|Home | About | Journals | Submit | Contact Us | Français|
Vaccinologists strive to harness immunity at mucosal sites of pathogen entry. We studied respiratory delivery of an attenuated vaccine against Blastomyces dermatitidis. We created a T cell receptor transgenic mouse responsive to vaccine yeast and found that mucosal vaccination led to poor T cell activation in the draining nodes and differentiation in the lung. Mucosal vaccination subverted lung T cell priming by inducing matrix metalloproteinase (MMP)2, which impaired the action of the chemokine CCL7 on egress of CCR2+ Ly6Chi inflammatory monocytes from the bone marrow and their recruitment to the lung. Studies in Mmp2−/− mice, or treatment with MMP inhibitor or rCCL7, restored recruitment of Ly6Chi monocytes to the lung, where they successfully primed CD4+ T-cells. Mucosal vaccination against fungi and perhaps other respiratory pathogens may require manipulation of host MMPs in order to alter chemokine signals needed to recruit Ly6Chi monocytes and prime T cells at the respiratory mucosa.
The rising rates of systemic fungal infections worldwide have stimulated interest in developing vaccines (Cutler et al., 2007). Although experimental vaccines are under study, none are in clinical trials or commercially available (Cutler et al., 2007). An understanding of the mode of action of fungal vaccines will enhance their application in human populations. Th1-cell responses mediated by IL-12 and IFNγ are believed to foster protective immunity to many pathogenic fungi (Wüthrich et al., 2011a). During the earliest stages of fungal infection, pattern recognition receptors on cells of myeloid lineage sense fungal pathogens and surfaces (Brown, 2011). Myeloid cells, particularly dendritic cells (DCs), bridge innate and adaptive immunity to pathogens. DCs have thus become a target for vaccine development strategies (Steinman, 2008).
The lung airways of all species are lined with an intraepithelial dendritic network of MHCIIhi CD11chi cells that are mostly CD11b− (Lambrecht and Hammad, 2009). The lamina propria contains MHCIIhi CD11chi cells that express CD11b and elaborate proinflammatory chemokines. CD11b+ and CD11b− subsets of DCs express high amounts of CD11c, and are viewed as conventional DCs, in contrast to a population of CD11cint plasmacytoid DCs (pDCs). Lung alveoli contain CD11chi MHCIIhi DCs, which are enriched in CD103+ subsets. Alveolar macrophages also express CD11c (not CD11b), confusing the analysis of lung DCs unless autofluorescence is used to identify macrophages (Vermaelen and Pauwels, 2004). Under inflammatory conditions, such as microbial challenge, CD11b+ monocyte-derived DCs are recruited that rapidly upregulate CD11c and retain Ly6C as a remnant of monocytic descent (Shi and Pamer, 2011).
Monocyte-derived DCs are important for priming T cell responses to microbes (Hohl et al., 2009; Serbina et al., 2008). These DCs prime Th1 responses and IFNγ producing Ag-specific CD4+ transgenic (tg) T cells in pulmonary aspergillosis (Hohl et al., 2009). In Aspergillus infection, CCR2+ Ly6Chi DCs are required to prime T cells in the lung, yet not in the spleen. We recently analyzed sequential stages during the induction of vaccine immunity to fungi after s.c. injection (Ersland et al., 2010). Monocyte-derived DCs initially take up the most vaccine yeast and traffic them to the skin-draining LNs. However, the direct priming of naive Ag-specific CD4 T cells in vivo is governed by LN-resident DCs and skin-derived DCs. In fact, other skin DCs compensate for monocyte derived DCs in Ccr2−/− mice lacking the cells.
Vaccinologists strive to harness protective immunity at mucosal sites of initial pathogen entry. We sought to immunize against fungal respiratory pathogens by delivering vaccines into the respiratory tract. Although delivery of an attenuated strain of Blastomyces dermatitidis(Bd) into the skin protects 100% of mice against a lethal pulmonary challenge, and most develop sterilizing immunity (Wüthrich et al., 2000), vaccine delivery intranasally (i.n.) failed to protect and 100% of vaccinees succumbed to infection. Herein, we studied the mechanism behind the failure in priming anti-fungal vaccine immunity at the respiratory mucosa.
