Visualizing Immunity and Inflammation in Tissues Using Dynamic and Multiplex Static Imaging - Ron Germain
Session 1 - Bridging to cellular immunity
Background: Previously, we observed a reduced switch from antigen-specific IgM to IgG autoantibodies in C5ar1-/- mice in a murine model of autoimmune skin disease, associated with strong upregulation of C5aR1 in Germinal Center B cells (GCB). Here, we used an in-vitro model of GCB induction to define the pathways downstream of the C5a/C5ar1 axis controlling the switch from IgM to IgG production in induced GCBs (iGCB).
Methods: We co-cultured 3T3 cells overexpressing CD40-ligand and B-cell-activating factor (BAFF) (40LB cells) (Haniuda and Kitamura 2019) with naïve B cells from either C57BL/6 WT or C5ar1-/- mice in the presence of IL-4 for 4 or 6 days. At both time points, we determined the expression of the GCB markers GL7 and CD95 (Fas), IL-4-receptor, IgG1 and IgM production as well as CD138+ plasma cell (PC) differentiation. Further, we assessed expression of C5 and C5aR1 as well as C5a generation in 40LB, naïve B cells and iGCBs.
Results: We observed strong intracellular expression of C5, C5a and C5aR1 in 40LB cells. A small fraction of naïve B cells from WT or C5ar1-/- mice expressed C5 but not C5a or C5aR1. After 4d of co-culture CD95 but not GL7 expression was reduced in iGCBs from C5ar1-/- as compared to WT mice. In iGCBs from WT mice, 65% expressed C5 and 50 % C5a at d4, which was significantly lower in iGCBs from C5ar1-/- mice. Strikingly, IL-4 receptor expression was downregulated in iGCBs from C5ar1-/- mice at d4, which was associated with a reduced IgM to IgG1 switch at d4 and d6 as compared with iGCBs from WT mice. Finally, we found induction of PCs from iGCBS at d6, the frequency of which was significantly lower in cells from C5ar1-/- as compared to those from WT mice.
Conclusion: Our findings suggest a model in which autocrine (in iGCBs) and/or paracrine (in 40LB cells) production of C5 and generation of C5a activates C5aR1 in iGCBs to drive their differentiation and expression of IL-4 receptor, which is crucial for the switch from IgM to IgG1 antibody production. Later, C5a/C5aR1 axis activation controls the differentiation of PCs from iGCBs.
Complement C5a has been shown to be critically involved in many autoimmune diseases through activation of its cognate receptors, C5aR1 and C5aR2, expressed on effector cells. While the role of C5aR1 in autoimmune disease is well understood, the function of C5aR2 remains elusive. Initially thought to be a decoy receptor that negatively regulates the functions of C5aR1, C5aR2 has recently been shown to exert both pro- and anti-inflammatory properties independent of C5aR1. Recent experiments by our group using C5aR2 reporter mice showed that B cells express C5aR2 but not C5aR1 in the steady state. We also observed an altered maturation of B cells in C5ar2–/– mice. These findings prompted us to further investigate the role of C5aR2 in B cells in the context of early autoimmunity. First, using flow cytometry to assess the role of C5aR2 for B cells under homeostatic conditions, we found that naive splenic B cells from C5ar2–/– mice showed a drastically increased expression of the decay accelerating factor (DAF, CD55) compared to B cells from wild-type and C5ar1–/– mice. Since the regulation of DAF expression has been shown to be critical for a functional germinal centre response, we hypothesised that the increased DAF expression in B cells of C5ar2–/– mice might affect their differentiation. To investigate this, we used an in vitro culture system in which we co-cultured splenic B cells with 40LB fibroblast feeder cells that exogenously express CD40‑ligand (CD40L) and B cell activating factor (BAFF), which promote B cell proliferation and survival and mimic the germinal center response. Flow cytometric analysis of the co-cultured B cells revealed attenuated activation, differentiation and class switching of B cells from C5ar2–/– mice compared to the wild-type. These results also suggest a potential role for C5aR2 in autoimmunity, which we will investigate in future experiments. Taken together, our findings highlight a potential role of C5aR2 as a novel checkpoint for B cell development, possibly by regulating DAF expression during B cell maturation.
Background: Complement factor H (FH) is a critical complement regulator and its absence results in kidney disease. FH deficiency renders C57BL6 mice susceptible to immune complex mediated glomerulonephritis. Significant macrophage infiltration occurs in this setting and understandably, depletion of macrophages reduced the disease pathology. Macrophages participate in defense, inflammation, restoration and fibrosis of the kidney.
Gap in knowledge: Macrophages express intracellular FH, but its role in these cells remains incompletely understood. The present study steps into this gap in our understanding and investigates the role of FH in macrophages.
Methods and Results: Wild type (WT) and FH knockout (KO) bone marrow derived macrophages were isolated, cultured and analyzed to understand the impact of FH deficiency. In FH deficient macrophages, nanostring gene expression analysis using mouse myeloid panel and stringent analysis revealed that of the 754 genes on the array, 73 genes were increased and 20 genes were decreased with at least 1.5 fold change and p<0.05. The significantly increased genes include the inflammatory genes TNF, Nlrp3, IL1B and IL6. In line with the increased mRNA expression of VEGF, ATF3, Egr3, Myc, Csf1R genes, macrophage trafficking and migration were increased compared to WT macrophages. Phagocytosis was reduced in FH deficient macrophages, and oxidative metabolism was decreased by 40%. Since transient receptor potential (TRP) channels modulate macrophage functions such as phagocytosis and cytokine generation, we studied the effect of FH on these channels. FH modulated the mRNA expression of the transient receptor potential cation channel, subfamily M, member 7 (TRPM7) in macrophages. In line with increase in mRNA expression of TRPM7, patch clamp assessment showed that TRPM7 current was significantly higher in FH deficient macrophages compared to wildtype macrophages. The TRPM7 current was blocked by FTY720, a specific TRPM7 inhibitor. FTY720 also significantly attenuated the FH regulated proliferation of macrophages.
