1. Immunology and Inflammation
Download icon

Squalene emulsion-based vaccine adjuvants stimulate CD8 T cell, but not antibody responses, through a RIPK3-dependent pathway

  1. Eui Ho Kim
  2. Matthew C Woodruff
  3. Lilit Grigoryan
  4. Barbara Maier
  5. Song Hee Lee
  6. Pratushya Mandal
  7. Mario Cortese
  8. Muktha S Natrajan
  9. Rajesh Ravindran
  10. Huailiang Ma
  11. Miriam Merad
  12. Alexander D Gitlin
  13. Edward S Mocarski
  14. Joshy Jacob
  15. Bali Pulendran  Is a corresponding author
  1. Emory Vaccine Center, Emory University, United States
  2. Yerkes National Primate Research Center, Emory University, United States
  3. Viral Immunology Laboratory, Institut Pasteur Korea, Republic of Korea
  4. Institute for Immunity, Transplantation and Infection, Stanford University School of Medicine, Stanford University, United States
  5. Department of Oncological Sciences, Tisch Cancer Institute and the Immunology Institute, Icahn School of Medicine at Mount Sinai, United States
  6. Department of Microbiology and Immunology, Emory Vaccine Center, School of Medicine, Emory University, United States
  7. Department of Physiological Chemistry, Genentech, United States
  8. Department of Pathology, Stanford University School of Medicine, Stanford University, United States
  9. Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford University, United States
Research Article
  • Cited 0
  • Views 1,252
  • Annotations
Cite this article as: eLife 2020;9:e52687 doi: 10.7554/eLife.52687

Abstract

The squalene-based oil-in-water emulsion (SE) vaccine adjuvant MF59 has been administered to more than 100 million people in more than 30 countries, in both seasonal and pandemic influenza vaccines. Despite its wide use and efficacy, its mechanisms of action remain unclear. In this study we demonstrate that immunization of mice with MF59 or its mimetic AddaVax (AV) plus soluble antigen results in robust antigen-specific antibody and CD8 T cell responses in lymph nodes and non-lymphoid tissues. Immunization triggered rapid RIPK3-kinase dependent necroptosis in the lymph node which peaked at 6 hr, followed by a sequential wave of apoptosis. Immunization with alum plus antigen did not induce RIPK3-dependent signaling. RIPK3-dependent signaling induced by MF59 or AV was essential for cross-presentation of antigen to CD8 T cells by Batf3-dependent CD8+ DCs. Consistent with this, RIPK3 deficient or Batf3 deficient mice were impaired in their ability to mount adjuvant-enhanced CD8 T cell responses. However, CD8 T cell responses were unaffected in mice deficient in MLKL, a downstream mediator of necroptosis. Surprisingly, antibody responses were unaffected in RIPK3-kinase or Batf3 deficient mice. In contrast, antibody responses were impaired by in vivo administration of the pan-caspase inhibitor Z-VAD-FMK, but normal in caspase-1 deficient mice, suggesting a contribution from apoptotic caspases, in the induction of antibody responses. These results demonstrate that squalene emulsion-based vaccine adjuvants induce antigen-specific CD8 T cell and antibody responses, through RIPK3-dependent and-independent pathways, respectively.

Introduction

The discovery of vaccine adjuvants which enhance the magnitude and durability of immune response to antigens has greatly facilitated the development of effective vaccines against diseases such as influenza, hepatitis B, malaria and shingles (Coffman et al., 2010; Levitz and Golenbock, 2012; McKee and Marrack, 2017; O'Hagan et al., 2017; Pulendran and Ahmed, 2011; Reed et al., 2013). Adjuvants comprise a variety of substances, ranging from emulsions such as aluminum salts (alum), to purified plant extracts such as QS21, to synthetic nanoparticles and small molecules that activate specific receptors in the innate immune system (Del Giudice et al., 2018; O'Hagan et al., 2017). The benchmark for adjuvant research has been alum which has been included as a safe and effective adjuvant in billions of doses of vaccines, administered to diverse populations throughout the world over the past 80 years. Alum has long remained the only adjuvant licensed for clinical use, but, after decades of slow progress, recent years have witnessed the licensure of several human vaccines containing novel adjuvants, such as the squalene emulsion-based adjuvants MF59 and AS03, the Toll-like receptor 4 (TLR4) agonist monophosphoryl lipid A (MPL) absorbed on alum (AS04), and combined immune stimulators such as QS21 and MPL (AS01) (Didierlaurent et al., 2017; Garçon and Di Pasquale, 2017; MacLeod et al., 2011). Many adjuvants can induce activation of the innate immune system, which programs the magnitude, quality and durability of the adaptive immune response (Coffman et al., 2010; Levitz and Golenbock, 2012; O'Hagan et al., 2017; Pulendran and Ahmed, 2011; Reed et al., 2013). Importantly, seminal advances in our understanding of innate immunity over the past two decades (Beutler et al., 2006; Iwasaki and Medzhitov, 2015; Satoh and Akira, 2016; Temizoz et al., 2018) have facilitated the evaluation of novel synthetic adjuvants that target innate immune receptors such as TLRs (Hanson et al., 2015; Hou et al., 2011; Kasturi et al., 2011; Kurche et al., 2012; Lynn et al., 2015; MacLeod et al., 2011; Petitdemange et al., 2019; Reed et al., 2013; Yamamoto et al., 2019).

Whilst there has been much progress in understanding the cellular and molecular mechanisms of action of synthetic adjuvants such as TLR ligands, substantial knowledge gaps exist in our understanding of the mechanisms of action of classic adjuvants such as alum and MF59. Although it has long been a prevailing view that alum was ‘immunologically inert,’ and mediated its adjuvant effects via a ‘depot effect’ of slow release of antigens, injection of alum is known to rapidly recruit various cells, including neutrophils which expel neutrophil extracellular traps (NETs) composed of chromatin (Munks et al., 2010; Walls, 1977). Moreover, DNA released in NETs has been reported to mediate the adjuvant activity of alum (Marichal et al., 2011; McKee et al., 2013), although a recent study suggests this may in part be due to contaminations in the DNA preparations (Noges et al., 2016). Furthermore, alum is known to rapidly activate NALP3 inflammasome (Eisenbarth et al., 2008; Li et al., 2007; McKee et al., 2009), although opinions vary about the relative importance of this in mediating its immunogenicity (De Gregorio et al., 2008; Eisenbarth et al., 2008; Franchi and Núñez, 2008; Li et al., 2007; McKee et al., 2009; Schmitz et al., 2003).

In the case of MF59, recent studies demonstrate that it induces a broader range of cytokines and chemokines than alum or CpG DNA, and more rapidly recruits CD11b+ inflammatory cells to the site of injection (Mosca et al., 2008). Interestingly, while the adjuvant effects of MF59 appear to be independent of the NALP3 inflammasome, mice deficient in ASC – a protein necessary for inflammasome activation – are impaired in their ability to mount antibody responses to MF59 adjuvanted proteins, arguing for an NALP3-independent role for ASC in mediating MF59 adjuvanticity (Ellebedy et al., 2011). Furthermore, although MF59 does not activate TLRs in vitro, MyD88 deficient mice were impaired in their ability to mount bactericidal antibody responses to the MF59 adjuvanted rMenB vaccine (Seubert et al., 2011). In addition, it has been demonstrated that transient ATP release by muscle is necessary for MF59-induced immune responses (Vono et al., 2013). Despite these important observations, there is a paucity of understanding about the cellular and molecular mechanisms that mediate the adjuvant effects of MF59. In the present study, we demonstrate that MF59 and its SE mimetic AddaVax can induce CD8 T cell responses through RIPK3-dependent cell signaling in LN-resident macrophages. Further, we show that RIPK3 signaling in these macrophages is essential for Batf3+ dendritic cell-dependent Ag cross-presentation. Surprisingly, while the RIPK3 pathway was critical for adjuvant-triggered CD8 T cell response, it was dispensable for antibody responses that instead relied on apoptosis and damage-associated molecular pattern (DAMP) signaling. These observations reveal new mechanistic insights about MF59 and demonstrate that cellular immunity and humoral immunity can be differentially regulated by RIPK3-dependent and -independent pathways, respectively.

Results

MF59 and AddaVax induce robust antibody and CD8 T cell responses

We compared the adjuvant effects of two types of clinically licensed vaccine adjuvants, alum and SEs (MF59 or its mimetic AddaVax (AV)) in mice (Figure 1A). We immunized and boosted C57BL/6 (B6) mice with chicken ovalbumin (Ova) adjuvanted with either alum, AV or MF59 (Figure 1—figure supplement 1A), and assessed the antibody response. The titers of Ova-specific IgG1 were comparable between the alum and SE groups, while titers of IgG2b and IgG2c (representing TH1 IgG response) were significantly elevated in SE groups (Figure 1B). Consistent with this observation, the frequency of IFN-γ-producing OVA-specific CD4 T (TH1) cells was increased in AV group (Figure 1—figure supplement 1B). Moreover, the generation of follicular helper T cells (TFH) was also enhanced in SE groups (Figure 1C), consistent with a higher frequency of germinal center (GC) B cells (Figure 1D and E). These data suggest that SE adjuvants elicit stronger IgG2b and IgG2c responses by potentiating helper TH1 and TFH cell activity and GC formation.

Figure 1 with 1 supplement see all
MF59 and AddaVax induce robust antibody and CD8 T cell responses.

(A) Formulation comparison of two SE adjuvants; AddaVax and MF59 (B) WT B6 mice were primed and boosted with Ova mixed with MF59, AV or alum at indicated time points (arrows). Levels of Ova-specific IgG1, IgG2b and IgG2c in serum were determined by ELISA. *p<0.05 (t-test). (C–J) After the prime-boost vaccination with Ova plus MF59, AV or alum, immune responses were measured at day 7 post-boost. (C) Frequency of follicular helper T cells (CXCR5+ PD1+ CD4+) *p<0.05 (t-test) (D and E) Frequency of germinal center B cells *p<0.05 (t-test). **p<0.01 (t-test) (F) Ova-specific CD8 T cell response was assessed by class I MHC-peptide tetramer staining. ***p<0.001 (t-test). (G) Granzyme B expression was measured among Ova-specific CD8 T cells. ***p<0.001 (t-test). (H, I and J) FACS plots and bar graphs displaying Ova-specific CD8 T cell responses in different organs. (K) 10 days after the prime-boost vaccination with Ova alone or Ova plus AV, WT B6 mice were challenged with Ova-expressing B16 melanoma. Graphs show size of tumor (upper panel) and survival of mice (lower panel). *p<0.05 (ANOVA). For the survival curve, p value was determined by Log-rank (Mantel-Cox) test. Data are representative of two to three independent experiments (mean and s.e.m.).