We report that fungi subvert the induction of vaccine immunity at the respiratory mucosa by inducing lung matrix metalloproteinase 2 (MMP2), which suppresses the action of CCL7 and impairs recruitment and maturation of Ly6Chi inflammatory monocyte-derived DCs in the lung. Elimination of mature Ly6hi DCs at this site retarded the activation, expansion and differentiation of IFNγ+ CD4+ T cells. Conversely, administering an MMP2 inhibitor or CCL7 during vaccine delivery restored recruitment of Ly6Chi monocytes to the lung and priming of CD4+ T cells. Our results pinpoint mechanisms that underpin vaccination against fungi at the respiratory mucosa. They also highlight host and microbial strategies that must be overcome to engineer fungal and other vaccines that induce respiratory mucosal immunity. Mucosal vaccination against respiratory agents may require manipulation of host MMPs that alter chemokine signals needed to recruit Ly6Chi inflammatory monocytes and prime CD4+ T cells at the respiratory mucosa.
Subcutaneous (s.c.) injection of mice with a live attenuated Bd strain engenders 100% survival against a lethal pulmonary challenge (Wüthrich et al., 2000), but inconsistent sterilizing immunity. Since the natural route of infection is inhalation of spores, we sought to enhance the vaccine’s efficacy by delivering it into the respiratory tract. All mice vaccinated i.n. (Wüthrich et al., 2000) or intratracheally (i.t.) (data not shown) were unable to control infection and died after pulmonary challenge. This scenario contrasts with that for Histoplasma capsulatum (Hc), in which primary pulmonary infection induces protective immunity and resistance against lethal pulmonary challenge (Deepe and Seder, 1998). Since these two fungal infections require cellular immunity for resistance, we compared the priming of CD4+ T cells for each of them to uncover the reasons for failure vs. success in priming of T cells at the respiratory mucosa.
Th1 differentiation occurs fully after CD4+ T cells migrate to the lung (Rivera et al., 2006). Although i.t. vaccination with Bd induced activated CD4+ T cells (CD44+) cells in the lung, Th1 cells failed to accrue and <1% produced IFNγ (Fig. 1A). There were ≈1,000-fold less IFNγ+ CD4 T cells in the lung after i.t. vaccination with Bd compared with Hc. Hc induced a 1,000-fold increase in the number of IFNγ+ CD4 T cells in the lung upon mucosal vaccination, with nearly 14% producing IFNγ, whereas Bd induced little increase. In contrast, s.c. administration of Bd, as well as Hc, lead to marked expansion of IFNγ+ cells during a recall response after challenge. Mice given Bd s.c. had 100-fold more IFNγ+ cells than unvaccinated controls, and over 7% of CD4+ T cells produced this cytokine (Fig. S1A).
Several mechanisms could explain the small number of IFNγ++ cells and the inability of attenuated Bd to vaccinate at the respiratory mucosa. First, the vaccine may not induce proliferation of Bd specific CD4+ T cells or promote their survival. Second it may not induce differentiation of Ag-specific T cells in the draining MLN. Third, Th1 CD4+ T cells may not be recruited from MLN into the lung airways. Last, CD4 T cells may not fully differentiate or mature into Th1 cells in the lung. To distinguish among these possibilities, and interrogate T cell priming, expansion, differentiation and trafficking, we generated a TCR tg mouse specific for Bd.
The Bd 1807 TCR tg mouse was engineered (Fig. S1C–F; see supplemental experimental procedures) from a CD4+ T cell clone that confers protective immunity against lethal pulmonary challenge in mice (Wüthrich et al., 2007). Bd 1807 mice have an increased prevalence of Vα2+ CD4+ T cells in the peripheral blood, spleen and LNs vs. wild-type B6 mice (Fig. S1E). Naïve CD4+ T cells from Bd 1807 mice became activated and proliferated in response to cell wall-membrane antigen (CW/M), whereas CD4+ T cells from naïve wild-type mice did not respond to CW/M (Fig. S1F). Thus, Bd 1807 cells are specific and responsive to Bd in vitro.