Conclusion: For the first time our results show that FH is of functional relevance in macrophages, and establishes a critical link between intrinsic FH and macrophage effector functions through TRPM7, which could serve as a potential therapeutic target in FH modulated disease settings.
Background: Complement and B cells are implicated in the pathophysiology of rheumatoid arthritis, an autoimmune disease characterized by synovial inflammation. Loss of complement regulator CD55 leads to delayed onset of inflammatory arthritis in the collagen-induced mouse model of arthritis (CIA).1 Marginal zone (MZ) B cells, an innate-like B cell population, serve as first responders to blood-borne pathogens by rapidly differentiating into antibody-producing plasmablasts, and are important initiators of CIA.2 Whether and how CD55 regulates MZ B cells in CIA pathogenesis is unknown.
Methods: We examined naïve and sorted MZ B cell populations in 3-5 week-old and 10-12 week-old CD55-/- and compared to wildtype (WT) C57BL/6 mice using flow cytometry, immunofluorescence, bromodeoxyuridine (BrdU) incorporation, and bulk RNAseq. Differential gene expression analysis was performed, followed by Gene Sequence Enrichment Analysis (GSEA).
Results: We found that loss of CD55 decreased the number of MZ B cells in 10-12 week-old, but not 3-5 week-old mice. The precursors to MZ B cells were unchanged at both endpoints, indicating that loss of MZ B cells was not due to failure of MZ B cell differentiation. En vivo BrdU incorporation showed no difference between CD55-/- and WT, indicating MZ B cell differences are not due to loss of proliferative capacity. Differential gene expression analysis showed decreased expression of genes involved in preventing apoptosis in MZ B cells of CD55-/- compared to WT mice (HSP90b1, STIP1, AHSA1, and FKBP4). GSEA showed decreased expression of cell cycle/cell survival pathways in CD55-/- cells relative to WT. Flow cytometry revealed that CD55-/- MZ B cells had corresponding increases in activated caspase-3 and caspase-7, as well as increases in cellular reactive oxygen species relative to WT cells.
Conclusion: Loss of CD55 led to increased reactive oxygen species and activation of caspases in MZ B cell populations, which may prevent the normal expansion of the MZ B cell subset over time. These studies suggest a role for the complement regulator CD55 in MZ B cell homeostasis and survival. Future experiments are aimed at defining whether loss of CD55 impacts MZ B cell survival through complement-dependent or independent mechanisms.
Session 2 - Preclinical models in kidney diseases
Complement mediated atypical haemolytic uraemic syndrome (c-aHUS) arises from impaired regulation or gain-of-function changes in the alternative pathway (AP). Eculizumab is the gold standard therapy for c-aHUS but there is room for improvement in patient care. The unique mouse model of aHUS based on a single D1115N amino acid substitution in C3, (C3N/N, PMID: 30714990) provides a testbed to evaluate targeted complement inhibition and we aimed to establish new effective therapies in c-aHUS.
Therapeutic blockade of the AP using anti-properdin monoclonal antibodies (mAb H4, 14E1), CR2-FH, homodimeric minimal-FH (HDM-FH – protein and gene therapy) or FH1-5, and terminal pathway (TP) blockade with an anti-C7 mAb or an oral C5aR antagonist (ACT-1014-6470) was carried out using previously published dosing strategies. A prophylactic (from weaning) or rescue strategy (after 2 days of haem at 25 ery/ul) was used as appropriate. Upon study completion, animals were culled, tissue collected, and analysed for histological evidence of renal disease, fibrin deposition and glomerular complement deposition. Plasma was analysed for renal impairment (BUN), haemolysis indices (Haemoglobin, reticulocyte and platelet count) and complement levels.
No C3N/N mice succumbed to aHUS during prophylactic use of the oral C5aR1 antagonist (up to 80 days) which mirrored genetic deletion of C5aR1 (as previously reported1). Similarly, anti-C7 mAb therapy rapidly reduced evidence of renal damage and no mice succumbed to disease. Anti-properdin mAb 14E1 (similar to genetic deletion of CfP2) and HDM-FH (protein) were also highly effective in restoring renal function but CR2-FH (even when given daily) and FH1-5 largely failed to prevent acute TMA and renal damage in C3N/N mice. Therapeutic use of the oral C5aR antagonist and prophylactic gene therapy using HDM-FH is ongoing with no mice succumbing to TMA thus far.
Our data suggest that therapeutic targeting of either arm of the TP and disruption of the C5 convertase can significantly reduce renal damage and prevent c-aHUS (in the context of a C3 gain-of-function change) providing novel insight into disease mechanisms as well as illuminating potentially new disease maintenance and treatment strategies.