In addition to antibody production, Ova-specific CD8 T cell response in the AV group was significantly higher than in the alum group at day 7 post-boost (Figure 1F). Moreover, AV-induced Ova-specific CD8 T cells displayed enhanced expression of granzyme B, a surrogate indicator of cytotoxic function (Figure 1G). Notably, Ova-specific CD8 T cells in the SE groups could be found systemically in non-lymphoid organs, but did not accumulate in the gut (Figure 1H,I and J). The role of this cytotoxic T cell population was tested with a tumor challenge model using B16 melanoma where the tumor cell was engineered to express chicken Ova, and could be cleared by functional Ova-specific CD8 T cells. Using this system, mice vaccinated with Ova+AV were fully protected from tumor challenge, while Ova+PBS vaccinated mice were not (Figure 1K). Together, these data demonstrate that SE adjuvants can generate robust IgG responses and protective CD8 T cell responses.

MF59 and AV elicit robust innate immune responses

In order to understand the mechanisms driving the robust adjuvant activity of SE adjuvants, we analyzed the innate immune responses in the draining lymph nodes at early time points following vaccination in both alum- and SE adjuvants-immunized groups (Figure 2—figure supplement 1A). SE groups displayed a significant increase of dLN size from 24 hr post-vaccination, while no increase in cellularity could be detected in alum-immunized dLNs even at 48 hr (Figure 2A and Figure 2—figure supplement 1B). The increase of dLN cellularity was due to the rapid recruitment of monocytes and neutrophils, followed by late influx of DCs and lymphocytes including B and T cells (Figure 2B). Next, we measured the capacity of antigen (Ag) uptake using fluorochrome-conjugated Ova (Ova-AF647) in various populations. Following vaccination, total LN cells in the SE groups showed higher Ag uptake when compared to LN cells from mock or alum groups (Figure 2C and Figure 2—figure supplement 1C). Neutrophils, monocytes and DCs were mainly responsible for Ag uptake at early time points, but DCs dominated later on (Figure 2D). Additionally, we examined the activation of DCs by measuring the surface expression of co-stimulatory molecules CD80 and CD86. At 24 hr post-vaccination, migratory DCs (mDCs) and resident DCs (rDCs) in both AV and MF59 groups had significantly increased the expression of CD80 and CD86 (Figure 2E and Figure 2—figure supplement 1D; Idoyaga et al., 2013). Kinetic analysis of CD80 expression on mDCs and rDCs showed a gradual increase in AV group over time, while no significant alteration was detected in alum group until 48 hr post-vaccination (Figure 2F). Taken together, SE adjuvants induce enhanced leukocyte recruitment, increased Ag uptake, and stronger DC activation, potentially explaining the potency of subsequent adaptive immune responses (Calabro et al., 2011).

Figure 2 with 1 supplement see all
AV elicits robust innate immune responses.

WT B6 mice were vaccinated with Ova mixed with AV or alum, and innate immune responses were assessed in dLNs during the first 48 hr period. (A) Total cell numbers of inguinal LNs. (B) Absolute numbers of different immune cell populations. (C) At 12 hr post-vaccination, the uptake of antigen was measured using AF647-conjugated Ova. (D) Total numbers of Ova+ immune cell subsets were plotted. (E) Histograms show surface expression of CD80 and CD86 on migratory DCs and resident DCs. (F) Kinetic analysis of CD80 expression. Data are representative of two independent experiments (mean and s.e.m.). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (ANOVA).

LN-resident macrophages are perturbed following MF59 and AV uptake

To examine how the SE adjuvants stimulate enhanced innate immune responses, we investigated the role of SE adjuvants in dLNs at early time points. Previous studies have established SIGN-R1medullary macrophages within medullary inter-follicular regions as a primary collection point of lymph-borne antigen due to their proximal nature to the afferent lymphatics and high levels of phagocytosis (Gray and Cyster, 2012; Woodruff et al., 2014). Indeed, at 2 hr post-immunization, lipophilic dye-labeled AV could be detected primarily within the medullary region of the dLN (Figure 3A). Flow cytometry analysis further verified that medullary cord macrophages (CD169- F4/80+, MCMs) and medullary sinus macrophages (CD169+ F4/80+, MSMs) were the most abundantly represented cell types (up to 50%) within the AV+ populations, although lower frequencies of other innate immune cells including neutrophils, monocytes, DCs and subcapsular sinus macrophages (CD169+ F4/80-, SSMs) were also AV+ (Figure 3B).

Figure 3 with 1 supplement see all
LN-resident macrophages uptake SE adjuvant and are eliminated.

(A and B) WT B6 mice were vaccinated with Ova together with AV-Did, and the uptake of AV-Did was measured in dLNs by immunofluorescence at 2 hr (A) and flow cytometry at 3 hr (B). (C and D) Immunofluorescence of dLNs at 24 hr post-vaccination with PBS, alum or AV. (D) Quantification of MMs (upper) and SSMs (lower) from the acquired confocal images using CellProfiler software. *p<0.05, ***p<0.001 (ANOVA). (E–G) Kinetic changes of dLN-resident macrophage subsets were analyzed by flow cytometry. FACS plots are pre-gated on CD11b+CD11cSiglec-F- Ly6G- Ly6C- population. Frequencies (F) and absolute numbers (G) of each macrophage subset were plotted. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (ANOVA). Data are representative of two to three independent experiments (mean and s.e.m.).

Next, we wondered if the homeostasis of macrophages in dLN were affected by the uptake of SE adjuvants. To this end, the presence of medullary macrophages (MMs) and SSMs in differently immunized dLNs were assessed by confocal microscopy. At 24 hr after vaccination, PBS-injected dLNs displayed abundant MMs and a clear lining of SSMs along the subcapsular sinus. However, dLNs in the AV and MF59 groups displayed a remarkable decrease of both MM and SSM populations with an almost complete loss of the subcapsular sinus lining, while there were no such reductions in mice immunized with alum plus Ova (Figure 3C and D and Figure 3—figure supplement 1A). To dissect this phenomenon further, we performed a time course analysis on macrophage populations in dLNs from alum and AV groups during the first 2 days post-vaccination by flow cytometry. Consistent with confocal data, alum did not cause significant alteration in any of the macrophage subsets, even until 48 hr post-immunization (Figure 3E and F). However, in the AV group at 4 hr and 12 hr post-vaccination, there was a striking reduction in the frequencies of both types of medullary macrophages (MCMs and MSMs). Moreover, there was a reduction in the frequencies of SSM albeit with slower kinetics, starting at 12 hr and dropping to the lowest levels at 48 hr (Figure 3E and F). Similarly, MF59 also caused a significant reduction of LN-resident macrophages at 24 hr post-immunization (Figure 3—figure supplement 1B and C). In terms of absolute number, MSMs displayed a significant and persistent reduction, while MCMs were reduced transiently and SSMs displayed a mild reduction (Figure 3G). The discrepancy between frequencies and absolute numbers is explained by the dramatic increase of dLN cellularity after 24 hr post-immunization (Figure 2A). Collectively, these data show that the efficient phagocytosis of SE adjuvants is associated with the early loss of MMs including MCMs and MSMs, while the relatively slow accumulation of AV in SSMs is associated with the delayed and less pronounced loss.

MF59 and AV trigger sequential waves of regulated necrosis and apoptosis in LN macrophages

We sought to understand whether the loss of LN-resident macrophages was due to cell death or any other mechanism. Using bone marrow-derived macrophages (BMMs), the kinetics of in vitro cell death was assessed in the presence of AV. AV induced dynamic changes such as dominant necrosis (AnnexinV- PI+) at early time points (up to 6 hr), followed by delayed apoptosis (AnnexinV+ PI-) (Figure 4A). In order to examine if this is the case in vivo, dLNs were collected after vaccination with alum or AV, and subjected to western blot of whole LN lysate. Consistent with the in vitro cell death data, phosphorylated-MLKL (p-MLKL) and cleaved caspase 1 (markers of necroptosis and inflammasome activation/pyroptosis respectively) were visible at 3 hr, and were strongly induced at 6 hr in AV-immunized dLNs (Figure 4B). Immuno-fluorescent staining of phosphorylated RIPK3 and MLKL verified necroptosis signaling in the medullar following AV immunization (Figure 4C). Induction of MLKL phosphorylation was confirmed to be RIPK3-dependent as no positive p-MLKL signal could be observed within AV-immunized RIPK3-KO animals (Figure 4—figure supplement 1A). Meanwhile no MLKL phosphorylation and lower levels of cleaved caspase 1 were detected in alum-immunized dLNs. Furthermore, there was increased apoptosis in the dLNs of mice immunized with alum, as evidenced by enhanced cleavage of caspase 3 at 3 hr and 6 hr, while AV injection displayed a strong activation of caspase 3 at 24 hr (Figure 4B). These cell death phenotypes in dLNs were also true for MF59 vaccination (Figure 4—figure supplement 1B). Thus, although these data do not give information on which cell types are undergoing cell death, they suggest that after SE-adjuvanted immunization, there is early regulated necrosis such as MLKL-mediated necroptosis and pyroptosis, followed by delayed apoptosis in dLNs.

Figure 4 with 1 supplement see all
SE adjuvant triggers sequential waves of regulated necrosis and apoptosis in LN macrophages.

(A) Bone marrow-derived macrophages were incubated with AV for different time periods. Using co-staining of AnnexinV and Propidium Iodide, kinetic changes of necrotic and apoptotic cells were determined by flow cytometry. (B) After WT B6 mice were immunized with alum or AV, dLNs from each group were collected, and total dLN lysates were subjected to western blot analysis at indicated time points. (C) Immunofluorescence of dLNs at indicated time points to detect necroptosis signaling. (D) Western blot analysis with dLN lysates from control and CLL-treated WT B6 mice. (E) Immunofluorescence of dLNs at 0 hr, 0.5 hr and 2 hr post-vaccination. (F) Electron micrographs displaying AV-vaccinated dLNs at different time points. Red arrows represent cells with necrotic phenotype. (G) Serum levels of different cytokines were assessed by ELISA. *p<0.05, ****p<0.0001 (ANOVA). (H) At 24 hr post-immunization, the release of dsDNA and uric acid was determined in serum. **p<0.01, ***p<0.001 (t-test). Data are representative of two independent experiments (mean and s.e.m.).