Bd 1807 cells respond to Bd and cross-react with Hc after s.c. vaccination (Wüthrich et al., 2011b), they become activated in the draining LNs, differentiate into Th1 effectors on migration to the vaccine site, and exhibit memory and recall to lung upon lethal pulmonary infection. Bd 1807 cells not only show fidelity with polyclonal CD4+ T cells, but they also report the behavior of CD4+ T cells that confer protective immunity (Wüthrich et al., 2007). Thus, Bd 1807 cells enable interrogation of CD4+ T cell priming, expansion and trafficking in response to both fungi, and comparisons of these events between them.
We compared the priming of Bd 1807 tg CD4+ T cells in response to vaccination with Bd or Hc in the lung. Although Bd 1807 cells respond vigorously to both fungi after vaccination s.c. (above), when vaccine is instead given i.t., the activation, expansion, and differentiation into IFNγ effectors is sharply reduced in mice given Bd vs. Hc (Fig. 1B). The discrepancies between the two groups were most exaggerated for IFNγ+ 1807 cells in the lung. Hc-vaccinated mice had >100-fold more 1807 Th1 effectors compared to Bd. In lung-draining or mediastinal LNs (MLN), Bd-vaccinated mice also had 10- to 100-fold fewer activated cells. In contrast, after s.c. vaccine, Bd 1807 cells were activated (CD44+) in the sdLNs efficiently in response to both fungi, and similar numbers of antigen-specific IFNγ+ T1 effectors were recalled to the lung (Fig. S1B).
Thus, Bd1807 cells let us pinpoint defects in priming CD4+ T cells after delivery of Bd vaccine in the lung, which are linked to 1) impaired activation and expansion in the lung-draining MLN, and 2) profound failure to differentiate fully into Th1 IFNγ producing effectors upon migration back to lung after mucosal vaccination. These deficits likely underpin failed vaccination by Bd at the respiratory mucosa. The results contrast sharply with the robust priming, expansion and differentiation of Bd 1807 cells in response to Hc delivered via the respiratory route and to Bd vaccine delivered by the s.c. route. These differences beg the question: why does Bd vaccine fail to prime T cells at the respiratory mucosa and what mechanism accounts for the sharp disparity in vaccine priming in the lung by these two closely related fungi?
We analyzed the myeloid APCs recruited to the lung after delivery of vaccine yeast into the respiratory tract. We looked at CD103+ DCs, CD11b+ DCs, including inflammatory monocyte-derived Ly6Chi DCs, alveolar macrophages, and neutrophils. We saw sharp differences in the influx of APCs and uptake of yeast for the two fungi. CD11b+ DCs were among the most prominent APC recruited to the lungs by both fungi (Figs. 2A and S2C–F). The sharpest difference between the two fungi was in the influx of Ly6Chi CD11bhi cells. Ly6Chi monocytes accounted for ≈10% of the lung cells in response to Hc vs. <1% for Bd (Fig. 2B). By staining yeast with PKH26, we honed in on lung APC that harbored yeast (Figs. 2C&D; and S2D–F). CD103+DC contained few yeast (Fig. 3C). In contrast, for both fungi, most yeast resided with CD11b+ DCs (Figs. 2C and S2D), and also in macrophages and neutrophils (Fig. S2E–F). Strikingly, there were >10-fold more Ly6Chi CD11bhi cells in the lung harboring Hc vs. Bd (Fig. 2C&D).
The same disparity between these fungi was seen in the draining MLN (Fig. 2E–H). Ly6Chi CD11bhi cells were nearly absent in the MLN of Bd-vaccinated mice (0.045%), whereas the proportion of these cells in the MLN of Hc-vaccinated mice was 30-fold higher (1.4%) (Fig. 2E&F). Most of the DCs that harbored yeast in MLN were CD11b+ DC, although some CD103+ cells also harbored yeast for both fungi (Fig. 2G). Still, the biggest difference between the two fungi was again in Ly6Chi CD11b+ cells; the proportion of these cells harboring Hc (7.2%) was ≈12-fold higher than that for Bd (0.59%) (Fig. 2G). Consequently, in the MLN, there was >1000-fold more Ly6Chi inflammatory monocytes with Hc (≥103 cells) than Bd (≤10 cells) (Fig. 2H). Thus, the numbers of Ly6Chi monocytes that are recruited to the lung and migrate into the draining MLN, and the numbers of these DCs harboring yeast, are sharply reduced in mice that received Bd compared to Hc.
