Meeting report by Cecile Piot who was awarded a travel grant by the CFCD
Cecile Piot is a Post Doctoral Fellow working at Caetano Reis e Sousa laboratory at the Francis Crick Institute in London, United Kingdom
The 2023 CFCD meeting took place in Bordeaux, France, over the course of two days. The conference covered a wide range of topics of great interest for the field of DC and macrophage biology. Notably, the talks centered on the role of these cells in regulating anti-tumour immunity and in tissue homeostasis, multi-omics approaches to study DCs and macrophages in various contexts and the interplay between metabolism and the functions of these cells. Below are summarized selected findings that were presented during the meeting.
The concept that cDCs are important for anti-tumour immunity both in tumour-draining lymph nodes (tdLNs) and in tumours was first introduced by Stefani Spranger (Massachusetts Institute of Technology, Cambridge, USA) in the “DCs and macrophages in anti-tumor immunity” session. During her talk, she showed that certain tumours, such as some non-small cell lung cancer (NSCLC) tumours, are highly infiltrated by CD8+ T cells but fail to respond to immune checkpoint blockades (ICBs). In these tumours, her team showed that CD8+ T cells in tumours were “dysfunctional” (clonally expanded but expressing low levels of effector and exhausted molecules), which was due to aberrant priming of CD8+ T cells by cDC1s in tdLNs. Mechanistically, cDC1s were suppressed by Th1-like Tregs in tdLNs which impaired cDC1’s abilities to induce potent cytotoxic CD8+ T cell responses. This was highlighted by comparing the immune response mounted in flank versus orthotopic lung tumours, in which tumours are controlled or outgrow, respectively. Interestingly, in both scenarios cDC1s were able to take up tumour-derived antigens and migrate to the tdLNs, but Th1-like Tregs within tdLNs of lung tumours induced a reduction in IL-12 and co-stimulatory molecules in cDC1s. These are important findings as they might explain why some patients that have clonally expanded CD8+ T cells in tumours do not respond to ICBs. In addition, she presented another important concept revolving around the fact that a state of cDC2s in tumours, called the ISG-DC, can interact with CD8+ T cells in tumours by acquiring intact MHCI-I:peptide from tumour cells (a process called cross-dressing). “ISG-DCs” express high levels of interferon-induced genes (ISGs), and this state was induced by type 1 interferons. Although cDC1 is considered the main population inducing CD8+ T cells in tumours, ISG-DCs could compensate for the absence of cDC1s in Batf3-/- mice. Interestingly, exogenous addition of IFN-β could augment DCs’ cross-dressing and control regressor tumours, which might represent an axis by which the CD8+ T cell response could be boosted in patients.
In the same session, Mikael Pittet (School of Medicine, Geneva, Switzerland) presented some work uncovering how the ratio of two states of macrophages, associated with CXCL9 or SPP1 expression (that he defined as the “CS” ratio) can be used as marker to predict tumour progression in cancer patients. First, he showed that the gene expression profile of various immune and stromal cells correlated with the patient’s prognosis (first in head and neck cancer patients but also in other cancer type) using univariate analyses. However, when using multivariate analyses, only the state of macrophages and of mast cells in tumours retained independent effect to predict tumour progression, suggesting that the other cell types do correlate with macrophage states and tumour progression but are redundant to predict prognosis. At the two ends of the macrophage states spectrum, CXCL9 and SPP1 macrophages were shown to correlate with the response positively and negatively, respectively. Pathways associated with a CS-high ratio (dominated by CXCL9 macrophages) included interferon-gamma and alpha response, while a CS-low ratio was associated with hypoxia. Spatially, CXCL9 and SPP1 macrophages were in different “hubs” of the tumour, and interestingly, cells surrounding CXCL9 macrophages were also more likely expressing CXCL9, and similar for the SPP1. This work suggests that the CS ratio can be used to stratify patients, and future work will reveal how these states dictate the response, and how this further shape the states of macrophages and of other cells in the tumour.