Clodronate-loaded liposomes (CLL) are known to induce selective apoptosis in macrophages (van Rooijen and Hendrikx, 2010). Consistent with the literature, subcutaneous injection of CLL resulted in efficient depletion of MSM and SSM subsets in dLNs (Figure 4—figure supplement 1C). Interestingly, following macrophage depletion, Ova+AV vaccination resulted in significantly decreased MLKL phosphorylation and altered kinetics of caspase1 cleavage (Figure 4D). These data, combined with the observed loss of the LN-resident macrophage populations by both flow cytometry and immunofluorescence (Figure 3C–G) suggests that LN-resident macrophages contribute substantially to the observed necroptosis in dLNs after AV vaccination. In addition to putative LN-resident macrophages undergoing necroptosis, MMs displaying cleaved caspase 3 were also detected after the AV vaccination (Figure 4E). Morphologically, necrotic cell death is characterized by organelle swelling, membrane disruption, and cytoplasmic release as opposed to the chromatin condensation and membrane blebbing phenotype typical of apoptosis. In contrast to the minimal level of dying cells in dLNs of naïve mice, necrotic cells (red arrows), including necrotic monocytes/macrophages (1 hr) and a necrotic macrophage (4 hr) were observed predominantly at the medulla of dLNs after 1 hr of AV injection (Figure 4F). Together these data consistently support the idea that LN-resident macrophages undergo early necrosis and delayed apoptosis upon AV immunization.

An extensive literature has outlined the role of necrosis in the generation of inflammatory responses through the secretion of pro-inflammatory cytokines and the released danger signals that can act as DAMPs (Bergsbaken et al., 2009; Blander, 2014; Mocarski et al., 2014; Pasparakis and Vandenabeele, 2015). Therefore, we asked if these signals could be detected systemically following AV vaccination. Indeed, cytokines such as IL-1β, IL-6, IL-12 and TNF were released into serum by AV vaccination (Figure 4G). IL-6, IL-12 and TNF were particularly elevated by AV vaccination in comparison with the alum group. Moreover, elevated levels of double-stranded DNA (dsDNA) and uric acid, well described DAMPs could be increasingly detected in serum of AV-immunized mice at 24 hr (Figure 4H; Ko et al., 2016). Collectively, these data indicate that the subcutaneous injection of SE adjuvants can induce mixed types of cell death signaling pathways which in turn set up a highly immuno-stimulatory microenvironment.

Macrophages and Batf3+ DCs cooperatively stimulate CD8 T cell response upon SE-adjuvanted immunization

To determine the importance of macrophage perturbation in the induction of innate and adaptive immune responses, both pharmacological and genetic approaches were used to deplete LN-resident macrophages. Following macrophage depletion by CLL treatment, Ova+AV vaccination resulted in reduced activation of the mDC subset, compared to control liposome treatment at 24 hr (Figure 5—figure supplement 1A). Production of IL-6 was also dramatically attenuated (Figure 5—figure supplement 1B). Importantly, Ova-specific CD8 T cell responses were significantly decreased in the CLL-treated group at day 7 post-boost (Figure 5A).

Figure 5 with 1 supplement see all
Macrophages and Batf3+ DCs cooperatively stimulate CD8 T cell response.

(A) WT B6 mice were injected with control or clodronate-loaded liposomes. 5 days later, mice were subsequently immunized with Ova and AV. Ova-specific CD8 T cell response at day 7 post-boost in different tissues. *p<0.05, **p<0.01 (t-test). (B and C) Mice got intraperitoneal injection with 400 ng of diphtheria toxin per mouse, and two days later, they were subsequently immunized by Ova plus AV. (B) The depletion of LN-resident macrophages was measured in CD169-DTR mice two days after the DT injection. *p<0.05 (t-test). (C) Ova-specific CD8 T cell responses in WT B6 and CD169-DTR mice at day 7 post-boost. *p<0.05 (t-test). (D) Ova-specific CD8 T cell response in WT B6 and Batf3 KO mice. *p<0.05, **p<0.01 (t-test). Data are representative of two independent experiments (mean and s.e.m.).

In order to more specifically dissect the effect of macrophage population on CD8 T cell responses, we used two different genetic models of macrophage depletion: CD169-DTR and LysM-iDTR mice. Upon intraperitoneal DT injection, CD169-DTR mice showed efficient depletion in MSM and SSM subsets, while LysM-iDTR mice displayed only SSM depletion (Figure 5B and Figure 5—figure supplement 1C; Gupta et al., 2016; Shaabani et al., 2016). Surprisingly, in contrast to normal CD8 T cell responses in the LysM-iDTR group (Figure 5—figure supplement 1D), the CD169-DTR group showed significant reduction of CD8 T cell responses in spleen and lung (Figure 5C). These data demonstrate that LN macrophages are required for the efficient activation of innate immune response and the optimal induction of CD8 T cell response by SE adjuvants. Of note, the recent description of a conventional DC2 (cDC2, CD11b+ CD11c+) population opens the possibility of an additional actor in this response pathway (Ciavarra et al., 2005). While lymph node macrophages have been directly tested here, it is unclear if the cDC2 population might also play a role as their function has not yet been fully validated across these various model systems.

With respect to the cell types involved in Ag cross-presentation to CD8 T cells, it is known that CD8a+ DCs are the key mediators of this process (den Haan et al., 2000; Hildner et al., 2008). It has been reported that Batf3+ DCs such as Xcr1+ CD103+ DCs essential for cross-presentation of skin-derived Ags to CD8 T cells (Bedoui et al., 2009). In contrast, a recent study suggested that CD169+ macrophages are capable of Ag cross-presentation to CD8 T cells without DCs in specific experimental settings (Bernhard et al., 2015). To investigate which cell type is responsible for Ag cross-presentation to CD8 T cells in the context of the current study, we asked if Batf3+ DCs were required for the AV-mediated CD8 T cell responses using Batf3-deficient mice. In the absence of Batf3+ DC populations (Figure 5—figure supplement 1E), the Ova-specific CD8 T cell response was substantially impaired (Figure 5D). This implies that cross-presentation by Batf3+ DCs is critical for the generation of CD8 T cell responses in SE-adjuvanted vaccination. Taken together, these data demonstrate that LN macrophages and Batf3-dependent CD103+ DCs are necessary for optimal induction of antigen-specific CD8 T cell response to SE adjuvants.

Kinase activity of RIPK3 in macrophages is critical for the induction of CD8 T cell response by SE adjuvants

We sought to understand the molecular pathways governing the CD8 T cell response in SE-adjuvanted vaccination model. To this end, TLR signaling pathways were investigated utilizing different KO mice. Surprisingly, neither individual TLR deficiency (TLR3, TLR4, TLR7, and TLR9) nor the absence of adaptor proteins such as MyD88 and TRIF affected AV-induced CD8 T cell response. (Figure 6—figure supplement 1A and B). Since SE adjuvants induce RIPK3-dependent macrophage death (Figure 4D and Figure 4—figure supplement 1A), and dLN-resident macrophages are necessary for CD8 T cell responses (Figure 5A-C), we investigated whether different types of cell death pathways are involved in this process. To probe the role of inflammasomes and pyroptosis, we examined the CD8 T cell response in ASC KO and Caspase 1 KO mice and observed that the CD8 T cell response was normal in these mice (Figure 6—figure supplement 1C). In addition, in vivo inhibition of pan-caspase activity by z-VAD-fmk did not influence the normal CD8 T cell response (Figure 6—figure supplement 1D), suggesting that neither inflammasome activation/pyroptosis nor apoptosis is crucial for AV-induced CD8 T cell responses. A previous report by Yatim et al. demonstrated that an injection of necroptotic cells rather than apoptotic cells generated potent CD8 T cell response (Yatim et al., 2015). To elucidate the role of RIPK3 signaling in SE-adjuvanted vaccination, we assessed immune responses in RIPK3 KO mice upon Ova+AV vaccination. At 24 hr after vaccination, RIPK3 KO mice displayed significantly impaired secretion of IL-6 and TNF in serum (Figure 6A). Remarkably, in the absence of RIPK3, CD8 T cell response in various tissues was significantly reduced by AV vaccination, as compared to WT mice (Figure 6B). Consistent with AV data, vaccination with MF59 in RIPK3 KO mice resulted in a significant decrease of CD8 T cells (Figure 6—figure supplement 2A).

Figure 6 with 2 supplements see all
RIPK3 kinase-dependent signaling is critical for the induction of CD8 T cell response.

(A) WT B6 and RIPK3 KO mice were primed with Ova and AV, and serum cytokine levels were quantified at 24 hr. (B) At day 7 post-boost, Ova-specific CD8 T cell response was determined from different tissues in WT B6 and RIPK3 KO mice. *p<0.05, ***p<0.001 (t-test). (C) Ova-specific CD8 T cell responses in WT B6, RIPK3 KO and RIPK3 K51A kinase-dead mice. Data were pooled from two independent experiments. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 (t-test). (D) Ova-specific CD8 T cell responses in WT B6 and MLKL KO mice. (E) Histograms represent OT-I cell proliferation upon Ag cross-presentation (1h- or 3h-incubation of BMMs with AV). Gray, blue and red colors represent control, WT B6 BMM and RIPK3 KO BMM groups, respectively. Data are representative of two to three independent experiments (mean and s.e.m.).

Recent studies have proposed multiple roles for the RIPK3 protein – the kinase-dependent functions including the execution of necroptosis and induction of inflammatory cytokines, and kinase-independent scaffolding functions (Kang et al., 2013; Lawlor et al., 2015; Muendlein et al., 2020; Najjar et al., 2016; Zhu et al., 2018). To further dissect the role of kinase activity of RIPK3 in the SE-adjuvanted immunization, we utilized a RIPK3 kinase-inactive knock-in mutant (Ripk3K51A/K51A) mouse model. RIPK3 K51A mice were also significantly impaired in their capacity to induce Ova-specific CD8 T cell responses in multiple tissues, albeit to a lesser degree than RIPK3 KO mice (Figure 6C), suggesting that the kinase-dependent activity of RIPK3 is necessary for induction of CD8 T cell response by SE adjuvants. Next, we examined whether the execution of necroptosis is essential for the AV-induced CD8 T cell response. Notably, deficiency of MLKL, an effector molecule of necroptosis, did not affect Ova-specific CD8 T cell response after Ova+AV vaccination (Figure 6D), implying an alternate pathway for RIPK3 function in this system outside of the classically described execution of necroptosis.

