regulate the induction of Th2-polarized responses. 21 However, all the MNP populations in the intestine are transiently depleted in CD11c-DTR mice after diphtheria toxin treatment. 23 This may have influenced disease susceptibility as intestinal macrophages also contribute to immunity to H. polygyrus infection. 24 It is also plausible that the use of lethal irradiation may have adversely affected gut integrity and the microarchitecture of the secondary lymphoid organs.
Whether CXCR5-expressing cDC are important for the induction of protective immunity to other helminth pathogens such as T. muris was not known. Therefore, in the current study, a novel compound transgenic mouse model was used in which CXCR5 deficiency was specifically restricted to CD11c + cells, including cDC. 25 These mice were used to test the hypothesis that CXCR5-expressing CD11c + cells such as cDC are required for the induction of protective immune responses to T. muris infection.

| Mice
The following mouse strains were used in this study where indicated: CD11c-Cre 26 (strain Tg(Itgax-cre)1-1Reiz) and CXCR5 F/F (strain Cxcr5 tm1.Namt ), which have loxP sites flanking exon 2 of the Cxcr5 gene. 25 All mice were bred and maintained on C57BL/6J mice background, maintained under SPF conditions and used at 8-12 weeks of age. All studies and regulatory licences were approved by the University of Edinburgh's Ethics Committee and carried out under the authority of a UK Home Office Project Licence.
The genotypes of all mice used in this study were confirmed by the analysis of genomic or cDNA extracted from ear punch biopsies. DNA samples were analysed for the presence of CD11c-Cre using the following primers: ACTTGGCAGCTGTCTCCAAG and GCGAACATCTTCAGGTTCTG; and CXCR5 F and recombined CXCR5 F (Cxcr5 de-flox ) using the following primers: AGGAGGCCATTTCCTCAGTT; GGCTTAGGGATTGCAGTCAG; and TTCCTTAGAGCCTGGAAAAGG.

| Parasite-specific immunoglobulin (Ig) ELISA
Serum T. muris antigen-specific IgG1 and IgG2c levels were determined by ELISA as previously described. 28

| Histopathology and immunohistochemistry (IHC)
Intestines, MLN and spleens were snap-frozen at the temperature of

| Image analysis
For morphometric analysis, images were analysed using ImageJ software (http://rsb.info.nih.gov/ij/) as described on coded sections. 29 Crypt and cell counting were performed manually using the FiJi cell counter plug-in. In each instance, data were typically obtained from 14 to 37 crypts/mouse, from the intestines of 4 to 8 mice/group.
Details of the sample sizes for each parameter analysed are provided in the figure legends.

| Statistical analysis
Unless indicated otherwise, data are presented as mean ± SEM and significant differences between groups were sought using Student's

| Altered positioning of CD11c + cells in the secondary lymphoid organs of CXCR5 ΔDC mice
Throughout this study, a novel compound transgenic mouse model was used in which CXCR5 deficiency was specifically restricted to CD11c + MNP. 25 The expression of Cre recombinase under the control of the Itgax locus (encoding CD11c) in CD11c-Cre mice has been used in a variety of studies to conditionally control gene expression in cDC. 3,25,26,30,31 These mice were crossed to CXCR5 F/F mice to generate CXCR5 ΔDC mice. 25 We have previously shown that in these mice, the Cre recombinase-mediated recombination of Cxcr5 is restricted to CD11c + cDC and that the migration of their cDC towards CXCL13 is specifically impeded. 25 In the MLN and

| Steady-state serum IgG1 and IgG2c Ig levels are not altered in CXCR5 ΔDC mice
Next, the relative concentrations of total IgG1 and IgG2c Ig isotype levels were compared in the serum of uninfected CXCR5 F/F mice and CXCR5 ΔDC mice. Similar levels of total IgG1 (P = 0.590) and IgG2c (P = 0.946) were detected in mice from each group, indicating no constitutive difference in the ability to produce either Ig isotype in the steady state ( Figure 1C).

| Enhanced susceptibility to T. muris infection in CXCR5 ΔDC mice
Groups of CXCR5 F/F mice and CXCR5 ΔDC mice were next orally and IgG2c were undetectable in the sera of uninfected mice from both genotypes as expected ( Figure 2B,C, respectively).