References: Conference abstracts by 1. Smith-Jackson et al and 2. Kamala et al, see 10.1016/j.molimm.2022.07.003
Atypical haemolytic uraemic syndrome(aHUS) is rare kidney disease, caused by thrombotic microangiopathy (TMA). Mutations in complement regulators and activators can ‘trigger’ TMA but exact mechanisms remain unclear. Recently, we showed that a C3 gain-of-function point mutation (C3D1115N) when in homozygous (or C3N/N) led to TMA and aHUS in mice. However, as patients seldom have homozygous mutations, we sought to investigate whether aHUS could be ‘precipitated’ in C3+/N (heterozygous) mice following various stimuli, that would mimic infections long suspected to be linked to aHUS in patients.
C3+/N mice were injected intraperitoneally with O111:B4 lipopolysaccharide (LPS) at various concentrations (0.1, 0.2 or 0.5 mg/kg) and were monitored (weight and urine analysis) until day (D)3. In additional studies, mice received 0.2 mg/kg LPS on D0 and D7, mice were tracked until D10. Alternatively, C3+/Nand wild type mice received an intra nasal dose of influenza strain X31 (7.8×105 plaque forming units) and monitored until D14. In our final approach, a Freund’s adjuvant free method of nephrotoxic serum (NTS) nephritis was performed. Briefly, mice received two priming doses of sheep IgG on D0 and D7, before receiving 25, 50, 100 or 200µl of NTS on D14, mice were monitored to D17. In all cases, serum was analysed at cull for evidence of renal disease and complement activation split products. Kidneys were assessed for TMA by histological approaches.
While there was evidence that C3+/N mice where more susceptible to these agents, with increased proteinuria/weight loss being common, no mice succumbed to TMA or exhibited significant haematuria. Furthermore, while altered B and T cell populations were found and greater levels of C3 breakdown products detectable in the serum and kidney of the C3+/N mice, this did not translate into aHUS.
These data confirm that excess complement activation systemically, through viral infection or immune complex is not sufficient to precipitate aHUS in C3+/N mice. These data indicate the mouse is potentially more robust than man with respect to development of aHUS and CRRY may play a role in this situation. Although further investigation is needed to clarify this finding.
Background: Mannose-binding lectin-associated serine protease-1 (MASP-1) and MASP-3, transcribed form the common Masp1 gene, play critical roles in the lectin and alternative complement pathway activation, respectively. We previously reported that Masp1-knockout (deficient for both MASP-1 and MASP-3) lupus-prone MRL/lpr mice showed significantly reduced glomerulonephritis and albuminuria compared to their wild-type littermates. The data suggested that MASP-1 and/or MASP-3 play an important role in the development of lupus-like glomerulonephritis in these mice via activation of the lectin and/or alternative pathways. Here, we aimed to clarify the role of MASP-1 in the development of renal disease in lupus by analyzing disease expression in MRL/lpr mice monospecifically deficient for MASP-1.
Methods: MASP-1-deficient mice were generated by the CRISPR/Cas9-mediated genome editing in the C57BL/6 background and then backcrossed to the MRL/lpr background for eight generations. Sera and urine were collected biweekly from groups of MASP-1-deficient and wild-type MRL/lpr mice starting at 12 weeks of age. Serum C3, anti-dsDNA IgG antibody and urinary albumin excretion levels were measured by ELISA. Circulating immune-complex levels were measured by ELISA utilizing anti-C3 and anti-mouse IgG antibodies. Mice were sacrificed at 24 weeks of age and kidneys were collected for histopathological analysis.
Results: MASP-1-deficient MRL/lpr mice showed no significant differences in serum C3, anti-dsDNA IgG antibody and circulating immune-complex levels compared to their wild-type littermates. Wild-type MRL/lpr mice developed albuminuria as early as 12 weeks of age, whereas MASP-1-deficient MRL/lpr mice developed albuminuria around 16 weeks of age. Wild-type MRL/lpr mice started to die as early as 16 weeks of age, whereas no MASP-1-deficient MRL/lprmice died until 20 weeks of age (p = 0.019 at 19 weeks of age). However, histopathological analysis of the kidneys collected at 24 weeks of age showed no significant differences in C3 deposition levels and in the development of lupus-like glomerulonephritis including proliferative, necrotic and crescentic changes between the two groups.
Conclusion: Our results suggest that MASP-1 accelerates the development of albuminuria and mortality in MRL/lpr mice, most likely via activation of the lectin pathway.
Leptospirosis is a neglected zoonosis affecting approximately 1 million people each year worldwide and causing near 5% deaths. Most patients are asymptomatic or have only mild clinical manifestations. However, some patients develop severe chronic and end-stage kidney disease responsible, which may be fatal. We are investigating the role of C3 in the chronic leptospirosis by using C57BL/6 wild-type (WT) and C57BL/6 C3-knockout (C3KO) male mice infected with 108 L. interrogans serovar Copenhageni strain Fiocruz L1-130 (LIC) or PBS (negative control) after 15, 30, 60, 90 and 180 d.p.i.. We observed the following results: i) all infected mice survived. However, more leptospires were observed in the kidney of LIC-C3KO mice up to 90 d.p.i., when compared to LIC-WT mice, by immunohistochemistry analysis; ii) LIC-infected mice lost more body weight during the first 5 days of infection when compared to PBS, independent of the presence of C3; iii) no significant differences in the spleen/total body weight ratio were observed between LIC-WT and LIC-C3KO after 15 d.p.i. or 30 d.p.i.; iv) LIC-C3KO mice presented higher scores of histopathological alterations in the liver and kidney, when compared to LIC-WT mice at 30 d.p.i up to 180 d.p.i.; v)) higher percentage of fibrosis was observed in the kidney from LIC-C3KO mice when compared to LIC-WT at 30 d.p.i., after Red Sirius staining, which reveals the presence of collagen I and III in the tissue. In conclusion, C3 did not affect the survival of LIC-infected C57Black/6 mice but it was important for the in vivo killing of leptospires. Leptospira remains for longer time in the liver and kidney of LIC-C3KO mice when compared do WT in the same conditions. In addition, the lack of this protein may trigger renal fibrosis in the infected mice at 30 d.p.i. suggesting that the presence of local C3 may control tissue repair after damage by pathogens.