The blunted entry of Ly6Chi monocytes into the lungs of Bd-vaccinated mice could be due to either a failure to induce the recruitment of these cells, or alternatively, an active process of blocking their recruitment. To distinguish between the two possibilities, we performed a mixing experiment in which we added Bd vaccine yeast to the inoculum of Hc given i.t. The addition of Bd curtailed the recruitment of Ly6Chi monocytes into the lung by Hc (Fig. 3A). In the mixed infection, the distribution and numbers of DCs showed a paucity of Ly6Chi CD11b+ DCs (0.26%) that was similar to mice vaccinated with Bd (0.19%) and much lower than mice that received Hc (13.7%) (Fig. 3A); the numbers of total Ly6Chi DCs in the lung followed similar trends with almost identical patterns for Bd and mixed infections, each much lower than for Hc vaccination (Fig. 3B). To see if mixed infection perturbed the numbers of APC harboring yeast, we stained the yeast with PKH26. The total number of yeast-loaded APC in the lung was increased in the mixed infection (Fig. S3B), although the numbers of Ly6Chi (and Ly6Chi CD11c+) cells that harbored yeast was greatly reduced in the mixed-infection and Bd-alone groups, as compared to the H. capasulatum group (Fig. 3B). The patterns in the MLN mirrored those in the lung (Fig. 3C&D). Thus, Bd interferes with either recruitment or entry and migration of Ly6Chi monocytes in the lung and MLN, and can do so even in the case of a strong inducing stimulus like respiratory exposure to Hc.
Ly6Chi monocytes that enter the lung in response to infection mature upon transition to DCs, involving down-regulation of Ly6C and up-regulation of CD11c and MHC class II (Osterholzer et al., 2009). We explored this process in the three groups: either yeast alone or mixed together. For Ly6Chi monocytes that entered the lung in response to mucosal vaccination with Bd, both the Ly6C expression and the CD11c expression were much lower than that compared to the Hc group, and the mixed infection yielded a pattern similar to that for Bd alone (Fig. 3E). Thus, in the setting of Bd vaccine, low Ly6C expression was not due to the maturation to CD11c+ DC; the yeast instead blunted the maturation of CD11cDCs.
Ly6Chi inflammatory monocytes trafficking to organs may require CCR2 expression (Serbina et al., 2008). Over 90% of the Ly6Chi monocytes that migrate into the MLNs in response to Hc display CCR2 (Fig. 3F). Among the CD11c+ Ly6C+ population in the MLN in this setting, 90% express CCR2. Importantly, Ly6Chi CD11c+ CCR2+ cells numbered ≈5-fold more after the mucosal delivery of Hc, compared to Bd or mixed infection (Fig. 3G).
We investigated mechanisms behind the impaired influx of Ly6C hi inflammatory monocytes upon respiratory vaccination with Bd. CCR2+ inflammatory monocytes emanate from the bone marrow in response to chemokine signals, so we first analyzed if the defect in cell recruitment is due to failed egress of cells from the bone marrow or failed entry into the lung. In wild-type mice, the percentage of Ly6Chi monocytes in the marrow was ≈2-fold higher in animals that got Bd or mixed infection, compared to Hc (Fig. 4A). This increase could be due to trapping of cells in the marrow or increased production. To distinguish these possibilities, we studied Ccr2−/− mice where these cells cannot exit the marrow (Serbina and Pamer, 2006). Production is increased in the Hc group relative to the other groups (Fig. S4A), suggesting that over-production does not explain the accumulation for Bd relative to Hc in wild type mice. We did not see a difference between groups in the number of circulating monocytes during 2–5 days after vaccination (data not shown). Thus, mucosal delivery of Bd leads to failure of egress and trapping of Ly6Chi monocytes in the marrow rather than defective monocyte extravasation into the lung.
