Introduction

Phosphatidylinositol 3-kinases (PI3K) are a family of lipid kinases that phosphorylate phosphatidylinositides (PtdIns) at the 3′-hydroxyl group¹. Upon activation, PI3K phosphorylates phosphatidylinositol 4,5-bisphosphate (PIP₂) to generate phosphatidylinositol 3,4,5-trisphosphate (PIP₃). PIP3 serves as a second messenger that recruits proteins containing pleckstrin homology (PH) domains, such as Akt (also known as protein kinase B)^{2, 3}. This activation of PI3K regulate various cellular functions, including cell proliferation, growth, survival, motility, inflammation, and metabolism, among others^{1, 4}. In macrophages, the activation of PI3K-Akt signalling is crucial to restrict inflammation and to promote anti-inflammatory responses in Toll Like Receptors (TLR)-induced macrophages, contributing to macrophage polarization^{5, 6}.

PI3Ks are heterodimeric lipid kinases composed of catalytic and adaptor/regulatory subunits that can be categorized into three classes based on their structures and substrate specificities^3,7. Class I catalytic isoforms, including p110α, p110β, p110γ, and p110δ, play essential roles in integrating signals from growth factors, cytokines, and other environmental cues. While p110α and p110β are ubiquitously expressed, p110δ and p110γ are largely restricted to the myeloid and lymphoid lineages^7–9. While p110α is primarily associated with cell growth regulation and survival in epithelial cells, it must be considered that this isoform is also expressed in immune cells, including macrophages. The role played by p110α in macrophages is not well understood, although some studies have suggested that it might regulate the survival and the regulation of the phagocytic activity of macrophages¹⁰. In the context of cancer, impairment of RAS binding to p110α in macrophages results in reduced recruitment of macrophages to the tumour site ^{11, 12}. Additionally, this disruption leads to a change in macrophage polarization, favouring a more proinflammatory M1 state ¹². These findings suggest that p110α plays a crucial role in regulating macrophage-dependent functions. However, despite these insights, the precise impact of p110α on macrophage function and the underlying molecular mechanisms influencing the inflammatory response are not yet fully understood.

Inflammation is a complex and tightly regulated series of events triggered by various stimuli such as pathogens, harmful mechanical and chemical agents, and autoimmune reactions. The inflammatory response primarily occurs in vascularized connective tissues, involving a dynamic interplay of plasma components, circulating cells, blood vessels, and cellular and extracellular factors^13–17. During inflammation, mediators released by recruited leukocytes orchestrate a response that aims to facilitate tissue repair and protect the body against harmful stimuli^16–18.

Macrophages play a vital role in the inflammatory response by performing functions such as antigen presentation, phagocytosis, and immunomodulation^{16, 17}. Their role begin with the active recruitment of monocytes from the bloodstream to the site of infections¹⁹ where they differentiate into macrophages and recognize microbes and cellular debris through specific mechanisms. Macrophages subsequently actively participate in phagocytosis, a vital process involving the internalization and elimination of pathogens. Microbial destruction predominantly takes place within lysosomes and phagolysosomes^20–23. At later stages of inflammation, macrophages contribute to the resolution of inflammation, thus preventing progression from acute to persistent inflammation that would cause additional tissue damage.

In this study we have used a combination of cell biology techniques and animal models to better understand the role of RAS-dependent activation of p110α at the different stages of the inflammatory response. Our findings show that RAS-p110α signalling plays a key role in the initial stages of inflammation, facilitating the extravasation of monocytes from the bloodstream by promoting the necessary cytoskeletal changes. Subsequently, RAS-p110α has a crucial role in lysosomal acidification and activation of cathepsins, which are indispensable for efficient degradation of lysosomal cargo; when these functions are impaired, prolonged acute inflammatory responses and delayed resolution steps are observed. These results significantly enhance our understanding of the complex mechanisms governing the immune response to inflammation, emphasizing the pivotal role played by RAS-p110α signalling in orchestrating proper monocyte extravasation and maintaining optimal lysosomal function.