To directly address if RIPK3-dependent signaling in macrophages is necessary for the induction of CD8 T cell response, we designed an in vitro cross-presentation assay in which bone marrow-derived macrophages (BMMs) incubated with Ova and AV were subsequently co-cultured with DCs, followed by the addition of CFSE-labeled OT-I cells (Figure 6—figure supplement 2B). Strikingly, BMMs from the RIPK3 KO mouse resulted in the significantly decreased proliferation of OT-I cells, as compared to WT BMMs (Figure 6E). This result was observed regardless of different duration (1 hr and 3 hr) of SE adjuvant incubation with BMMs. Therefore, these data demonstrate that RIPK3-dependent signaling in dLN-resident macrophages is necessary for optimal CD8 T cell responses to SE-adjuvanted immunization.

IgG responses by SE adjuvants are not dependent on the RIPK3 signaling pathway

Vaccine-induced antibody responses are essential for protective immunity. In addition to the previously described CD8 T cell response here, SE adjuvants are known to elicit strong humoral immune responses, consistent with the results observed in the current study (Figure 1B). In order to understand how these adjuvants trigger potent antibody responses, we determined if dLN-resident macrophages are also associated with AV-induced IgG responses. Consistent with the role of macrophages in CD8 T cell responses, the depletion of both SSM and MSM in CD169-DTR mice resulted in the significant reduction of Ova-specific IgG responses (Figure 7A). In contrast, there were normal IgG responses in CLL-treated mice and LysM-iDTR mice (Data not shown and Figure 7—figure supplement 1A, respectively). This phenotypic discrepancy between CD8 T cell response and IgG response in CLL-treated mice could be due to CLL’s direct adjuvant effect on B cells (Tonti et al., 2013). Furthermore, the absence of Batf3+ DCs did not alter IgG responses (Figure 7—figure supplement 1B).

Figure 7 with 1 supplement see all
Caspase activity and DAMPs promote IgG responses.

Levels of IgG1 and IgG2b were assessed in serum samples at day 7 post-boost. (A) IgG levels in WT B6 and CD169-DTR mice at day 7 post-boost. (B) WT B6 mice were immunized with AV in the presence or absence of Nec-1s or z-VAD-fmk. (C) IgG levels in WT B6 and RIPK3 KO mice. (D) WT B6 mice were primed and boosted with AV in the presence or absence of DNaseI. (E) A working model describing distinctive mechanisms governing the IgG and CD8 T cell responses induced by the SE adjuvant. *p<0.05, ***p<0.001 (t-test). Data are representative of two to three independent experiments (mean and s.e.m.).

Next, we explored the effect of different types of cell death on IgG responses. In contrast to the CD8 T cell response, in vivo treatment with the pan-caspase inhibitor, z-VAD-fmk, resulted in a profound decrease in antigen-specific IgG1 titers induced by AV (Figure 7B). However, the inhibition of RIPK1 by Nec-1s injection or RIPK3 deficiency had no effect on IgG responses in AV vaccination (Figure 7B and C). In addition, impaired inflammasome formation in ASC KO and caspase 1 KO mice resulted in normal IgG responses (Figure 7—figure supplement 1C). This suggests that caspase-dependent apoptosis, rather than a RIPK3-dependent pathway might be primarily important for antibody responses induced by AV vaccination. Intriguingly, previous studies have reported that the modulation of DAMPs such as dsDNA in various vaccination models affects humoral immune response (Marichal et al., 2011). In the AV-adjuvanted vaccination model, the efficient generation of IgG responses was also associated with the release of dsDNA since the co-injection of DNaseI significantly decreased the IgG responses by AV (Figure 7D). In alignments with a previous report (Seubert et al., 2011), the MyD88 pathway was required for the generation of optimal IgG responses by SE adjuvant although some of individual TLRs are moderately associated (Figure 7—figure supplement 1D and E). Altogether, these data suggest that the caspase activity in macrophages and subsequent release of DAMPs might play important roles in the SE adjuvant-mediated IgG responses.

Discussion

Recently cellular stress and damage have become increasingly recognized as potent drivers of inflammation and the adaptive immune response (Bettigole and Glimcher, 2015; Blander, 2014; Chovatiya and Medzhitov, 2014; Matzinger, 1994; Osorio et al., 2014; Pulendran, 2015; Ravindran et al., 2014; Zhang and Kaufman, 2008). It has also become clear that cell death and associated signaling pathway can modulate immune outcomes (Bergsbaken et al., 2009; Blander, 2014; Chan et al., 2015; Murakami et al., 2014; Pasparakis and Vandenabeele, 2015). For example, apoptosis in the absence of foreign antigen (Ag) is generally considered as immunologically silent, while regulated necrosis such as RIPK3-dependent necroptosis and pyroptosis results in the release of pro-inflammatory cytokines and DAMPs (Kim et al., 2019; Kolb et al., 2017; Berghe et al., 2014). Injection of necroptotic cells rather than apoptotic cells promoted potent CD8 T cell responses in the presence of foreign Ag although it is noteworthy that death-associated signaling rather than necroptosis itself was critical (Yatim et al., 2015). This ‘sterile inflammation’ concept has revitalized the autoimmunity and oncology fields, but little is yet known about the role of these pathways in regulating the immune system in the context of vaccine response.

In this study, we present evidence that induction of antigen-specific CD8 T cell responses in mice immunized with MF59 or its mimetic AV occurs through a mechanism dependent on RIPK3 signaling in LN macrophages. Using several independent models, we demonstrate that the SE adjuvants are primarily acquired by LN-resident macrophages, resulting in their elimination through various cell death pathways including apoptosis and regulated necrosis. These infection/vaccination-mediated damage response doesn’t seem to be rare event because recent studies have demonstrated that Staphylococcus aureus infection, Modified Vaccinia Ankara (MVA) vaccination or the saponin based adjuvant caused the disruption of SSM layer in dLNs, which in turn significantly modulate subsequent B cell or T cell responses, respectively (Detienne et al., 2016; Gaya et al., 2015; Sagoo et al., 2016). In this SE adjuvant vaccination model, it is worth noting that the MSM subset, rather than SSMs, is critical for the uptake of the SE adjuvant, inducing cell death-associated signaling, and ultimately required for optimal adaptive immune responses. Importantly, we demonstrate the induction of optimal antigen-specific antibody responses by SE adjuvants is associated with caspase activity and dsDNA release (Figure 7E). Thus, while RIPK3-dependent signaling is critical for optimal CD8 T cell response, it was not necessary for IgG responses.

RIPK3 is known to be essential in the induction of MLKL-dependent necroptosis, but recent studies have established additional roles for RIPK3 in kinase-dependent cytokine induction and kinase-independent scaffolding function (Kang et al., 2013; Lawlor et al., 2015; Muendlein et al., 2020; Najjar et al., 2016; Zhu et al., 2018). We demonstrate that the kinase activity of RIPK3 was critical for SE adjuvant-mediated CD8 T cell response since RIPK3 kinase-dead (RIPK3 K51A) mice abrogated the response. Further investigation using MLKL-deficient mice has revealed the dispensability of the execution step of necroptosis in eliciting CD8 T cell response although RIPK3’s kinase activity is still required. Thus we propose that the increased secretion of proinflammatory cytokines by RIPK3-associated signaling is essential for the optimal CD8 T cell response to SE-adjuvanted immunization (Snyder et al., 2019; Yatim et al., 2015). In contrast, our data suggest that caspase activity is necessary for the efficient induction of IgG response by SE adjuvants. Caspase family proteins have two well-known functions, as triggers of apoptosis and in the activation of inflammasomes, depending on the caspase family member. The redundant role of the canonical inflammasome pathway in SE adjuvant-mediated IgG responses was demonstrated using caspase 1 and ASC deficient mouse models, raising the possibility of additional caspases’ roles in the AV-induced antibody response such as apoptosis induction or previously unappreciated function.

Our data also demonstrate that the SE-adjuvanted immunization establishes a highly immuno-stimulatory environment. We observed increased secretion of pro-inflammatory cytokines and DAMPs such as IL-6, IL-12 and TNF, and dsDNA by the SE adjuvant. It is possible that the release of pro-inflammatory mediators provides a feed-forward mechanism in the further elimination of macrophages and the efficient activation of DCs.