The major signals that induce egress of CCR2+ monocytes from the bone marrow are CCL2 and CCL7 (Jia et al., 2008; Tsou et al., 2007). To see if Bd affected these chemokines, we analyzed levels in the lungs and serum of mice from the three groups. The most striking difference was elevated serum levels of CCL7 in those that received Bd i.t. (Fig. 4B). We tested the activity of these sera in promoting migration of Ly6Chi monocytes in vitro. Sera from mice that got Hc promoted the migration of Ly6Chi monocytes, whereas sera from the Bd group did not (Fig. 4C). Thus, the elevated levels of CCL7 in Bd sera failed to induce chemotaxis.
Chemokines can be inactivated by serine proteases of mammalian or microbial origin. To test if inactive CCL7 might explain the defects in bone marrow egress of Bd vaccinated mice, we vaccinated the mice together with recombinant CCL7. Recombinant CCL7 given to Bd-vaccinated mice enhanced the egress of Ly6Chi monocytes out of the marrow and induced their recruitment into the lungs (Fig. 4D). Thus, the elevated endogenous levels of CCL7 in serum were not functional in vivo.
Inactive chemokines can desensitize their receptor (Ali et al., 2005). We studied the migration of bone marrow monocytes in vitro in response to both CCL7 and CCL2. Ly6Chi monocytes from naïve mice showed a ≈3-fold increase in migration toward these chemokines compared to medium alone, while cells from Bd-vaccinated mice showed significantly less migration toward the ligands (Fig. S4B). We further analyzed receptor desensitization by measuring Ca2+ flux. The flux of bone marrow Ly6Chi monocytes in response to CCL7 as well as CCL2 is curtailed in cells from Bd-vaccinated mice vs. naïve mice (Fig. 4E; data not shown). Thus, Ly6Chi monocytes from Bd-vaccinated mice showed reduced sensitivity to their ligands, promoting their trapping in the marrow and poor recruitment to the lung.
Chemokines can be inactivated by mammalian MMPs. The gelatinase MMP2 acts on CCL7 converting it into an inactive or antagonistic form (McQuibban et al., 2000; McQuibban et al., 2002). We investigated lung MMP2 in response to vaccination at the respiratory mucosa. The level of active MMP2 in bronchoalveolar lavage (BAL) fluid was several-fold higher in mice vaccinated with Bd vs. Hc(Fig. 5A). To see if the MMP2 levels affected chemokine-mediated egress of Ly6Chi monocytes from the marrow to the lung, we inhibited MMPs during vaccine administration. The broad MMP inhibitor, GM6001, and the MMP2-selective inhibitor promoted egress of Ly6Chi monocytes from the bone marrow (Fig. S5A) and increased their recruitment to the lung in mice vaccinated i.t. with Bd (Fig. 5B). These inhibitors did not affect the growth of the fungus in lung (data not shown). Vaccination of Mmp2−/− mice also augmented the release of Ly6Chi monocytes from the marrow (Fig. S5B) and influx of these cells into the lung compared to vaccinated wild-type mice (Fig. 5C). Thus, induction of lung MMP2 and inactive CCL7 underpin the trapping of Ly6Chi monocytes in the bone marrow and their failure to traffic to the lung in response to fungal vaccine administration at the respiratory mucosa.
We examined how impaired influx of Ly6C+ CD11b+ CD11c+CCR2 + cells upon vaccination affects priming of protective CD4+ T cells in the lung, including Bd 1807 cells. We used Ccr2−/− mice to establish the contribution of Ly6Chi monocytes (Fig. 6A). Expansion of Bd 1807 cells in the MLN was markedly impaired in wild-type mice that received Bd or mixed infection i.t., compared to Hc (Fig 6B). In the Ccr2−/− mice, 1807 cells again failed to expand in response to Bd. Slightly higher numbers of 1807 cells expanded in response to Hc in Ccr2−/− vs. wild-type mice, but the difference was insignificant. However, the differentiation of IFNγ producing Th1 1807 cells in response to Hc was sharply impaired in Ccr2−/− mice in the MLN (Fig. 6C&D), and especially after these T cells exited the nodes and migrated back into the lung (Fig. 6F&G). In addition to impairing the differentiation of 1807 into Th1 cells, the loss of CCR2 changed the ratio of cytokine producing 1807 cells in favor of T13 and T17 cells in response to Hc (Fig. 6G). Thus, the loss of CCR2+ Ly6Chi CD11b+ DCs, due to Bd or CCR2 deletion, curtails activation, proliferation and Th1 differentiation of 1807 cells in the MLN and lung.