In the “Omics” session, Shalin Naik (The Walter and Eliza Hall Institute of Medical Research, Australia) presented work and techniques uncovering clonal heterogeneity in immune progenitors. Using barcoding of progenitors and by analysing the resulting progeny in tissues, one stricking finding was the fact that only a few lympho-myeloid primed progenitor (LMPPs) clones can indeed give rise to multiple immune cell types in tissues (cDCs, B cells, neutrophils, etc.), but that an impressively high number of LMPP clones give rise to a single cell type (for example only cDC, only B cells, etc.). Therefore, although LMPPs are multipotent as a whole (= able to give rise to multiple cell types), at the single cell level, single clones of LMPP are already committed towards specific lineages. By analysing the progeny of “sister” progenitors, he showed that their output in terms of cell types generated was strikingly similar between sister cells, concluding that “progenitors know what they are going to make and when”. To understand the components that dictate the clonality of one progenitor, a challenge to overcome is that both the progenitors and the resulting progeny (or fate) need to be analysed. To overcome this challenge, his team developed “SIS-seq”, in which sister progenitors are used either for scRNAseq analysis or for fate analysis in culture. Using this method, his team identified Bcor as a gene important for the clonal fate of progenitors to the DC lineage. This method will likely uncover other genes and pathways important for the development of other lineages.
An important aspect to understand how cDCs and macrophages participate in immunity is to characterize how these cells interact with other cells. In the third session of the meeting, Giulia Pasqual (Università degli Studi di Padova, Padova, Italy) presented a mouse model that allows to track cell-cell interactions using a system of transfer label in closely interacting cells, where “donor” cells are engineered to express the transpeptidase sortase A (SrtA) that can transfer a peptide substrate onto a pentaglycine (G5) on interacting cells (acceptor cells). This “Universal LIPSTIC” mouse builds on the previously described LIPSTIC system (Pasqual et al., 2018) but differs in the fact that SrtA is expressed in selected “donor” cells upon Cre recombination (and therefore bypass the need of making SrtA fused to specific ligands such as CD40L in the original LIPSTIC paper) and that all cells are “receptor” cells (through expression of G5-Th1.1 expressed in all acceptors cells). For example, by crossing these mice with Clec9a-Cre mice immunized with OVA in alum, OT-II transferred could be labelled by cDC1s (the donors), and the labelling was abolished when giving MHC-II neutralizing antibody as expected. In addition, using Foxp3-CreERT2 mice, she showed that Tregs interact with migratory cDCs in LNs and that this interaction was only partially reduced upon ablation of MHC-II interaction, suggesting that Tregs might also interact with these cells in an MHC-II-independent manner. scRNAseq analysis of LIPSITC-labelled cDCs by T cells in LNs showed that migratory cDCs are preferentially labelled in the LNs, and in particular 3 clusters of cDCs expressing genes reminiscent of cross-presentation or of cDC activation that were completely LIPSIC-labelled. This technology will be of high interest to study interacting cells, such as DCs with T cells, across various immune contexts.
In the 4th session about “tissue-specific functions of DCs and macrophages”, Ruth Franklin (Harvard Stem Cell Institute, Boston, USA) presented work revolving around the balance between host defence and tissue repair in the context of flu infection, where the immune response needs to balance between eradication of the infection while avoiding immunopathology. The talk was focused on the IL-6 family cytokine Oncostatin M (OSM) that is expressed by macrophages but whose role is unknown. Strikingly, OSM-deficient mice succumbed from flu infection but not because of absence of viral control. She showed that the drastic loss of body weight in OSM-deficient mice was associated with higher levels of type 1 interferons and ISGs expression. Strikingly, OSM-deficient mice also succumbed to Poly IC, phenocopying flu challenge but in the absence of infection. Blocking IFNA rescued the mice, suggesting that this cytokine induced immunopathology in these contexts. Altogether, a model emerged where macrophages produce OSM to limit type 1 interferons-driven immunopathology in the context of flu infection, and in potentially other systems.