In conclusion, our results demonstrate that the CD8 T cell response, by MF59 and its analog AV, can be triggered via RIPK3-dependent signaling in dLN macrophages, whilst antibody responses occur independently of this pathway. These results suggest that pharmacological or genetic manipulation of these pathways may provide novel mediators of vaccine immunity.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Genetic reagent
(M. musculus)
C57BL/6JThe Jackson LaboratoryStock No: 000664
Genetic reagent
(M. musculus)
Batf3 KOThe Jackson LaboratoryStock No: 013755
Genetic reagent
(M. musculus)
Caspase 1 KOThe Jackson LaboratoryStock No: 016621
Genetic reagent
(M. musculus)
TLR3 KOThe Jackson LaboratoryStock No: 005217
Genetic reagent
(M. musculus)
TRIF KOThe Jackson LaboratoryStock No: 005037
Genetic reagent
(M. musculus)
RIPK3 KOPMID:14749364
Genetic reagent
(M. musculus)
RIPK3K51A/K51APMID:25459880
Genetic reagent
(M. musculus)
CD169-DTRPMID:23601688
Genetic reagent
(M. musculus)
LysM-iDTRPMID:20176743
Genetic reagent
(M. musculus)
ASC KOPMID:15190255
Genetic reagent
(M. musculus)
MLKL KOPMID:24012422
Genetic reagent
(M. musculus)
TLR4 KOPMID:16461338
Genetic reagent
(M. musculus)
TLR7 KOPMID:16461338
Genetic reagent
(M. musculus)
TLR9 KOPMID:20962088
Genetic reagent
(M. musculus)
MyD88 KOPMID:16461338
AntibodyCD11c
(Hamster monoclonal)
BioLegend117330FCM 1:400
AntibodyCD169
(Rat monoclonal)
BioLegend142413FCM 1:400
AntibodyCD19
(Rat monoclonal)
BioLegend115541FCM 1:800
AntibodyTCR-b
(Hamster monoclonal)
BioLegend109224FCM 1:400
AntibodyCD4
(Rat monoclonal)
BioLegend100557FCM 1:800
AntibodyCD44
(Rat monoclonal)
BioLegend103047FCM 1:800
AntibodyCD45R (B220)
(Rat monoclonal)
BioLegend103244
103229
FCM 1:800
IF 1:400
AntibodyCD80
(Hamster monoclonal)
BioLegend104729FCM 1:800
AntibodyCD86
(Rat monoclonal)
BioLegend105006FCM 1:800
AntibodyCD8a
(Rat monoclonal)
BioLegend100750FCM 1:800
AntibodyIFN-g
(Rat monoclonal)
BioLegend505806FCM 1:800
AntibodyLy6G
(Rat monoclonal)
BioLegend127624FCM 1:800
AntibodyRabbit IgG AF647
(Donkey polyclonal)
BioLegend406421IF 1:400
AntibodyFc block
(Rat monoclonal)
BD Biosciences553142FCM 1:800
AntibodyCD103
(Rat monoclonal)
BD Biosciences564322FCM 1:400
AntibodyCD11b
(Rat monoclonal)
BD Biosciences563402FCM 1:800
AntibodyCD279 (PD-1)
(Hamster monoclonal)
BD Biosciences563059FCM 1:200
AntibodyCD45
(Rat monoclonal)
BD Biosciences563053FCM 1:800
AntibodyCD95
(Hamster monoclonal)
BD Biosciences563646FCM 1:800
AntibodyCXCR5
(Rat monoclonal)
BD Biosciences551961FCM 1:100
AntibodyGL7
(Rat monoclonal)
BD Biosciences562967FCM 1:400
AntibodyF4/80
(Rat monoclonal)
eBioscience25-4801-82FCM 1:400
AntibodyLy6C
(Rat monoclonal)
eBioscience45-5932-82FCM 1:800
AntibodyMHC II (I-A/I-E)
(Mouse monoclonal)
eBioscience56-5321-82FCM 1:800
AntibodyCD209b (SIGN-R1)
(Hamster monoclonal)
eBioscience16-2093-82IF 1:200
AntibodyAnti-Rabbit IgG
(Goat polyclonal)
Life TechnologiesA-11070FCM 1:1000
AntibodyGranzyme B
(Mouse monoclonal)
InvitrogenGRB05FCM 5 μl
Antibodyp-RIP3 (T231/S232)
(Rabbit monoclonal)
Cell Signaling Technology91702IF 1:200
Antibodyp-MLKL (S345)
(Rabbit monoclonal)
Cell Signaling Technology37333IF 1:200
AntibodyCleaved caspase-3
(Rabbit polyclonal)
Cell Signaling Technology9661IF 1:400
AntibodyMLKL
(Rabbit
monoclonal)
Cell Signaling Technology37705WB 1:1000
AntibodyCleaved caspase-1
(Rabbit polyclonal)
Santa Cruz Biotechnologysc-514WB 1:500
AntibodyBeta-actin
(Rabbit polyclonal)
Cell Signaling Technology4967WB 1:1000
Antibodyp-MLKL (S345)
(Rabbit monoclonal)
Abcamab196436WB 1:1000
Antibodymouse IgG1-HRP
(Goat polyclonal)
Southern Biotech1070–05ELISA 1:5000
Antibodymouse IgG2b-HRP
(Goat polyclonal)
Southern Biotech1090–05ELISA 1:5000
Antibodymouse IgG2c-HRP
(Goat polyclonal)
Southern Biotech1079–05ELISA 1:5000
Peptide, recombinant proteinK(b)/Ova.SIINFEKL tetramerNIH Tetramer Core FacilityFCM 1:400
Peptide, recombinant proteinEndoGrade OvalbuminHyglos321001For injection
Peptide, recombinant proteinOva-AF647Life TechnologiesO34784
Peptide, recombinant proteinAlbumin, chicken egg white (Ovalbumin)Sigma-AldrichA2512For ELISA
Peptide, recombinant proteinUricase from Candida sp.Sigma-AldrichU0880-250UN
Peptide, recombinant proteinDNase IRoche10104159001
Peptide, recombinant proteinMurine GM-CSFPeproTech315–03
Peptide, recombinant proteinMurine M-CSFPeproTech315–02
Peptide, recombinant proteinCollagenase IVWorthington BiochemicalLS004189
Peptide, recombinant proteinStreptavidin-PEBD Biosciences554061
Chemical compound, drugOptEIA TMB SubstrateBD Biosciences555214
Peptide, recombinant proteinAnnexin VBiolegend640906FCM 5 μl
Chemical compound, drugMF59Novartis
Chemical compound, drugAddaVaxInvivoGenvac-adx-10
Chemical compound, drugAlhydogelInvivoGenvac-alu-250
Chemical compound, drugz-VAD-fmkCayman Chemical14463
Chemical compound, drugNec-1sBioVision2263–5
Chemical compound, drugLive/Dead Aqua stain kitInvitrogenL34957FCM 1:1000
Chemical compound, drugBlotting-grade blockerBio-Rad1706404
Chemical compound, drugProtease/phosphatase inhibitor cocktailCell Signaling Technology5872
Chemical compound, drugSuperSignalWest Femto/Dura chemiluminescent substrateThermo Scientific34094/34075
Commercial assay or kitMouse IL-1b ELISA kitBD Biosciences559603
Commercial assay or kitMouse IL-6 ELISA kitBD Biosciences555240
Commercial assay or kitMouse IL-12 (p70) ELISA SetBD Biosciences555256
Commercial assay or kitMouse TNF ELISA Set IIBD Biosciences558534
Commercial assay or kitQuanti-iT PicoGreen dsDNA assay kitInvitrogenP11496
Commercial assay or kitUric Acid Assay kitAbcamab65344
Commercial assay or kitStandard macrophage depletion kitEncapsula NanoSciences8901Clodronate liposome
Software, algorithmFlowJoBD
Software, algorithmPrismGraphPad
Software, algorithmCellProfilerBroad Institute

Mice and immunization

Request a detailed protocol

C57BL/6, Batf3 KO and Caspase 1 KO mice were purchased from Jackson Laboratories. RIPK3 KO (Ripk3-/-) (Newton et al., 2004) and RIPK3 kinase-inactive (Ripk3K51A/K51A) (Mandal et al., 2014) were kindly provided by E. Mocarski (Emory University). Tissues from immunized CD169-DTR and LysM-iDTR mice were kindly provided by M. Merad (Icahn School of Medicine at Mount Sinai). ASC KO mice and MLKL KO mice were originally obtained from V. M. Dixit (Genentech) and bred under specific pathogen-free conditions. Mice were immunized subcutaneously at the base of tail with endotoxin-free Ovalbumin (Hyglos) and following adjuvants: MF59 (Novartis), Addavax (Invivogen) and Alum (Alhydrogel; Invivogen). In case of co-administration of cell death inhibitors such as z-VAD-fmk (500 μg/mouse) and Nec-1s (100 μg/mouse) or enzymes including DNaseI (2,000U/mouse), those reagents were mixed with vaccine inoculums right before the subcutaneous injections. In order to deplete LN-resident macrophages, 0.25 mg (50 μl) of Clodrosome (CLL) and Encapsome (control liposome) were subcutaneously injected 5 days prior to each vaccination. Mice were maintained under specific-pathogen-free conditions in the vivarium of Emory Vaccine Center. All animal studies were conducted by following animal protocols reviewed and approved by the Institutional Animal Care and Use Committee of Emory University.

Antibodies and flow cytometry

Request a detailed protocol

Cell suspensions from PBMC, dLN, spleen, lung, liver and small intestine were stained with following fluorochrome-labeled antibodies. The following antibodies were obtained from BioLegend: CD11c (N418), CD169 (SER-4), CD19 (6D5), CD3 (17A2), CD317 (PDCA-1; 927), CD4 (RM4-5), CD44 (IM7), B220 (RA3-6B2), CD80 (16-10A1), CD86 (GL1), CD8 (53–6.7), IFN-g (XMG1.2), IL-5 (TRFK5), Ly6G (Gr-1; 1A8), TCR-b (H57-597), XCR1 (ZET). The following antibodies were obtained from BD Biosciences: CD103 (M290), CD11b (M1/70), CD279 (PD-1; J43), CD45 (30-F11), CD62L (MEL-14), CD95 (Fas; Jo2), CXCR5 (2G8), GL7 (GL7), KLRG1 (2F1). The following antibodies were obtained from eBioscience: F4/80 (BM8), Ly6C (HK1.4), MHC II (I-A/I-E; M5/114.15.2), pro-IL-1b (NJTEN3), CD209b (SIGN-R1; 22D1). Other reagents were Annexin V and 7-AAD (Biolegend), anti-Rabbit IgG and AF647-conjugated Ova (Life Technologies), granzyme B (GB11; invitrogen) and PE-conjugated streptavidin (BD). The following reagent was obtained through the NIH Tetramer Core Facility: K(b)/Ova.SIINFEKL.

To distinguish live/dead cells, cells were stained with Aqua (Thermo) or ef780 (eBioscience) prior to staining with antibodies. Fc Block (BD Biosciences) was added. For intracellular staining, cells were incubated in 100 ul of Lyse/Fix buffer (BD) and antibodies were added in 1x Cytofix/perm buffer (BD). Samples were acquired on a BD LSR II or BD LSRFortessa, and data were analyzed using FlowJo.

Antibody ELISA

Request a detailed protocol

Nunc Maxisorp 96-well plates were coated with 10 μg/mL Ovalbumin (grade VI, SIGMA-ALDRICH), goat anti- mouse IgG (1 μg/mL) (Southern Biotech) and incubated overnight at 4°C. Plates were washed 3 times with 0.5% Tween-20 in PBS and blocked with 200 μl of 2% blotting-grade blocker (Biorad). Mouse serum samples were serially diluted in blocking buffer and incubated on blocked plates. Antigen-specific serum antibodies were detected using horseradish peroxidase (HRP)-conjugated antibodies (anti-mouse IgG, anti-mouse IgG1, anti-mouse IgG2b, anti-mouse IgG2c, and anti-mouse IgE) (Southern Biotech) at 1:5000 dilution in blocking buffer. Incubation of serum samples or antibodies was conducted at room temperature. HRP activity was detected using 100 μL of tetramethylbenzidine (TMB) substrate (BD Biosciences) and stopped using 50 μL 2N H2SO4. Developed plates were recorded using BioRad spectrophotometer at 450 nm with correction at 595 nm by subtraction.