We tested whether resupply of Ly6Chi CD11b+ CD11c+ DCs into the lungs restored priming of Bd 1807 cells. We first established conditions that lead to optimal recruitment after i.t. delivery of Hc. These cells peaked by day 4 in the lung and day 7 in the MLNs (Fig. S6A&B). We collected Ly6Chi monocytes from mice 7 days after exposure to Hc. We enriched the cells by negative selection (Fig. S6C), after which Ly6Chi monocytes constituted ≈15% of the population (10–15-fold enrichment) and Ly6C+ cells made up ≈85%. About 1% to 2% of the Ly6Chi monocytes harbored PKH26+ yeast (data not shown). We transferred the Ly6C-enriched cells i.t. into Ccr2−/− or wild-type mice and also delivered Bd or Hc by this route (Fig S6C). We verified that congenic transferred DCs persisted in the lungs over several days (Fig. S6D). We analyzed the priming and differentiation of separately transferred Bd 1807 cells (and endogenous CD4+T cells) in lungs 9 days after DC transfer.
In Ccr2−/− mice, transfer of Ly6Chi monocytes enhanced the number of IFNγ+ 1807 cells after mucosal vaccination (Fig. 7A–C): they rose ≈100-fold in mice that received Bd (note Bd only vs. Bd + DCs in Fig. 7B), and ≈10-fold in mice that received Hc (note Hc only vs. Hc + DCs). Similar trends were observed among the polyclonal CD4+ T cells. Thus, the Ly6Chi monocytes transferred into Ccr2−/− recipients were functional and re-established activation, differentiation, and recruitment into the lungs of Th1 1807 cells.
Upon transfer into wild-type mice (Fig. 7D–F), Ly6Chi monocytes exerted a similar augmentation of 1807 cell function in the lungs. Transfer of Ly6Chi monocytes into unvaccinated mice enhanced the number of IFNγ+ 1807 cells, when compared to naïve mice. Transfer of these cells into mice that had been vaccinated i.t. with Bd had a more pronounced effect, increasing the number of IFNγ producing 1807 cells in the lung by nearly 100-fold (Fig. 7E; note Bd only vs. Bd + DCs), and the percentage of these cells by nearly 10-fold (Fig. 7D; note Bd only vs. Bd + DCs). Comparable trends were observed in the endogenous CD4+ T cell pool (Figs. 7D and F). Thus, impaired recruitment and maturation of Ly6Chi monocytes led to failed priming of CD4 T cells and skewing of their differentiation away from Th1 cells upon mucosal vaccination with Bd. These events were remediated by the adoptive transfer of Ly6Chi monocytes.
CD4+ T cells that produce IFNγ are pivotal in immunity to fungi (Cutler et al., 2007). Here, we found that respiratory mucosal vaccination with Bd did not prime IFNγ producing endogenous CD4+ T cells. We used a novel TCR tg mouse to track Ag-specific T cells to decipher the mechanism. The host response to attenuated fungus, involving exuberant MMP and restrained inflammation, paradoxically undermined downstream priming of T cells. This finding offers a cautionary note to vaccinologists harnessing vaccine immunity at the respiratory mucosa.
Subcutaneous delivery of attenuated Bd recruits Ly6Chi DCs to the injection site, which ferry yeast into the sdLNs (Ersland et al., 2010). Respiratory mucosal delivery of the vaccine yeast instead failed to recruit the cells. There were low numbers of these cells in the lung and MLN, and few harbored yeast. Bd blocked recruitment and maturation of Ly6Chi DCs in the lung, rather than simply failing to induce their influx and maturation. The impact of Bd on Ly6Chi monocytes was selective. Uptake of these yeast into APCs and their trafficking into MLNs was not generally impaired, since the number of APCs harboring Bd in lung and MLN was comparable to that in mice exposed to Hc alone or in a mixed infection. Thus, Bd led to a redistribution of yeast from Ly6Chi monocytes, into other lung APCs.