In his talk about “Biology of DC’s and macrophages in the lung”, Bart Lambrecht (VIB Center for Inflammation Research, Ghent, Belgium) presented work on cDC2s in the context of Th2 immunity, in particular in the model of papain which enzymatic activities mimics those of other allergens such as house dust mites. To understand which antigen presenting cells (APCs) induce T cells in this model, Bart presented work where his group imaged papain antigens-specific T cells (transgenic PAPER T cells), that co-localized with Sirpa+ cells in LNs (including cDC2s and monocyte-derived cells) and potentially representing the APCs that prime PAPER T cells. Although monocytes are often mistakenly considered as not able to migrate to LNs because of their lack of CCR7 expression, Bart Lambrecht presented data showing that monocytes do migrate to LNs in a CCR5-dependent manner, following CCL5 gradient produced by CCR7+ cDC1s as they migrate to LNs. However, he also showed that although they migrate to the LNs, monocytes actually do not present papain peptide to PAPER CD4+ T cells ex vivo, but that cDCs do. At this point of the talk, Bart Lambrecht provided a word of caution when comparing the ability of different APCs to present and activate T cells in vitro: although the majority of APCs might be able to activate T cells when adding peptides to the culture, it does not mean that this happens in vivo. APCs should be retrieved and analysed ex vivo, which allows to consider the activation and antigen processing that the APC has experienced within tissue and in the right immune context. Next, looking at the immune context polarizing the Th2 response in the context of papain, an important point was that the source of IL-4 did not come from cDCs but from the T cells themselves, and that papain can degrade IL-12 made by cDCs. This work sheds light on the processes that lead to Th2 immunity and allergy, and on how cDCs play a role in this context. Last, Bart Lambrecht discussed an interesting connection between respiratory infection (RSV) and the development of allergy in neonates: since RSV can induce expression FcR on APCs, these cells might be able to take up IgGA-bound allergens present in the breast milk of allergic mothers and instigate a Th2 response against the allergen.
Pasqual, G., Chudnovskiy, A., Tas, J. et al. Monitoring T cell–dendritic cell interactions in vivo by intercellular enzymatic labelling. Nature 553, 496–500 (2018). https://doi.org/10.1038/nature25442
Cecile Piot is a Post Doctoral Fellow working at Caetano Reis e Sousa laboratory at the Francis Crick Institute in London, United Kingdom
The 2023 CFCD meeting took place in Bordeaux, France, over the course of two days. The conference covered a wide range of topics of great interest for the field of DC and macrophage biology. Notably, the talks centered on the role of these cells in regulating anti-tumour immunity and in tissue homeostasis, multi-omics approaches to study DCs and macrophages in various contexts and the interplay between metabolism and the functions of these cells. Below are summarized selected findings that were presented during the meeting.
The concept that cDCs are important for anti-tumour immunity both in tumour-draining lymph nodes (tdLNs) and in tumours was first introduced by Stefani Spranger (Massachusetts Institute of Technology, Cambridge, USA) in the “DCs and macrophages in anti-tumor immunity” session. During her talk, she showed that certain tumours, such as some non-small cell lung cancer (NSCLC) tumours, are highly infiltrated by CD8+ T cells but fail to respond to immune checkpoint blockades (ICBs). In these tumours, her team showed that CD8+ T cells in tumours were “dysfunctional” (clonally expanded but expressing low levels of effector and exhausted molecules), which was due to aberrant priming of CD8+ T cells by cDC1s in tdLNs. Mechanistically, cDC1s were suppressed by Th1-like Tregs in tdLNs which impaired cDC1’s abilities to induce potent cytotoxic CD8+ T cell responses. This was highlighted by comparing the immune response mounted in flank versus orthotopic lung tumours, in which tumours are controlled or outgrow, respectively. Interestingly, in both scenarios cDC1s were able to take up tumour-derived antigens and migrate to the tdLNs, but Th1-like Tregs within tdLNs of lung tumours induced a reduction in IL-12 and co-stimulatory molecules in cDC1s. These are important findings as they might explain why some patients that have clonally expanded CD8+ T cells in tumours do not respond to ICBs. In addition, she presented another important concept revolving around the fact that a state of cDC2s in tumours, called the ISG-DC, can interact with CD8+ T cells in tumours by acquiring intact MHCI-I:peptide from tumour cells (a process called cross-dressing). “ISG-DCs” express high levels of interferon-induced genes (ISGs), and this state was induced by type 1 interferons. Although cDC1 is considered the main population inducing CD8+ T cells in tumours, ISG-DCs could compensate for the absence of cDC1s in Batf3-/- mice. Interestingly, exogenous addition of IFN-β could augment DCs’ cross-dressing and control regressor tumours, which might represent an axis by which the CD8+ T cell response could be boosted in patients.