Cytokine ELISA and DAMPs detection

Request a detailed protocol

Serum samples were collected from immunized mice. Cytokine levels were determined in serum samples by sandwich ELISA. IL-1β, IL-6, TNFα and IL-12p70 were measured using OptEIA ELISA sets (BD Biosciences). ELISAs were performed according to manufacturer’s instructions. Serum levels of dsDNA were determined by Quant-iT PicoGreen dsDNA Kit (invitrogen). Uric acid amount in serum was measured by Uric Acid Assay Kit (abcam).

Immunohistochemistry and confocal microscopy

Request a detailed protocol

Isolated inguinal LNs were fixed for 2 hr in 4% paraformaldehyde (PFA) and were saturated in 30% sucrose overnight. LNs were embedded in optimal cutting temperature (OCT) medium, then were frozen and cut into 10–20 μm thick sections. Sections were stained with fluorochrome-conjugated antibodies (SIGN-R1, CD11b, B220, CD169 and cleaved caspase 3) and were imaged on a Leica SP8 confocal microscope with integrated ‘tunable’ filter sets under identical conditions in the same imaging session. Fluorophores were selected such that each fluorophore is tied to a dedicated laser to limit crossover excitation. Sequential scanning and image-recompiling were used to reduce fluorophore bleed through, and filter sets were set not to exceed 40 nm to ensure tight control over independent fluorophore signal capture. Laser power was set independently for each fluorophore to identify potential oversaturation (preventing reliable image quantitation) or bleed-through issues into other channels. Signals deemed at-risk for bleed through were quantified to ensure signal independence. phospho-RIPK3, phospho-MLKL, Caspase 3 signal (AF488) and SIGNR-1 (AF568) were assessed using total fluorescence measurements across the image and shown to be independent. Acquired images were analyzed with CellProfiler software (Broad Institute) for the identification of individual macrophages in each section.

Electron microscopy

Request a detailed protocol

Mouse lymph node samples were fixed with 2.5% glutaraldehyde in 0.1M cacodylate buffer (pH 7.4). Samples were then washed and post-fixed with 1% osmium tetroxide in the same buffer for 1 hr. After rinsing with deionized water, samples were dehydrated through an ethanol series and then placed in 100% ethanol. Following dehydration, lymph node samples were placed in propylene oxide for 10 min before infiltration with propylene oxide and Eponate 12 resin (Ted Pella, Inc, Redding, CA) at a 1:1 ratio overnight. After additional infiltration in pure Epnonate 12 resin, lymph node samples were placed in labeled Beem capsule and polymerized in a 60°C oven. Ultrathin sections were cut at 70–80 nm thick on a Leica UltraCut S ultramicrotome (Leica Microsystems Inc, Buffalo Grove, IL). Grids with ultrathin sections were stained with 5% uranyl acetate and 2% lead citrate. Ultrathin sections were imaged on a JEOL JEM-1400 transmission electron microscope (JEOL Ltd., Tokyo, Japan) equipped with a Gatan US1000 CCD camera (Gatan, Pleasanton, CA).

Antigen Cross-presentation assay

Request a detailed protocol

For bone marrow-derived macrophages (BMMs) and DCs (BMDCs), tibiae and femurs were harvested from C57BL/6 mice or RIPK3 KO mice. Bones were washed in 70% ethanol and flushed with ice-cold HBSS through a 70 μm cell strainer. After red blood cell lysis, BM cells were pelleted and plated at a density of 5–7 × 106 bone marrow cells per 10 cm Petri dish in (RPMI complete) in the presence of M-CSF (20 ng/ml, Peprotech) or GM-CSF (20 ng/ml, Peprotech). At day 5, the BMM culture was supplemented with fresh media with M-CSF, and harvested by StemPro Accutase (Life Technologies) treatment and gentle flushing at day 7. For the BMDCs, media was replaced with fresh media supplemented with GM-CSF at day 4 and 6, and cells were harvested in the same way with BMM at day 8. BMMs were seeded in 24 well plate (tissue culture non-treated), and were cultured with different concentration of AV with Ova for indicated times. BMM cells were washed, collected and added to BMDC culture (round-bottom 96-well plate, tissue culture non-treated) for 20 hr. After that, BMDCs were thoroughly washed with PBS, and added with CFSE-labeled naïve OT-I cells. 3 days later, the proliferation of OT-I cells was assessed by flow cytometry.

Western blot

Request a detailed protocol

LNs were collected in ice-cold RIPA lysis buffer supplemented with 1x protease/phosphatase inhibitor cocktail (Cell Signaling Technology). After snap freeze and thaw by liquid nitrogen, LNs were homogenized via BioMasher II (RIP Research Products), followed by sonication. Equal amounts of protein from whole LN lysates were run on an SDS-PAGE and transferred onto nitrocellulose membranes. After blocking with 5% fat-free milk, the membranes were incubated at 4°C with the following primary antibodies: anti-mouse p-MLKL (S345; ab196436, Abcam), MLKL (28640, Cell Signaling Technology), cleaved caspase 1 (M-20, Santa Cruz), cleaved caspase 3 (9661, Cell Signaling Technology) and β-actin. The membranes were then washed and incubated with Horseradish peroxidase–conjugated secondary antibody (Cell Signaling Technology). Proteins were visualized with SuperSignal West Femto or Dura chemiluminescent substrate (Pierce). Signals were acquired and analyzed via Odyssey Fc (LI-COR).

Statistical analysis

Request a detailed protocol

All results are displayed as mean ± s.e.m. Biological replicates were used in all experiments unless stated otherwise. Statistical significance was determined by t-test, one- way ANOVA, or two-way ANOVA using Prism software (GraphPad) depending on the experimental layout. Survival analysis was determined by Log-rank (Mantel-Cox) test. Probability values of p<0.05 were considered significant and denoted by *. Where indicated, ** denotes p<0.01, ***p<0.001 and ***p<0.0001.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78

Decision letter

  1. Tadatsugu Taniguchi
    Senior Editor; Institute of Industrial Science, The University of Tokyo, Tokyo, Japan
  2. John W Schoggins
    Reviewing Editor; University of Texas Southwestern Medical Center, Dallas, United States
  3. Andrew Oberst
    Reviewer; University of Washington, United States
  4. Ross Kedl
    Reviewer; University of Colorado Denver, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Thank you for submitting your article "Squalene-based adjuvants stimulate CD8 T-cell responses, but not antibody responses, through a RIP3k necroptotic pathway" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Tadatsugu Taniguchi as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Andrew Oberst (Reviewer #1); Ross Kedl (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

This manuscript presents a role for the necroptotic death of lymph node resident macrophages as a key mechanism by which squalene-based adjuvants elicit CD8+ T cell immunity. The authors walk their way through a well-designed series of experiments to support the conclusions that: 1) squalene-based adjuvants induce both necrosis and apoptosis in the draining lymph node macrophages, 2) necroptotic medullary sinus marophages are picked up by Batf3+ dendritic cells and antigen is cross presented to CD8+ T cells, and 3) these CD8+ T cell responses are RIP3K-dependent whereas B cell responses are unaffected by necroptosis, instead relying on apoptosis of macrophages to augment antibody responses. The conclusions, which are well supported by the data, will be of high interest to the biological community.

Essential revisions:

Below are compiled comments from the reviewers, each one numbered by general topic. The reviewers have requested that additional experiments be performed to address #1-2. If possible, experiments to address #3 are highly desirable, unless the authors have sufficient justification to decide otherwise.

1) Studies in MLKL knockouts

Reviewer 1: An emerging paradigm in RIPK3 signaling is that cell death is not the only, or even most important, immunological output of this pathway. The paper by Yatim et al. cited by the authors actually indicates that it is RIPK1- and NF-ΚB-dependent cytokines produced following RIPK3 activation that underlies the observed effect. Further evidence is provided by K. Newton et al. Cell Death Differ 2016. Given this, it would be of great interest to clarify whether the observed effect actually requires necroptotic cell death and "DAMP" release at all. The RIPK3 K51A mice don't really address this, as they lack both the death and the cytokine outputs of RIPK signaling. The simplest way to do this would be to run the vaccination model in mice lacking MLKL, the effector of lytic necroptosis. Alternatives could include using the in vitro co-culture model described by the authors in combination with NF-κB and/or MLKL inhibition. Assaying direct de novo production of inflammatory cytokines from WT or RIPK3 KO LN macrophages upon AV treatment would also be illuminating.

Reviewer 2: An additional experiment is the MLKL KO response, as MLKL is the terminal effector molecule of the necroptotic pathway… if the predictions are correct, that should have poor CD8 responses but OK Ab.

2) More direct demonstration of necroptosis in vivo.

Reviewer 1: I find the IF data presented in Figure 4D difficult to understand. The previous figures show an accumulation of cleaved Casp3 at late timepoints only, which is preceded by MLKL phosphorylation. However, this figure appress to show very early caspase-3 cleavage upon AV immunization. More to the point, robust antibodies exist to phosphorylated RIPK3 and MLKL, which work well in IF and IHC. Use of these reagents, and demonstration of in vivo necroptosis induced by AV, would much more clearly make the authors' point.

3) An in vivo readout to demonstrate the importance of RIPK3 in AddaVax-induced CD8 T cell responses.

Reviewer 1: While the assays used point to a requirement for RIPK3 in the emergence of CD8+ T cell immunity upon AV vaccination, the paper currently lacks an in vivo readout that this difference is important. A simple experiment using the systems the authors already have in place would be to repeat the B16 tumor clearance experiment in RIPK3 knockout mice. Showing that alum-based vaccination does not protect mice from this model would likewise strengthen the conclusions.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Squalene-based adjuvants stimulate CD8 T cell, but not antibody responses, through a RIPK3-dependent pathway" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Tadatsugu Taniguchi as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, we are asking editors to accept without delay manuscripts, like yours, that they judge can stand as eLife papers without additional data, even if they feel that they would make the manuscript stronger. Thus the revisions requested below only address clarity and presentation.

Summary:

In this revised manuscript, the authors have addressed most of the specific points I raised, and there is consensus that requesting additional work or major modifications at this stage would be both unfair and unnecessary.