Committee meeting ECCO
Session 3 - C5a and receptors in health and disease
Background: C5a regulates the differentiation and function of dendritic cell (DC) subsets from bone marrow-derived precursors through C5aR1 activation. Here, we determined the impact of the C5a/C5aR1 axis on FMS-like tyrosine kinase 3 ligand (Flt3L)-induced in vivo generation and function of splenic conventional DC (cDC) subsets.
Methods: Flt3L expressing B16 melanoma cells were subcutaneously injected into wildtype and C5ar1-/- mice. After 10 days spleens were harvested. cDC1 and cDC2 subsets were FACS-sorted, incubated with LPS-free ovalbumin (3 or 10 ug/ml) ± TLR ligands (LPS, PAM3) and co-cultured for 72h with ovalbumin-specific TCR transgenic OT-II CFSE-labelled CD4 T cells. Then we determined the generation of effector memory (TEM) and effector T cells (TEFF) and their differentiation towards Th1 (IFN-y) and Th17 (IL-17A) effector lineages.
Results: Flt3L induced mobilization of cDC1 and cDC2 cells in the spleen was independent of C5aR1 expression. Ovalbumin stimulation ex vivo resulted in strong TEFF and less TEM generation by cDC2, which was markedly reduced in cDC2 from C5ar1-/- mice at low ovalbumin concentration. Surprisingly, TEFF and TEM generation was absent using cDC1 cells. Additional LPS/PAM3 stimulation of cDC2 cells from either wildtype or C5ar1-/- mice increased the dominant TEFF cell differentiation. In contrast, TEM and TEFF induction was similar using cDC1 cells from wildtype or C5aR1-deficient mice. Strikingly, we found a strong and dominant IFN-y production upon wildtype cDC2 cell stimulation with OVA ± TLR ligands and OT-II cell co-culture, which was significantly reduced using C5ar1-/- cDC2 cells.
Conclusion: Our findings suggest that cDC2 but not cDC1 cells are critical for antigen-driven T cell proliferation and TEFF differentiation in the absence of TLR ligands. Further, C5aR1 activation seems to be crucial for this effect at low but not at high doses of antigen. Also, while C5aR1 activation did not affect TEFF or TEM cell differentiation induced by cDC1 or cDC2 following OVA ± TLR ligand stimulation, it controlled the dominant Th1 induction mediated by cDC2 cells. Thus, we identified a novel role for C5aR1 in antigen-driven TEFF differentiation by splenic cDC2 cells and subsequent differentiation into Th1 cells in response to pattern recognition receptor activation.
Mycobacterium avium (Mav) is one of the most successful pathogens infecting billions of people with daily exposure to water and soil. Mav belongs to the group of non-tuberculous mycobacteria (NTM) which affects the elderly and immunocompromised individuals. With the evasion and alteration of host inflammatory response, Mav persistence leads to severe pathogenicity resulting in interstitial chronic lung disease. Given that the C5a peptide is pivotal in chronic inflammatory and infectious diseases, we studied the role of C5a receptor 1 (C5ar1) signaling response in the in vivo Mav pathogenicity. We infected C5ar1-deficient mice (C5ar1KOC57B6/J), and the wildtype littermates (WTC57B6/J) with the aerosol exposure of a clinical strain ofMav(5×108 CFU/ml) and examined the immunopathologic changes up to 90 days post-infection (dpi). The Ashcroft scores obtained using Masson’s trichrome staining revealed significant fibrogenic burden withMav granulomas in the 30-dpi Mav-WT lung tissues that increased through 60 and 90-dpi than their C5ar1-deficient counterparts. Significantly increased inflammatory cytokines expressions including Il-1β and Tnfα were observed in the Mav- WT lungs at 60 and 90-dpi, which also corroborated with significantly increased proinflammatory helper T cell (Th1) specific marker Ifnγ, and its regulator T-Box transcription factor 21, Tbx21expressions in the BAL fluid cells. These markers were significantly low in the Mav-C5ar1KO mice correlating with low Ifnγ levels measured by ELISA. Importantly, marked Tbx21/T-bet immunofluorescent staining indicated Th1-specific immune cell subset in the Mav-WT lungs. Finally, we confirmed the role of C5ar1-dependent Th1 response in Mav-induced fibrogenic changes using splenocytes isolated from WT and C5ar1KO mice. In the absence of C5ar1 signaling, Mav-infected splenocytes demonstrated an impaired helper T cell response with a significant decrease in the Ifnγ and Tbx21 transcript levels, and decreased CD4+ /CCR5+-proinflammatory Th1 cell subsets were measured flow-cytometrically. Therefore, our data indicated a C5ar1-dependent proinflammatory Th1 response during the fibrotic tissue remodeling in Mav-infected mouse lungs.