In the same session, Mikael Pittet (School of Medicine, Geneva, Switzerland) presented some work uncovering how the ratio of two states of macrophages, associated with CXCL9 or SPP1 expression (that he defined as the “CS” ratio) can be used as marker to predict tumour progression in cancer patients. First, he showed that the gene expression profile of various immune and stromal cells correlated with the patient’s prognosis (first in head and neck cancer patients but also in other cancer type) using univariate analyses. However, when using multivariate analyses, only the state of macrophages and of mast cells in tumours retained independent effect to predict tumour progression, suggesting that the other cell types do correlate with macrophage states and tumour progression but are redundant to predict prognosis. At the two ends of the macrophage states spectrum, CXCL9 and SPP1 macrophages were shown to correlate with the response positively and negatively, respectively. Pathways associated with a CS-high ratio (dominated by CXCL9 macrophages) included interferon-gamma and alpha response, while a CS-low ratio was associated with hypoxia. Spatially, CXCL9 and SPP1 macrophages were in different “hubs” of the tumour, and interestingly, cells surrounding CXCL9 macrophages were also more likely expressing CXCL9, and similar for the SPP1. This work suggests that the CS ratio can be used to stratify patients, and future work will reveal how these states dictate the response, and how this further shape the states of macrophages and of other cells in the tumour.
In the “Omics” session, Shalin Naik (The Walter and Eliza Hall Institute of Medical Research, Australia) presented work and techniques uncovering clonal heterogeneity in immune progenitors. Using barcoding of progenitors and by analysing the resulting progeny in tissues, one stricking finding was the fact that only a few lympho-myeloid primed progenitor (LMPPs) clones can indeed give rise to multiple immune cell types in tissues (cDCs, B cells, neutrophils, etc.), but that an impressively high number of LMPP clones give rise to a single cell type (for example only cDC, only B cells, etc.). Therefore, although LMPPs are multipotent as a whole (= able to give rise to multiple cell types), at the single cell level, single clones of LMPP are already committed towards specific lineages. By analysing the progeny of “sister” progenitors, he showed that their output in terms of cell types generated was strikingly similar between sister cells, concluding that “progenitors know what they are going to make and when”. To understand the components that dictate the clonality of one progenitor, a challenge to overcome is that both the progenitors and the resulting progeny (or fate) need to be analysed. To overcome this challenge, his team developed “SIS-seq”, in which sister progenitors are used either for scRNAseq analysis or for fate analysis in culture. Using this method, his team identified Bcor as a gene important for the clonal fate of progenitors to the DC lineage. This method will likely uncover other genes and pathways important for the development of other lineages.