Revisions:

Some remaining issues that would benefit from adding to the text/Discussion, or where appropriate, providing a response to reviewers are:

1) The uric acid release and the role for RIPK3 in CD8+ T cell responses may very well be true, true and unrelated. I am not convinced it is sufficiently grounded in causation to include. Especially since it is also highly released in response to Alum, it cannot contribute to understanding the specific mechanism of the AV. This fact is not addressed in the uricase treatment in Figure 6 because it is only AV immunization and no alum is included. As both alum and AV induce copious amounts of uric acid, were uricase treatment to have an impact on AVC but not alum immunized mice, this would nicely serve to implicate it in the AV-based mechanism of CD8 induction. Otherwise we are back to true true and unrelated… at least unrelated to mechanisms that delineate AV from alum.

2) Chlodronate depletes both macrophages and splenic cDC2s. (Ciavarra et al. 2005). This is worth noting. It is not clear to me what impact the various DTR mice (other than the BatF3) have on cDC2s, but it seems at least worth an acknowledgement that they may have a role that may not be able to be ruled out by the given experiments.

3) The CFSE data (6E) are still problematic if only percentages are going to be shown and not total numbers. In their response to the review, the authors stated that the number of undivided cells are different between wt and KO BMMs. However, only percentages are shown, a result totally consistent with identical numbers of undivided cells (which I expect it probably true) but differential accumulation of cells that are divided. If the data are going to be kept then total numbers need to be added and the exact impact of their KO on CD8 division clarified.

4) Again, uricase treatment in the last figure is of highly questionable value to the specific mechanism of AV as compared to alum given that both induce copious production of uric acid (Figure 4H). Did the authors do alum side by side here?… if there is a difference in the AV response, but not the alum response, after uricase treatment, then please show it. If not, then we can all rule out Uric acid as the DAMP that is specific to the AV-RIPK axis of T cell induction. This does not compromise the fact that there appears to be a role for RIPK in Cd8 responses that was previously unidentified.

5) As their MLKL data indicate that actual death by necroptosis has no impact on the immune responses, doesn't their model at the end need modification? Shouldn't it rely more on the signaling of RIPK, in either T cells or APC, than any significant role for necroptotic cell death? Wasn't that the point of the MLKL mice? Similarly, is the nec-1 data in the last figure necessary? The MLKL data are far more encompassing… is there no antibody data form the MLKL immunizations? If not I guess this is a decent proxy, but then it should be stated that this is unsurprising, given the lack of any effect in the MLKL KO. Lastly, it seems to me that the Discussion should be modified to temper the conclusion that over necroptosis is a critical component of their RIPK-dependent mechanism.

https://doi.org/10.7554/eLife.52687.sa1

Author response

Essential revisions:

Below are compiled comments from the reviewers, each one numbered by general topic. The reviewers have requested that additional experiments be performed to address #1-2. If possible, experiments to address #3 are highly desirable, unless the authors have sufficient justification to decide otherwise.

1) Studies in MLKL knockouts

Reviewer 1: An emerging paradigm in RIPK3 signaling is that cell death is not the only, or even most important, immunological output of this pathway. The paper by Yatim et al. cited by the authors actually indicates that it is RIPK1- and NF-ΚB-dependent cytokines produced following RIPK3 activation that underlies the observed effect. Further evidence is provided by K. Newton et al. Cell Death Differ 2016. Given this, it would be of great interest to clarify whether the observed effect actually requires necroptotic cell death and "DAMP" release at all. The RIPK3 K51A mice don't really address this, as they lack both the death and the cytokine outputs of RIPK signaling. The simplest way to do this would be to run the vaccination model in mice lacking MLKL, the effector of lytic necroptosis. Alternatives could include using the in vitro co-culture model described by the authors in combination with NF-κB and/or MLKL inhibition. Assaying direct de novo production of inflammatory cytokines from WT or RIPK3 KO LN macrophages upon AV treatment would also be illuminating.

Reviewer 2: An additional experiment is the MLKL KO response, as MLKL is the terminal effector molecule of the necroptotic pathway… if the predictions are correct, that should have poor CD8 responses but OK Ab.

The authors thank the reviewers for their shared perspective on the benefit of incorporating studies using the MLKL KO model into this manuscript. We agree that previous reports by Yatim et al. demonstrated that NF-ΚB-dependent cytokines (perhaps IL-6 or TNF) following RIPK3 activation induced CD8 T cell response in their experimental system. Therefore, the authors agree that the incorporation of a model targeting the downstream effector of necroptosis is of inherent interest to the manuscript regardless of the outcome. As a result, we have incorporated new vaccination data into the manuscript comparing the CD8+ T cell response in Wild type, vs MLKL KO vaccine recipients. The results of those data indicate that MLKL, an executioner of necroptosis, is not required for the SE adjuvant-induced CD8 T cell response (Figure 6D). This result is consistent with previous reports the reviewers mentioned, and the activation of RIPK3 signalling is critical for the CD8 T cell response possibly via secretion of inflammatory cytokines like IL-6 and TNF.

Meanwhile, the authors think that a certain DAMP is required for this response since in vivo blocking uric acid by uricase treatment significantly impaired the AV-induced CD8 T cell response. Mulay et al. has reported that MSU can induce RIPK3-dependent necroptosis, so it is likely that SE adjuvant-derived uric acid stimulates RIPK3 signalling, followed by the optimal induction of CD8 T cell response.

2) More direct demonstration of necroptosis in vivo.

Reviewer 1: I find the IF data presented in Figure 4D difficult to understand. The previous figures show an accumulation of cleaved Casp3 at late timepoints only, which is preceded by MLKL phosphorylation. However, this figure appress to show very early caspase-3 cleavage upon AV immunization. More to the point, robust antibodies exist to phosphorylated RIPK3 and MLKL, which work well in IF and IHC. Use of these reagents, and demonstration of in vivo necroptosis induced by AV, would much more clearly make the authors' point.

The authors appreciate the perspective that directly staining mediators of necroptosis would be a more direct test of necroptosis pathway execution at the time points indicated by other figures in the paper. Indeed, the use of the cleaved caspase-3 IF data implying cell death signalling at later time points, while supporting the idea of ongoing death in the lymph node, was ultimately out of step with the core messaging of the manuscript.

To address these concerns, we have replaced the cleaved caspase IF data with a more relevant time course showing the robust phosporylation of both RIP3K and MLKL by 6 hrs post vaccination and persisting to at least 24h (Figure 4C). These new data have replaced Figure 4D, with an additional panel included to show the lack of MLKL phosphorylation in RIPK3 KO controls to help validate the specificity of the staining.

It should be pointed out that while the necroptosis staining is identified (as expected) in the medulla of the draining inguinal lymph nodes following vaccination in line with the extensively described loss of medullary macrophages in this manuscript, there was an unexpected loss of SIGNR-1 surface staining on pMLKL and pRIPK3 + cells. To address this issue, we have incorporated additional staining parameters in an attempt to accurately identify the necroptotic cells as medullary macrophages. Through the use of CD11b, F4/80, MHCII, and CD169, we have.

3) An in vivo readout to demonstrate the importance of RIPK3 in AddaVax-induced CD8 T cell responses.

Reviewer 1: While the assays used point to a requirement for RIPK3 in the emergence of CD8+ T cell immunity upon AV vaccination, the paper currently lacks an in vivo readout that this difference is important. A simple experiment using the systems the authors already have in place would be to repeat the B16 tumor clearance experiment in RIPK3 knockout mice. Showing that alum-based vaccination does not protect mice from this model would likewise strengthen the conclusions.

The authors agree that the suggested experiments may strengthen the RIPK3-dependent mechanism for SE-induced CD8 T cell response. Actually, several additional experiments were carried out to verify if the adjuvant-induced CD8 T cell response was essential for tumor rejection. For example, Ova+alum vaccination also provided partial protection (which was weaker than Ova+AV group) to the tumor challenge possibly due to Ova-specific antibody derived ADCC effect (Data not shown). In addition, depletion of CD8 T cells by rat IgG (53-6.7) followed by Ova+AV vaccination failed to generate normal protection after the tumor challenge (Data not shown).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Revisions:

Some remaining issues that would benefit from adding to the text/Discussion, or where appropriate, providing a response to reviewers are:

1) The uric acid release and the role for RIPK3 in CD8+ T cell responses may very well be true, true and unrelated. I am not convinced it is sufficiently grounded in causation to include. Especially since it is also highly released in response to Alum, it cannot contribute to understanding the specific mechanism of the AV. This fact is not addressed in the uricase treatment in Figure 6 because it is only AV immunization and no alum is included. As both alum and AV induce copious amounts of uric acid, were uricase treatment to have an impact on AVC but not alum immunized mice, this would nicely serve to implicate it in the AV-based mechanism of CD8 induction. Otherwise we are back to true true and unrelated… at least unrelated to mechanisms that delineate AV from alum.

After combing through the presented data, the authors still believe that uric acid likely plays a role in the induction of the CD8 T cell response in the draining lymph nodes following AV administration. However, they also understand the reviewers concerns about the specificity of uric acid in the differentiation between alum and AV responses as administration of both leads to detectible uric acid release in the serum. We suspect that uric acid may serve as a general driver of responses that are polarized by the different lymph node microenvironments resulting from alum versus AV administration. We have not, however, described those other putative environmental factors in detail sufficiently to warrant inclusion within this manuscript. Therefore, In line with the reviewer’s recommendation, we have decided to remove our descriptions of uric acid as a specific factor in driving the RIPK3-dependent signaling pathway and have substantially altered both the manuscript and the figures to that effect.

2) Chlodronate depletes both macrophages and splenic cDC2s. (Ciavarra et al. 2005). This is worth noting. It is not clear to me what impact the various DTR mice (other than the BatF3) have on cDC2s, but it seems at least worth an acknowledgement that they may have a role that may not be able to be ruled out by the given experiments.

The reviewer brings up an interesting point about the new characterization of cDC2s, and the authors agree with the caveat of macrophage depletion using chlodronate-loaded liposome (CLL) method. Here, we first used CLL to generally assess the role of mostly macrophages (and possibly some other phagocytes such as monocytes and CD8- DCs). After seeing impaired phenotype in CD8 T cell response, we wanted to more specifically characterize the role of macrophage subsets in immune response using LysM-iDTR and CD169-DTR. Together, we believe that our model still strongly points to macrophages as the likely target of this pathway, but we cannot fully rule out other population’s contributions as we have not directly tested their status across our various knockout models. As a result, we have modified the text to leave open this possibility.

3) The CFSE data (6E) are still problematic if only percentages are going to be shown and not total numbers. In their response to the review, the authors stated that the number of undivided cells are different between wt and KO BMMs. However, only percentages are shown, a result totally consistent with identical numbers of undivided cells (which I expect it probably true) but differential accumulation of cells that are divided. If the data are going to be kept then total numbers need to be added and the exact impact of their KO on CD8 division clarified.