Streptococcus pneumoniae (Spn) is one of the leading bacterial pathogens causing mortality among adults. Complement activation plays a pivotal role in the clearance of encapsulated Spn and the pathogenesis of ARDS. The subsequent release of the anaphylatoxin C5a promotes the recruitment and activation of inflammatory cells including neutrophils after C5a recognition by its two homologous receptors, C5aR1 and C5aR2. The blockade of C5a attenuates pro-inflammatory responses and rescues lethality in various experimental models of sepsis and inflammation. However, there is insufficient evidence whether C5a neutralization exerts protective or harmful effects during Spn infection.
We established a novel homozygous C5aR1/2-/- double-knockout mouse strain using CRISPR/Cas9 guided gene editing to assess the direct role of C5aR1 and C5aR2 during Spn infection. To further profile alveolar cell populations at the single cell level, we performed TOTAL-seq on FACSorted live CD45+ BALF cells using ~200 oligonucleotide-conjugated antibodies enabling simultaneous detection of surface protein markers and RNA.
The C5a-induced influx of neutrophils to the airways was completely abrogated in C5aR1/2-/- mice compared to C57BL/6J wild-type (WT) mice. In striking contrast, C5aR1/2-/- mice showed a stronger inflammatory response in the alveolar spaces after Spn infection, as indicated by enhanced neutrophil infiltration, increased cytokine/chemokine secretion, and exacerbated lung vascular permeability. Interestingly, the neutrophils recruited in C5aR1/2-/- mice released more neutrophil extracellular traps (NETs) but exhibited impaired bactericidal function as suggested by higher CFUs in BALF. Single cell proteotranscriptomics revealed 8 distinct subclusters of alveolar macrophages and 7 subclusters of neutrophils after Spn infection – which displayed heterogeneous levels of C5aR1 protein and C5aR2 RNA expression in WT mice. Consistently, neutrophil clusters were more abundant in C5aR1/2-/- mice but macrophage counts were reduced. Furthermore, C5aR1/2-/- neutrophils developed signs of exhaustion resembling human septic neutrophils with elevated expression of PD-L1 and ICAM-1, and decreased expression of CXCR2 and CD62L. These findings would explain the reduced capacity of neutrophils to control bacterial growth in C5aR1/2-/- mice, creating a potent chemotactic local environment for a persistent neutrophil influx and therefore, resulting in greater lung damage.
Concomitant inhibition of the C5a-C5aR1/2 axes could drive neutrophil exhaustion and lung injury after Spn infection.
Background: The anaphylatoxin receptors play a critical role in the pathogenesis of various autoimmune diseases. In this context, the enigmatic second receptor for the anaphylatoxin C5a, C5aR2, is often overlooked. In this study, we aimed to elucidate the impact of C5aR2 on the pathogenesis of epidermolysis bullosa acquisita (EBA), a prototypical autoimmune skin disease characterized by neutrophil-dependent skin blistering in which complement activation is a prerequisite.
Methods: We used mice with a global or a LysM-specific deficiency of C5ar2 in a passive model of EBA. In this model, mice were subcutaneously injected with antibodies against type VII collagen (COL7) to induce neutrophil-dependent blistering of the skin. In addition, we performed in vitro assays to assess how targeted deletion of C5ar2 affects C5a-induced activation and effector functions of neutrophils. In these experiments, neutrophils from wild-type and C5ar1–/–mice served as controls. We employed single-cell RNA-sequencing to identify differentially expressed genes and elucidate the molecular mechanisms underlying C5a-induced activation in neutrophils.
Results: Mice lacking C5aR2 globally or specifically in LysM-positive cells were found to have a significantly ameliorated disease phenotype compared with wild-type mice. The relative C5a-mediated activation and effector functions of neutrophils in these mice were correspondingly decreased. This reduced responsiveness to C5a stimulation was associated with a lower number of differentially expressed genes and a changed transcriptome profile compared with neutrophils from wild-type mice. Notably, neutrophils lacking C5ar1 expression did not respond to C5a stimulation, which was reflected in an unchanged transcriptome profile.
Conclusion: These results suggest that the often-neglected C5aR2 is a critical contributor to neutrophil-driven autoimmune diseases such as EBA. While our in vitro assays clearly indicate a dependence of C5aR2 on C5aR1, our single-cell RNA-sequencing results shed light on how signaling through C5aR2 affects gene expression. However, to fully understand the signaling cascade downstream of C5aR2 and its dependence on C5aR1, further research is needed. This will also help determine the potential efficacy of targeting C5aR2 to treat neutrophil-driven diseases such as EBA.
Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is a major cause of death in RA patients with 40% mortality rate within 72 months from the time of clinical diagnosis. RA-ILD is manifested with usual and non-specific interstitial pneumonia and lung fibrosis. The role of complement activation is poorly understood in the pathogenesis of RA-ILD. We hypothesized that complement anaphylatoxin C5a via its receptor C5ar1, may contribute to the pathogenesis of RA-ILD. To unravel the molecular mechanisms of C5ar1-signaling, we induced collagen-induced arthritis in C5ar1 deficient-C57B6/J (C5ar1KO) mice and their wildtype littermates (WT). A set of arthritic mice were challenged with intradermal C5a (2.5mg/50μl) at 24-hour intervals for 27 days. All mice were euthanized on day 28 and lung tissues were collected for histologic and molecular analyses. Masson’s trichrome staining revealed pleural hypertrophy and fibrogenic foci in the WT arthritic lungs. C5a-challenged WT arthritic mice revealed enhanced hyaluronan production and presence of myofibroblast activation marker, α-SMA as compared to the C5a-challenged C5ar1KO group. We further approached our hypothesis with Poly(A)-ClickSeq RNA sequencing analysis and the PolyA-miner algorithms to study alternative polyadenylation (APA) as a post-transcriptional mechanism of gene regulation involving 3’-UTR sequences. Distinct 3’-UTR landscapes in proinflammatory genes result in transcript isoforms lacking regulatory miRNA binding sites during chronic inflammation. Our data indicated increased 3’-UTR lengthening profile with 477 transcripts in arthritic C5ar1KO mice compared to their WT counterparts. This included the presence of cleavage and polyadenylation specificity subunit 5 (Cpsf5/ Nudt21), a key regulator of pulmonary fibrosis. C5a-pulsed WT mice showed 1240 3’-UTR-shortened genes including Notch4 and RNA-binding protein Zfp871, compared to their C5ar1KO counterparts. C5a-pulsed arthritic WT mice demonstrated 972 disease-specific 3’-UTR-shortened genes relating C5ar1 signaling with endosomal trafficking, histone modifications, cellular senescence, and fibrogenesis. Intriguingly, significant upregulation of miRNA-29a and miRNA-30 was observed in C5a-arthritic mice, given that miRNA-29a is involved in inhibiting histone modification and fibrogenesis; and miRNA-30 targets cellular senescence. Thus, our data offer a novel mechanism of C5ar1-dependent 3’-UTR shortening in profibrogenic modulators that attenuates regulatory targeting by miR29a and miR-30 during arthritic lung remodeling.
In ischemic tissue, platelets can modulate angiogenesis. The specific factors influencing this function, however, are poorly understood. Here, we characterized the complement anaphylatoxin C5a-mediated activation of C5a receptor 1 (C5aR1) expressed on platelets as a potent regulator of ischemia-driven revascularization and assessed its role if diseases featuring tissue ischemia in patients.
We have employed genetic mouse models to substantiate this. Indeed, the presence of C5aR1-expressing platelets was increased in the hindlimb ischemia model, a mouse model of ischemia-induced revascularization. Ischemia-driven angiogenesis was significantly improved in C5aR1-/- mice, but not in C5-/- mice suggesting a specific role of C5aR1. Experiments using supernatant of C5a-stimulated platelets suggested a paracrine mechanism of angiogenesis inhibition by platelets by means of antiangiogenic CXC chemokine ligand 4 (CXCL4, PF4). Lineage-specific C5aR1 deletion verified that the secretion of CXCL4 depends on C5aR1 ligation on platelets. WE characterized this secretion ot be a specific sub-alpha-granule specific event featuring a specific signaling cascade. Finally, we applied the C5aR1 inhibitor PMX-205 in vivo in the hindlimb ischemia model. This drug induced a phenotype of improved revascularization in mice.
Furthermore, we assessed the relevance of the anaphylatoxin receptor C5aR1 on platelets in coronary artery disease and found a correlation between specific but not all markers of platelet activation and C5aR1 expression on platelets. Furthermore, we have assessed the role of C5aR1 in peripheral artery disease (PAD) patients. We have matched a cohort with proven high-grade arterial stenoses of the leg or groin arteries but no ischemic walking pain with patients with PAD and typical symptoms. We hypothesized that lack of symptoms might be due to a better capacity for collateral artery growth in asymptomatic PAD patients. Strikingly, the C5aR1 expression on platelets and CXCL4 correlated with pain-free walking distance Thus, the C5aR1-CXCL4 axis seems to play a role in adaption to diseases featuring tissue ischemia also in cardiovascular patients.
We identified a novel mechanism for inhibition of neovascularization via platelet C5aR1, which was mediated by release of antiangiogenic CXCL4. Importantly, we can present evidence for the significance of this mechanism in cardiovascular patients suffering from diseases featuring tissue ischemia.
SVAR Prize: Awardees announced by SVAR representatives
Session 4 - Complement in cancer
Background: Loss of function STK11 mutations occur in 15-20% of non-small cell lung cancer (NSCLC) and are associated with poor survival and resistance to immune checkpoint inhibitors (ICI). We observed in human STK11-mutant NSCLC upregulation of C3, Factor D, Factor H, and CD55. We previously observed that in the tumor microenvironment (TME), neutrophils (PMN) acquire complement-dependent T cell inhibitory function. Less is known about the roles of tumor-derived complement in tumor progression. We hypothesized that tumor-derived complement will promote growth of STK11-mutant tumors in vivo.
Methods: CMT167 (syngeneic NSCLC) tumor cells with Stk11 deletion with and without C3 deletion were generated by CRISPR-Cas9. To evaluate the role of tumor-derived versus systemic complement in tumor progression, we compared growth of subcutaneous (s.c.) administered Stk11KOC3WT and Stk11C3KO tumor in WT versus C3-/- mice. To test the interaction of tumor-derived C3 and T cell immunity in regulating tumor growth, we compared tumor growth in WT versus nude mice. Finally, we evaluated whether PMN depletion with anti-Ly6G or Cxcr2 inhibition would overcome resistance of Stk11KO tumors to anti-PD-1 treatment.