| Altered expression of Th1/Th2 cytokines in the MLN of T. muris-infected CXCR5 ΔDC mice
The increased susceptibility of CXCR5 ΔDC mice to oral T. muris infection and their increased production of parasite-specific serum IgG2c suggested an altered Th1/Th2 balance. We therefore compared the expression of cytokine-encoding genes in mRNA from the MLN of T. muris-infected CXCR5 F/F mice and CXCR5 ΔDC mice.
The resistance of C57BL/6 mice to T. muris infection is associated with the induction of a Th2-polarized parasite-specific immune response. As anticipated, the expression of mRNA encoding the Th2 cytokines IL-4, IL-5, IL-9 and IL-13 was upregulated in the MLN of T. muris-infected CXCR5 F/F control mice when compared to uninfected mice ( Figure 3A). In contrast, the expression of Il4 and Il9 mRNA in the MLN of chronically infected CXCR5 ΔDC mice was significantly less increased when compared to uninfected control mice ( Figure 3A).
F I G U R E 1 Altered positioning of CD11c + cells in the secondary lymphoid organs of CXCR5 ΔDC mice. A, Immunohistochemical (IHC) analysis of the distribution of B cells (CD45R/B220 + cells; green, left-hand column), stromal FDC (CR1/CD35 + cells; green) and CD11c + cells (red) in the spleen and MLN of uninfected CXCR5 F/F control mice (upper panels) and CXCR5 ΔDC mice (lower panels). In the MLN and spleens of CXCR5 F/F mice, CD11c + cells were occasionally detected within the FDC-containing B-cell follicles (arrows). Few, if any, CD11c + cells were detected in the FDC-containing B-cell follicles of CXCR5 ΔDC mice. Scale bar, 100 μm. B, qRT-PCR analysis of cytokine-encoding genes in mRNA from the MLN of uninfected CXCR5 F/F control mice (closed circles) and CXCR5 ΔDC mice (open circles). Horizontal bars, median. C, Similar levels of total IgG1 and IgG2c were detected in the serum of uninfected CXCR5 F/F control mice (closed circles) and CXCR5 ΔDC mice (open circles). Data were derived from 6 mice/group In the MLN of T. muris-infected CXCR5 F/F control mice, only modest increases in mRNA encoding the Th1 cytokine IFNγ and the proinflammatory cytokines IL-1β and IL-6 were observed ( Figure 3B).
In contrast, the expression of Ifng was substantially and significantly elevated in the MLN of infected CXCR5 ΔDC mice (P < 0.0087, Mann-Whitney U test; Figure 3B). These data indicated that in the absence of CXCR5-expressing cDC, the Th1/Th2 cytokine balance in the MLN was disturbed with significantly reduced expression of IL-4 and IL-9 and significantly elevated expression of IFNγ. The expression levels of Il12a, Il12b and Tnfa (encoding the IL-12p35 and IL-12p40 subunits, and TNFα, respectively) and Il17a were similar in MLN from T. murisinfected CXCR5 F/F mice and CXCR5 ΔDC mice ( Figure 3C). A similar increase in goblet cell density following T. muris infection was also observed in CXCR5 ΔDC mice, revealing that the ability to induce goblet cell hyperplasia was unaffected.