The authors appreciate the reviewer’s point on the CFSE experiment presented in Figure 6E. Looking back and reanalyzing the original data, we realized that the max values of Y-axis of histograms were normalized, and that’s why the numbers of undivided cells (max peaks) look identical. In Author response image 1 is histograms using Y axis with actual cell counts (and normalized Y-axis) and numbers of undividing and dividing cells of each sample. In addition, we have updated the Figure 6E with the corrected histograms.

Author response image 1

4) Again, uricase treatment in the last figure is of highly questionable value to the specific mechanism of AV as compared to alum given that both induce copious production of uric acid (Figure 4H). Did the authors do alum side by side here?… if there is a difference in the AV response, but not the alum response, after uricase treatment, then please show it. If not, then we can all rule out Uric acid as the DAMP that is specific to the AV-RIPK axis of T cell induction. This does not compromise the fact that there appears to be a role for RIPK in Cd8 responses that was previously unidentified.

The authors collectively responded to the comments 1 and 4. Please see our response for 1.

5) As their MLKL data indicate that actual death by necroptosis has no impact on the immune responses, doesn't their model at the end need modification? Shouldn't it rely more on the signaling of RIPK, in either T cells or APC, than any significant role for necroptotic cell death? Wasn't that the point of the MLKL mice? Similarly, is the nec-1 data in the last figure necessary? The MLKL data are far more encompassing… is there no antibody data form the MLKL immunizations? If not I guess this is a decent proxy, but then it should be stated that this is unsurprising, given the lack of any effect in the MLKL KO. Lastly, it seems to me that the Discussion should be modified to temper the conclusion that over necroptosis is a critical component of their RIPK-dependent mechanism.

The authors appreciate the reviewer’s concern about the working model. We have revised the working model figure (Figure 7E) like below and also modified sentences in the Discussion to carefully describe what this study has revealed, especially emphasizing the role of RIPK3 signaling rather than “necroptosis” itself in the generation of CD8 T cell response by Ag+SE vaccination.

Regarding the antibody data, we did not measure Ag-specific antibody response in MLKL KO mice after the Ova+AV vaccination. Since IgG levels is normal in RIPK3 KO mice, we assume that the antibody response in MLKL KO mice would also be normal.

https://doi.org/10.7554/eLife.52687.sa2

Article and author information

Author details

  1. Eui Ho Kim

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Yerkes National Primate Research Center, Emory University, Atlanta, United States
    3. Viral Immunology Laboratory, Institut Pasteur Korea, Seongnam, Republic of Korea
    Contribution
    Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing
    Contributed equally with
    Matthew C Woodruff
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4235-5622
  2. Matthew C Woodruff

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Yerkes National Primate Research Center, Emory University, Atlanta, United States
    Contribution
    Data curation, Investigation, Methodology, Writing - review and editing
    Contributed equally with
    Eui Ho Kim
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5252-7539
  3. Lilit Grigoryan

    Institute for Immunity, Transplantation and Infection, Stanford University School of Medicine, Stanford University, Stanford, United States
    Contribution
    Investigation
    Competing interests
    No competing interests declared
  4. Barbara Maier

    Department of Oncological Sciences, Tisch Cancer Institute and the Immunology Institute, Icahn School of Medicine at Mount Sinai, New York City, United States
    Contribution
    Resources, Investigation, Methodology
    Competing interests
    No competing interests declared
  5. Song Hee Lee

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Yerkes National Primate Research Center, Emory University, Atlanta, United States
    Contribution
    Resources, Investigation
    Competing interests
    No competing interests declared
  6. Pratushya Mandal

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Department of Microbiology and Immunology, Emory Vaccine Center, School of Medicine, Emory University, Atlanta, United States
    Contribution
    Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
  7. Mario Cortese

    Institute for Immunity, Transplantation and Infection, Stanford University School of Medicine, Stanford University, Stanford, United States
    Contribution
    Resources, Project administration
    Competing interests
    No competing interests declared
  8. Muktha S Natrajan

    Emory Vaccine Center, Emory University, Atlanta, United States
    Contribution
    Resources, Project administration
    Competing interests
    No competing interests declared
  9. Rajesh Ravindran

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Yerkes National Primate Research Center, Emory University, Atlanta, United States
    Contribution
    Resources, Investigation
    Competing interests
    No competing interests declared
  10. Huailiang Ma

    Institute for Immunity, Transplantation and Infection, Stanford University School of Medicine, Stanford University, Stanford, United States
    Contribution
    Resources, Investigation
    Competing interests
    No competing interests declared
  11. Miriam Merad

    Department of Oncological Sciences, Tisch Cancer Institute and the Immunology Institute, Icahn School of Medicine at Mount Sinai, New York City, United States
    Contribution
    Resources
    Competing interests
    No competing interests declared
  12. Alexander D Gitlin

    1. Department of Physiological Chemistry, Genentech, South San Francisco, United States
    2. Department of Pathology, Stanford University School of Medicine, Stanford University, Stanford, United States
    Contribution
    Investigation, Writing - review and editing
    Competing interests
    No competing interests declared
  13. Edward S Mocarski

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Department of Microbiology and Immunology, Emory Vaccine Center, School of Medicine, Emory University, Atlanta, United States
    Contribution
    Conceptualization, Resources
    Competing interests
    No competing interests declared
  14. Joshy Jacob

    1. Emory Vaccine Center, Emory University, Atlanta, United States
    2. Yerkes National Primate Research Center, Emory University, Atlanta, United States
    3. Department of Microbiology and Immunology, Emory Vaccine Center, School of Medicine, Emory University, Atlanta, United States
    Contribution
    Resources, Supervision
    Competing interests
    No competing interests declared
  15. Bali Pulendran

    1. Institute for Immunity, Transplantation and Infection, Stanford University School of Medicine, Stanford University, Stanford, United States
    2. Department of Pathology, Stanford University School of Medicine, Stanford University, Stanford, United States
    3. Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford University, Stanford, United States
    Contribution
    Conceptualization, Supervision, Funding acquisition, Project administration
    For correspondence
    bpulend@stanford.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6517-4333

Funding

National Institutes of Health (R37 DK057665)

  • Bali Pulendran

National Institutes of Health (R37 AI048638)

  • Bali Pulendran

National Institutes of Health (U19 AI057266)

  • Bali Pulendran

National Institutes of Health (U19 AI090023)

  • Bali Pulendran

Bill and Melinda Gates Foundation

  • Bali Pulendran

Soffer Fund

  • Bali Pulendran

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Derek O’Hagan at GSK for providing us with MF59. We are grateful to Meera Trisal for her technical help with the MLKL knockout experiments. We acknowledge the NIH (grants R37 DK057665, R37 AI048638, U19 AI090023, and U19 AI057266) and the Bill and Melinda Gates Foundation, and the Soffer Fund endowment for supporting this work in Bali Pulendran’s lab. In addition, this study was supported in part by the Robert P Apkarian Integrated Electron Microscopy Core (RPAIEMC), which is subsidized by the Emory College of Arts and Sciences and the Emory University School of Medicine and is one of the Emory Integrated Core Facilities. Additional support was provided by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UL1TR000454. The content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institutes of Health. The data described here were gathered on the JEOL JEM-1400 120kV TEM supported by a National Institutes of Health Grant S10 RR025679.

Ethics

Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols (#2002593) of Emory University.

Senior Editor

  1. Tadatsugu Taniguchi, Institute of Industrial Science, The University of Tokyo, Tokyo, Japan

Reviewing Editor

  1. John W Schoggins, University of Texas Southwestern Medical Center, Dallas, United States

Reviewers

  1. Andrew Oberst, University of Washington, United States
  2. Ross Kedl, University of Colorado Denver, United States

Publication history

  1. Received: October 12, 2019
  2. Accepted: June 8, 2020
  3. Accepted Manuscript published: June 9, 2020 (version 1)
  4. Version of Record published: June 24, 2020 (version 2)

Copyright

© 2020, Kim et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,252
    Page views
  • 233
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Immunology and Inflammation
    2. Microbiology and Infectious Disease
    Patrick S Mitchell et al.
    Research Article

    Bacteria of the genus Shigella cause shigellosis, a severe gastrointestinal disease that is a major cause of diarrhea-associated mortality in humans. Mice are highly resistant to Shigella and the lack of a tractable physiological model of shigellosis has impeded our understanding of this important human disease. Here we propose that the differential susceptibility of mice and humans to Shigella is due to mouse-specific activation of the NAIP–NLRC4 inflammasome. We find that NAIP–NLRC4-deficient mice are highly susceptible to oral Shigella infection and recapitulate the clinical features of human shigellosis. Although inflammasomes are generally thought to promote Shigella pathogenesis, we instead demonstrate that intestinal epithelial cell (IEC)-specific NAIP–NLRC4 activity is sufficient to protect mice from shigellosis. In addition to describing a new mouse model of shigellosis, our results suggest that the lack of an inflammasome response in IECs may help explain the susceptibility of humans to shigellosis.

    1. Immunology and Inflammation
    2. Stem Cells and Regenerative Medicine
    Laura Alonso-Herranz et al.
    Research Article

    Macrophages (Mφs) produce factors that participate in cardiac repair and remodeling after myocardial infarction (MI); however, how these factors crosstalk with other cell types mediating repair is not fully understood. Here, we demonstrated that cardiac Mφs increased expression of Mmp14 (MT1-MMP) 7 days post-MI. We selectively inactivated the Mmp14 gene in Mφs using a genetic strategy (Mmp14f/f:Lyz2-Cre). This conditional KO (MAC-Mmp14 KO) resulted in attenuated post-MI cardiac dysfunction, reduced fibrosis, and preserved cardiac capillary network. Mechanistically, we showed that MT1-MMP activates latent TGFβ1 in Mφs, leading to paracrine SMAD2-mediated signaling in endothelial cells (ECs) and endothelial-to-mesenchymal transition (EndMT). Post-MI MAC-Mmp14 KO hearts contained fewer cells undergoing EndMT than their wild-type counterparts, and Mmp14-deficient Mφs showed a reduced ability to induce EndMT in co-cultures with ECs. Our results indicate the contribution of EndMT to cardiac fibrosis and adverse remodeling post-MI and identify Mφ MT1-MMP as a key regulator of this process.