Results: We observed high infiltration of PMN and low CD8 T cell in Stk11KO tumor in vivo. Systemic C3 deficiency resulted in modest delay of growth of Stk11WT tumors, but had no effect on growth of Stk11KO tumors. Deletion of C3 in CMT167-Stk11KO (CMT167 Stk11/C3KO) resulted in dramatic inhibition of tumor growth in immune competent mice, but had no effect in nude mice. While anti-PD1 treatment alone had modest to no effect in the growth of Stk11KO tumors, the addition of anti-Ly6G or Cxcr2 inhibitor resulted in improved control of tumor growth.
Conclusions: Tumor-derived C3 drives Stk11KO tumor growth in mice. Growth of Stk11KO tumor was strikingly dependent on tumor-derived C3 in immunocompetent mice but dispensable in nude mice. These results support a role for tumor-derived C3 suppressing T-cell immunity, potentially indirectly through recruitment of PMN or inducing PMN suppressor function. Our results also provide rationale for targeting tumor-derived complement and inhibiting Cxcr2 to enhance ICI efficacy as novel therapeutic approaches in STK11-mutant NSCLC.
Background: Normal cells are equipped with numerous membrane-bound complement inhibitors that protect them from misguided complement attack. However, soluble complement inhibitors are rarely produced in locations other than the liver. Previously, we reported the expression of complement factor I (FI) in non-small cell lung cancer (NSCLC) cell lines.
Methods: FI expression in cancer biopsies from lung adenocarcinoma and squamous cell carcinoma patients was assessed by immunohistochemical staining and associated with clinicopathological characteristics and clinical outcomes. To approach the question of whether the expression of FI by tumor cells was aimed to protect tumor cells from host innate immunity, the deposition of C4d – the end degradation product of FI-supported inactivation of active complement component C4b was analyzed in the same tissue. To elucidate the role of FI in lung cancer cell physiology, three human non-small lung cancer cell lines naturally expressing FI were engineered with CRISPR/Cas9 technology and the transcriptome of FI-deficient and FI-sufficient clones was compared in each cell line. These cells were also compared in in vitro colony-formation assay.
Results: FI immunohistochemical staining intensity did not correlate with age, smoking status, tumor size, differentiation grade, stage, T cell infiltrates or PD-L1 expression, but was associated with progression-free survival (PFS) and disease-specific survival (DSS). Multivariate Cox analysis of high vs. low expression of FI revealed HR 0.55, 95% CI 0.33-0.95, p=0.031 for PFS and HR 0.35, 95% CI 0.15-0.78, p=0.011 for DSS. Importantly, only negligible C4d staining was found in cancer tissue. RNA sequencing in FI knockout and wild-type NSCLC cells revealed differentially expressed genes of potential importance for intracellular signaling pathways controlling proliferation, apoptosis, and responsiveness to growth factors. In vitro colony-formation assays showed that FI-deficient cells formed smaller foci than FI-sufficient NSCLC cells, but their size increased when purified FI protein was added to the medium.
Conclusions: We postulate that a non-canonical activity of FI influences cellular physiology and contributes to the poor prognosis of lung cancer patients.
This project was supported by National Science Centre (Poland) grant no. 2014/14/E/NZ6/00182
Background: The complement system can exert pro- or anti-tumorigenic functions based on different tumour types and contexts, suggesting novel perspectives for therapeutic targeting in selected subgroups of patients (1). We aimed at developing an innovative prognostic tool based on the in situ quantification of several complement components for the prediction of malignant pleural mesothelioma (MPM) patient outcome, introducing the definition of “Complement Score”.
Methods: Bioinformatics analysis via GEPIA and UALCAN (TCGA-MESO); immunohistochemistry on MPM tissue microarrays (TMAs,n= 88) and whole tissue sections (n=17); clinical information collection and statistics analysis.
Results: Bioinformatics analysis of genes encoding for complement activation components (n= 27) and regulators/receptors (n= 29) allowed the selection of four promising markers in MPM: the mRNA expression of C1S, SERPING1, CFB and CFI resulted to be positively correlated with patient overall survival. C1q was included in further analysis due to its abundance in MPM microenvironment (2). Protein expression of the selected complement components was evaluated both on TMAs and whole tissue sections of MPM patients, being expressed as percentage of tumour cell positivity, immune cell positivity and deposit. C1s, C1INH, CFB and CFI displayed cytoplasmic positivity of tumour cells, but also of immune cells and deposit, whereas C1q presence was only detected as deposit or positivity of monocytes/macrophages. Univariate analyses were performed in order to correlate the expression of each complement component with MPM histotype, TILs (CD4+, CD8+, CD19+), tumor proliferative activity (Ki-67), PD-L1 expression, overall survival. Survival analysis showed that C1qHIGH (Log-rank test, χ2 = 6.01;p= 0.01) and C1INHHIGH(Log-rank test, χ2 = 5.13;p= 0.02) patients displayed significantly increased survival.
Conclusions: Complement Score could be hopefully used in the future as a tool to stratify MPM patients, directing each patient to a personalized treatment, and its application could be potentially extended to other solid tumours, unveiling the “double-edged sword” role of complement system in cancer.
1. Revelet al., Antibodies 2020, https://doi.org/10.3390/antib9040057
2. Agostiniset al., Front Immunol 2017, https://doi.org/10.3389/fimmu.2017.01559
Background: Inflammation plays a pivotal role in the development and progression of colorectal cancer (CRC). Emerging evidence suggests that complement anaphylatoxin C3a produced upon complement activation and acting via its receptor (C3aR) may play a role in intestinal homeostasis. However, its role in CRC is currently unknown.