| Influence of cDC-specific CXCR5 deficiency on the abundance of leucocytes in the large intestinal lamina propria
Leucocytes accumulate in the lamina propria of the large intestine during T. muris infection, and the characteristics of the response can differ between resistant and susceptible mouse strains. 35 We therefore compared the density of T cells and MNP within the lamina propria in the large intestines of T. muris-infected CXCR5 F/F mice and CXCR5 ΔDC mice ( Figure 5). Although similar levels of CD4 + lymphocytes were detected in the lamina propria of infected CXCR5 F/F mice and CXCR5 ΔDC mice, the number of CD8α + lymphocytes in the lamina propria of CXCR5 ΔDC mice was significantly reduced ( Figure 5A,B). Comparison of densities of CD11b + , CD11c + and CD68 + cells suggested that similar densities of CD68 + MNP were present in the large intestines of T. muris-infected CXCR5 F/F mice and CXCR5 ΔDC mice ( Figure 5C,D). However, the density of CD11b + and CD11c + MNP in T. muris-infected CXCR5 ΔDC mice was similar to uninfected CXCR5 F/F mice ( Figure 5C,D).
F I G U R E 2 Susceptibility to Trichuris muris infection is enhanced in the specific absence of CXCR5-expressing cDC in CXCR5 ΔDC mice. CXCR5 F/F mice (n = 8, closed circles) and CXCR5 ΔDC mice (n = 5, open circles) were orally infected with approximately 200 embryonated T. muris eggs and worm burdens in the large intestine compared at 30 dpi. Susceptibility to T. muris infection was significantly increased in CXCR5 ΔDC mice when compared to CXCR5 F/F control mice (P < 0.008). (B&C) Serum T. muris E/S antigen-specific IgG1 (B) and IgG2c (C) levels were determined by ELISA. When compared to T. muris-infected CXCR5 F/F mice, the sera of T. muris-infected CXCR5 ΔDC mice contained significantly lower levels of parasite-specific IgG1 (B; P < 0.002) and significantly higher levels of IgG2c (C; P < 0.002) F I G U R E 3 The expression of Th1/Th2 cytokines in the MLN following Trichuris muris infection is altered in the specific absence of CXCR5-expressing cDC. CXCR5 F/F mice (n = 6) and CXCR5 ΔDC mice (n = 6) were orally infected with approximately 200 embryonated T. muris eggs, and at 30 dpi, the expression of cytokine-encoding genes in the MLN was compared by qRT-PCR analysis. A, Comparison of the expression of mRNA encoding the Th2 cytokines IL-4, IL-5, IL-9 and IL-13. B, Comparison of the expression of mRNA encoding the Th1 cytokine IFNγ (Ifng) and proinflammatory cytokines IL-1β (Il1b) and IL-6. C, The expression levels of Il17a (encoding IL-17) and D, Il21a, Il12b and Tnfa (encoding the IL-12p35 and IL-12p40 subunits, and TNFα, respectively) were similar in MLN from T. muris-infected CXCR5 F/F mice and CXCR5 ΔDC mice. Gene expression data show the relative expression level in infected mice compared to uninfected CXCR5 F/F control mice. Data were normalized so that the mean level in uninfected CXCR5 F/F control mice was 1.  CD103 + cDC are recruited to the colon in response to T. muris infection. 44 However, these Th1-polarizing cDC are dispensable for immunity to T. muris infection, 45 Figure S1). Further studies are required to determine whether the effects observed in the current study are due to the expression of CXCR5 in a specific cDC populations such as the cDC1, 7-10 cDC2 3-6 or other MNP 49,50 subsets.
The secretion of mucous by goblet cells plays a key role in the expulsion of intestinal helminths. 33 Expression of the Th2 cytokines IL-4 and IL-13 is considered important for the induction of goblet cell hyperplasia during helminth infections. 51 In the current study, goblet cell hyperplasia was unaffected in the large intestines of T. muris-infected CXCR5 ΔDC mice, despite the impaired expression of IL-4 in their MLN. Thus, although the Th1/Th2 balance was disturbed in T. muris-infected CXCR5 ΔDC mice, these data suggest that the remaining Th2 response was sufficient to induce the goblet cell hyperplasia. Data suggest that IL-22 also plays a central role in promoting goblet cell abundance and function during helminth infection. 34 Expression of IL-22 alone may be sufficient to enhance mucin production by the intestinal epithelium, 34 suggesting a potential mechanism through which this response may be maintained.
Although similar levels of goblet cell hyperplasia were observed in the intestines of infected CXCR5 ΔDC mice and CXCR5 F/F mice, the CXCR5 ΔDC mice were unable to clear the worm infection. This indicates that goblet cell hyperplasia on its own is insufficient to expel the worms from the intestine. Treatment of B-cell-deficient μMT mice with purified T. muris-specific IgG1 has been reported to restore resistance to infection. 52 Thus, it is plausible that the reduced serum levels of parasite-specific IgG1 may contribute to the increased susceptibility of CXCR5 ΔDC mice to T. muris infection.
Studies have suggested that cDC can capture and retain un-