Ly6Chi monocytes regulate the early host response to Aspergillus lung infection by taking up conidia and trafficking them into the draining MLN to prime CD4+ T cells (Hohl et al., 2009). Ly6Chi monocytes are dispensable in priming CD4+ T cells in the spleen of mice that are systemically infected i.v. with Aspergillus. Although Ly6Chi monocytes carry Bd yeast into the sdLNs after s.c. vaccination (Ersland et al., 2010), in Ccr2−/− mice lacking Ly6Chi DCs, other migratory skin and resident DCs compensate by delivering yeast into the draining nodes to prime T cells. In contrast, during mucosal vaccine delivery here, the modulation of Ly6Chi monocyte influx – either due to Bd alone or in mixed infection, or in Ccr2−/− mice – sharply reduced the numbers of vaccine yeast delivered to the MLN and impaired priming of T cells. Thus, Ly6Chi monocytes are indispensible in fungal vaccine delivery and antigen priming in the lung. Cryptococcus neoformans also induces the recruitment into lung of Ly6Chi monocytes (Osterholzer et al., 2009), which mature into CD11b+ CD11c+ DCs, down-regulate Ly6C, and promote resistance. In primary pulmonary histoplasmosis, Ccr2−/− mice have altered leukocyte influx in their lungs, and succumb to infection due to overproduction of IL-4 and alternative activation of pulmonary macrophages (Szymczak and Deepe, 2009).
Several chemokines can bind CCR2. CCL2 and CCL7 are the main CCR2 ligands in inflammatory settings (Jia et al., 2008; Tsou et al., 2007), and help mobilize monocytes from the bone marrow. Vaccination with Bd led to trapping of these cells in the marrow, and was also associated with high levels of CCL7 in serum. The serum CCL7 in these mice was inactive, failing to promote CCR2-dependent migration of Ly6chi monocytes in chemotactic assays. Further, treatment of these vaccinated mice with rCCL7 repaired the deficit, promoting egress of CCR2+ cells from the marrow and entry into the lung. We propose that CCL7 was rendered inactive by MMP2 since this product was elevated in Bd vaccinated mice and its inhibition relieved marrow trapping and promoted Ly6Chi cell recruitment to the lung. Removal of N-terminus residues of CCL7 by MMP2 impairs its activity, yet CCL7 can still bind its receptor and compete with CCL2 (McQuibban et al., 2000; McQuibban et al., 2002). Host MMP2 may have modified CCL7 impairing its action, perhaps also antagonizing CCL2. Inflammatory monocytes from the marrow of Bd vaccinated mice showed reduced sensitivity and response to CCL2. Although these defects could be partially corrected with exogenous rCCL7, we did not measure serum CCL7 levels after treatment, and cannot exclude a pharmacological effect. In hepatitis C, serum CXCL10, which recruits lymphocytes to the liver by binding surface CXCR3, is paradoxically elevated in the patients who fail therapy (Casrouge et al., 2011). Elevated levels of the protease dipeptidyl peptidase IV (DPP4) cleaves CXCL10, converting it from an agonist into an antagonist of CXCR3. While MMP2 likely acts similarly herein, we cannot exclude a role for other MMPs including of fungal origin.
Impaired recruitment and maturation of Ly6Chi monocytes by Bd blunted the activation, expansion and differentiation of IFNγ producing T cells in the draining MLN, and profoundly altered differentiation of T cells on migration to the lungs. Bd 1807 cells showed deficits in these functions after respiratory mucosal vaccination with Bd (but not after s.c. vaccination). Because Bd 1807 cells recognize a shared antigen in Bd and Hc, we could show that priming and Th1 differentiation of these T cells by Hc is impaired not only in Ccr2−/− mice, but also in wild-type mice upon mucosal delivery of mixed Bd and Hc. Thus, yeast modulation of Ly6Chi monocytes mediates failed T cell priming after mucosal vaccination with Bd. We substantiated the role of Ly6Chi monocytes by adoptive transfer, which restored priming, expansion and Th1 differentiation of Bd 1807 cells. Consistent with our findings, prior reports showed that inflammatory monocyte-derived DCs stimulate Th1 immune responses. In Ccr2−/− mice, Th1 CD4+ T cell responses are impaired due to reduced monocyte recruitment to inflamed LNs and diminished production of IL-12 at the time of CD4+ T cell priming (Nakano et al., 2009; Peters et al., 2000; Peters et al., 2001). Similarly, in pulmonary cryptococcosis, Ccr2−/− mice redirect CD4+ T cell differentiation from Th1 to Th2 cells (Blease et al., 2000; Traynor et al., 2000).