An important aspect to understand how cDCs and macrophages participate in immunity is to characterize how these cells interact with other cells. In the third session of the meeting, Giulia Pasqual (Università degli Studi di Padova, Padova, Italy) presented a mouse model that allows to track cell-cell interactions using a system of transfer label in closely interacting cells, where “donor” cells are engineered to express the transpeptidase sortase A (SrtA) that can transfer a peptide substrate onto a pentaglycine (G5) on interacting cells (acceptor cells). This “Universal LIPSTIC” mouse builds on the previously described LIPSTIC system (Pasqual et al., 2018) but differs in the fact that SrtA is expressed in selected “donor” cells upon Cre recombination (and therefore bypass the need of making SrtA fused to specific ligands such as CD40L in the original LIPSTIC paper) and that all cells are “receptor” cells (through expression of G5-Th1.1 expressed in all acceptors cells). For example, by crossing these mice with Clec9a-Cre mice immunized with OVA in alum, OT-II transferred could be labelled by cDC1s (the donors), and the labelling was abolished when giving MHC-II neutralizing antibody as expected. In addition, using Foxp3-CreERT2 mice, she showed that Tregs interact with migratory cDCs in LNs and that this interaction was only partially reduced upon ablation of MHC-II interaction, suggesting that Tregs might also interact with these cells in an MHC-II-independent manner. scRNAseq analysis of LIPSITC-labelled cDCs by T cells in LNs showed that migratory cDCs are preferentially labelled in the LNs, and in particular 3 clusters of cDCs expressing genes reminiscent of cross-presentation or of cDC activation that were completely LIPSIC-labelled. This technology will be of high interest to study interacting cells, such as DCs with T cells, across various immune contexts.
In the 4th session about “tissue-specific functions of DCs and macrophages”, Ruth Franklin (Harvard Stem Cell Institute, Boston, USA) presented work revolving around the balance between host defence and tissue repair in the context of flu infection, where the immune response needs to balance between eradication of the infection while avoiding immunopathology. The talk was focused on the IL-6 family cytokine Oncostatin M (OSM) that is expressed by macrophages but whose role is unknown. Strikingly, OSM-deficient mice succumbed from flu infection but not because of absence of viral control. She showed that the drastic loss of body weight in OSM-deficient mice was associated with higher levels of type 1 interferons and ISGs expression. Strikingly, OSM-deficient mice also succumbed to Poly IC, phenocopying flu challenge but in the absence of infection. Blocking IFNA rescued the mice, suggesting that this cytokine induced immunopathology in these contexts. Altogether, a model emerged where macrophages produce OSM to limit type 1 interferons-driven immunopathology in the context of flu infection, and in potentially other systems.
In his talk about “Biology of DC’s and macrophages in the lung”, Bart Lambrecht (VIB Center for Inflammation Research, Ghent, Belgium) presented work on cDC2s in the context of Th2 immunity, in particular in the model of papain which enzymatic activities mimics those of other allergens such as house dust mites. To understand which antigen presenting cells (APCs) induce T cells in this model, Bart presented work where his group imaged papain antigens-specific T cells (transgenic PAPER T cells), that co-localized with Sirpa+ cells in LNs (including cDC2s and monocyte-derived cells) and potentially representing the APCs that prime PAPER T cells. Although monocytes are often mistakenly considered as not able to migrate to LNs because of their lack of CCR7 expression, Bart Lambrecht presented data showing that monocytes do migrate to LNs in a CCR5-dependent manner, following CCL5 gradient produced by CCR7+ cDC1s as they migrate to LNs. However, he also showed that although they migrate to the LNs, monocytes actually do not present papain peptide to PAPER CD4+ T cells ex vivo, but that cDCs do. At this point of the talk, Bart Lambrecht provided a word of caution when comparing the ability of different APCs to present and activate T cells in vitro: although the majority of APCs might be able to activate T cells when adding peptides to the culture, it does not mean that this happens in vivo. APCs should be retrieved and analysed ex vivo, which allows to consider the activation and antigen processing that the APC has experienced within tissue and in the right immune context. Next, looking at the immune context polarizing the Th2 response in the context of papain, an important point was that the source of IL-4 did not come from cDCs but from the T cells themselves, and that papain can degrade IL-12 made by cDCs. This work sheds light on the processes that lead to Th2 immunity and allergy, and on how cDCs play a role in this context. Last, Bart Lambrecht discussed an interesting connection between respiratory infection (RSV) and the development of allergy in neonates: since RSV can induce expression FcR on APCs, these cells might be able to take up IgGA-bound allergens present in the breast milk of allergic mothers and instigate a Th2 response against the allergen.
Pasqual, G., Chudnovskiy, A., Tas, J. et al. Monitoring T cell–dendritic cell interactions in vivo by intercellular enzymatic labelling. Nature 553, 496–500 (2018). https://doi.org/10.1038/nature25442