Methods: We used a comprehensive approach encompassing analysis of publicly available human CRC datasets, inflammation-driven and newly generated spontaneous mouse models of CRC, and multi-platform high dimensional analysis of immune responses using microbiota sequencing, RNASeq, and mass cytometry.
Results: By mining publicly available datasets, we found that CpG island methylation of c3ar1 occurs in CRC patients and is associated with significant downregulation of C3aR. By reverse-translating this finding, we were able to shift in APCMin/+ mice the tumorigenesis from the small intestine to the colon, therefore, generating a novel mouse model which more closely mirrors CRC in humans. RNAseq analysis on the polyps from our newly developed mouse model (APCMin/+/C3aR-/-) revealed a significant increase in immune signatures. Interestingly, loss of C3aR significantly impacted the fecal and tumor-associated microbiota, which promoted enhanced immune infiltration in typically “cold” tumors. In line with our findings in the mouse model, human CRC with C3aR downregulation showed increased innate and adaptive immune cells. Since immune infiltration is often a favorable prognostic factor in CRC and predisposes to response to immune checkpoint blockade (ICB) therapy, we assessed whether the enhanced immune infiltrate could be exploited to treat the tumors of APCMin/+/C3aR-/- mice. We found that using anti-PD1 in APCMin/+/C3aR-/- but not APCMin/+ mice resulted in significant tumor reduction. Therefore, the lack of C3a in the colon activates a microbiota-mediated pro-inflammatory program, promoting tumors’ development with an immune signature that renders them responsive to ICB therapy.
Conclusions. The complement system in the gastrointestinal tract is essential to avoid overt inflammation in health. However, this regulatory mechanism may restrain the activation of immune responses during tumor development. Our findings reveal that C3aR may act as a previously unrecognized checkpoint to enhance anti-tumor immunity in CRC. C3aR can thus be exploited to overcome ICB resistance in a larger group of CRC patients.
Despite the substantial preclinical evidence for the role of complement in cancer, complement-based cancer therapies have not yet advanced to the clinic, likely, because of the failure to identify the optimal targets within the complement system and patients that will benefit from complement-based interventions. To address these needs, we identified the subsets of renal cell carcinoma (RCC) that are associated with upregulation of complement genes and can be potentially regulated by complement-mediated mechanisms. These mechanisms include complement anaphylatoxins-mediated inhibition of antitumor immunity and acceleration of angiogenesis in mouse model of cancer. The upregulation of complement genes in patients was associated with T cell exhaustion and myeloid cells’ transcriptomic signatures linked to immune suppression. The complement-dependent subsets of RCC responded poorly to immune checkpoint inhibitors (ICI) and antiangiogenics. Abundance of plasma complement proteins was associated with the response to ICI, pointing to plasma complement as a potential predictive biomarker. In low grade RCC tumors, complement protein expression and deposition were limited to stroma and infiltrating cells, consistent with complement functions in the regulation of infiltrating immune cells. However, tumor cells in aggressive high grade RCC expressed large amounts of C3. In addition, we found expression of complement proteins and evidence of complement activation in several cellular organelles of tumor cells from common human cancers. Downregulation of these genes led to reduced tumor cell proliferation and downregulation of the mammalian target of rapamycin complex 1 (mTORC1) downstream signaling, consistent with the role of intracellular complement activation in T cells. Interestingly, tumors with hyperactivation of mTORC1 expressed high levels of complement proteins, suggesting the signaling regulatory loop involving complement and mTORC1. In conclusion, complement in human malignancies appears to regulate both host components of the tumor microenvironment that contribute to immunosuppression and tumor cell signaling essential for tumor cell growth. Thus, cancer patients are likely to benefit from targeting extracellular complement to improve antitumor immunity and complosome to stop tumor cell proliferation.
Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive malignancies with a survival rate at 5 years very low. Chemotherapy agents causes only a modest enhancement in the survival and is usually associated to toxic effects. The proteoglycan GPC1 showed a cell surface localization in PDAC tissues but limited or absent expression in the most of adult tissues and in chronic pancreatitis. Exploiting this promising tumor-associated antigen, an anti-GPC1 monoclonal IgM was produced with the aim to induce a strong complement activation on PDAC cells and in tumor microenvironment.
The antibody was characterized both in vitro and in vivo. Biodistribution studies, using near-infrared optical imaging technology, confirmed the capacity of anti-GPC1 IgM to selectively bind its target in a localized PDAC model developed in Nude mice, with a pick after 4 days; this antibody was mainly eliminated by the liver. Moreover, a single injection of anti-GPC1 IgM induced a strong activation of the classical pathway of the mouse complement system; as expected, C1q, C3 and C9 deposits were documented in tumor microenvironment by immunofluorescence analysis, causing extended necrotic areas as well as the recruitment of CD14+ macrophages and CD56+ NK cells. Repeated injection of anti-GPC1 IgM (twice a week) controlled tumor growth in all tumor-bearing mice. All saline-treated animals died in 19 days after the first treatment while 60% of anti-GPC1 IgM treated mice survived at the end of the study (50 days after the first treatment).
Collectively, these findings showed the capacity of anti-GPC1 IgM to strongly activate the complement system on PDAC cells, causing cancer cell death, modifying tumor microenvironment and finally increasing mice survival.