Impaired recruitment and maturation of Ly6Chi monocytes skewed the differentiation of Bd 1807 cells toward IL-13 and IL-17 producing cells. This was especially evident in the 1807 cells that had fully differentiated after migrating back to the lung in Ccr2−/− mice that received Hc. This was also true for 1807 cells in mice that received mixed infection. Our findings suggest that other APCs in the lung of wild-type mice that received Bd or mixed infection, or in the Ccr2−/− that received Hc, promote the differentiation of CD4+ T cells into IL-13 and IL-17 producing cells. Symczak et al. (Szymczak and Deepe, 2009) found that lung IFNγ levels were unaltered in Hc infected Ccr2−/− mice, but elevated IL-4 levels instead led to their death. IL-17 levels were not reported in that study, but it was not possible to track Ag-specific T cells to assess skewed differentiation in the LNs or lung.
Priming Ag-specific immunity at mucosal sites of pathogen entry is an active area in vaccine development. Our study highlights pitfalls in applying this strategy broadly across pathogen kingdoms. There are 4 commercially available, mucosal vaccines: polio, rotavirus, Salmonella typhi, and influenza – each, a live attenuated vaccine. HIV vaccine development is focused on strategies for a mucosal vaccine. Intranasal vaccine delivery of respiratory viruses is thought to prime immunity in nasal mucosa-associated lympho-reticular tissue, and engage not only s-IGA, but also T cell immunity, particularly CTLs. Persistent antigen in the lung after influenza A infection fosters antigen uptake by specialized respiratory DC that prime memory T cells (Kim et al., 2010). Although CCR2+ cells govern resistance to M. tuberculosis (Peters et al., 2001), enhanced recruitment of inflammatory monocytes to the lungs of mice infected with M. tuberculosis using intranasal Poly-IC unexpectedly enhanced pathogen growth in this permissive population and lung tissue injury despite unaltered production of IFNγ (Antonelli et al., 2010).
These studies illustrate differences among pathogen kingdoms in vaccine or therapeutic strategies targeted to respiratory sites of pathogen entry. Our work sheds new light on this variability by highlighting the role of chemokines and their cellular targets in inducing mucosal immunity, and by unveiling the paradoxical effect of the host in undermining immunity to a mucosal vaccine. To optimize the efficacy of mucosal vaccination, suitable vaccine adjuvants may thus need to target host MMP responses that counter adaptive immunity at the lung mucosa. Targeted recruitment of leukocytes to the lung on vaccination may augment vaccine strategies.
The generation and characterization of Bd 1807 mice is described in detail in the accompanying supplemental experimental procedures.
To measure monocyte migration in response to serum or chemokine bone marrow cells were enriched for Ly6ChighCD11b+ cells and placed in the upper chamber of transwell plates and allowed to migrate through the trans-membrane. To measure Ca2+-flux, monocytes were loaded with Indo-1, stimulated with chemokine and the flux recorded over time by flow cytometry. For a detailed description, see Supplemental Experimental Procedures.
This work was supported by grants from the USPHS to BK and MW. We thank Drs. Amanda Starr and Christopher Overall from the University of British Columbia, Canada, for providing advice on the MMP2 assays; and Robert Gordon from the department of Pediatrics at the University of Wisconsin for assistance with illustrations.
Supplemental information includes supplemental figures, legends and experimental procedures (including adoptive transfer of Bd 1807 cells, vaccination and experimental infection, PKH26 staining of yeast, lung and mediastinal LN preparation, flow cytometry, adoptive transfer of inflammatory monocytes, generation of bone marrow dendritic cells, zymography, chemokine analysis and administration of recombinant CCL7 into mice, chemotaxis assay, calcium flux measurements, in vivo treatment with MMP2 inhibitors and statistics).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.