Results

Disruption of RAS-p110α causes prolonged and more acute responses to inflammatory stress in vivo

Previous data suggested that, in a tumoral setting, somatic disruption of RAS-p110α prevents macrophage recruitment to tumours^{11, 12} and favours polarization of macrophages to a proinflammatory phenotype¹², suggesting a possible role for p110α in macrophage function. To determine whether RAS-dependent activation of p110α participates in innate or adaptive immune responses to inflammation in macrophages, we used an established mouse model designed for the tamoxifen-inducible disruption of RAS binding to p110α^{11, 24} (Supplementary Fig. S1A). This mouse model introduces two-point mutations, T208D and K227A, in the RAS Binding Domain (RBD) of the endogenous Pik3ca allele (Pik3ca^RBD), enabling the selective disruption of the RAS-p110α interaction ²⁴. Wild-type (Pik3ca^WT) and Pik3ca^RBD mice were bred with mice containing a floxed Pik3ca allele²⁵, along with a strain containing a conditional Cre recombinase (Cre-ERT2) allele targeted to the ubiquitously expressed Rosa26 locus. The resulting Pik3ca^WT/Flox and Pik3ca^RBD/Flox mice displayed no discernible phenotype and exhibited behavior consistent with Pik3ca^WT mice^{11, 24}. Activation of Cre-recombinase by tamoxifen led to the excision of the floxed Pik3ca allele, resulting in mice expressing either one Pik3ca^WT allele (Pik3ca^WT/-) or one Pik3ca^RBD allele (Pik3ca^RBD/-)¹¹. This inducible genetic manipulation strategy allows us to selectively disrupt the RAS-p110α interaction in a controlled and temporally regulated manner, providing a valuable tool for dissecting the contributions of this pathway to immune responses in the context of inflammation.

Bone marrow derived macrophages (BMDMs) were generated from tibias and femurs of Pik3ca^WT/Flox and Pik3ca^RBD/Flox mice, ensuring efficient removal of the floxed allele (Supplementary Fig. S1B). Subsequently, these BMDMs were induced towards a proinflammatory state through stimulation with LPS and IFN-γ. To assess the impact of RAS-p110α binding deficiency on inflammatory intracellular signalling, we first examined Akt activation. The results revealed a decrease in Akt activation levels under inflammatory conditions in BMDMs lacking RAS-p110α binding, while no discernible change was observed in ERK activation, another well-known RAS effector (Fig. 1A and supplementary Fig. S1C and D). Interestingly, the decrease in Akt activation was accompanied by a decrease in the activation of NF-κB (Fig. 1B).

Figure 1.

Given the role of p65 in the expression of pro-inflammatory cytokines and chemokines²⁶ we next analyzed the expression of various inflammation mediators using a cytokine array in unstimulated and LPS- and IFN-γ-stimulated macrophages. Results showed that, under unstimulated conditions, RAS-PI3K disruption decreases the expression of IP-10, MIP-1α (Ccl3), JE (Ccl2/MCP1), IL-16, and IL-12p70, with no upregulated cytokines observed (Supplementary Fig. S1E). The downregulation of IP-10, MIP-1α, and JE indicates an impaired ability to recruit monocytes, macrophages, and T cells to sites of inflammation. This suggests that macrophages lacking RAS-PI3K interaction may have a reduced capacity to mount a robust immune response, particularly in recruiting and activating essential immune cells needed to combat pathogens or initiate inflammation.

Regarding LP +IFN-γ stimulated macrophages, results showed a decrease in the expression of IL-1β and IL-17, key drivers of pro-inflammatory responses, and an upregulation of IL-7, ILra, JE, BLC, I-309, Eotaxin, and G-CSF (Supplementary Fig. S1F). Elevated levels of these factors are associated with enhanced chemotactic signals and regulatory functions. Thus, the cytokine and chemokine expression profile observed in RAS-PI3K deficient macrophages suggests impaired pro-inflammatory responses and altered immune cell recruitment patterns, potentially influencing the resolution of inflammation and tissue repair processes.

Next we assessed whether the absence of RAS binding to p110α affects the ability of BMDMs to acquire a pro-inflammatory state characterized by increased expression of markers such as CD80, CD86, and MHCII^{27, 28}. To assess this, Pik3ca^RBD/-and Pik3ca^WT/- BMDMs were stimulated with LPS+IFN-γ, followed by flow cytometry analysis. Expression levels of CD80 (Supplementary Fig. S1G), CD86 (Supplementary Fig. S1H), and MCHII (Supplementary Fig. S1I) were examined. No significant differences were observed in the expression levels of any of these markers between the two genotypes under study.

Consequently, our next objective was to investigate whether in vivo disruption of RAS-p110α might lead to altered inflammatory responses. To address this, we administered tamoxifen to 10–12-week-old Pik3ca^WT/flox and Pik3ca^RBD/flox mice with tamoxifen and, after a two-week interval, conducted a paw swelling assay. In this assay, zymosan (10 μg/μl) or PBS was injected into the hind paws of Pik3ca^RBD/- and Pik3ca^WT/- mice with paw thickness measured at regular intervals over a 5-day period. The results revealed a significant increase in paw inflammation in Pik3ca^RBD/- mice compared to Pik3ca^WT/- mice, evident from the earlier time points measured and persisting throughout the experiment (Fig. 1D and 1E). Paws from the Pik3ca^WT/flox and Pik3ca^RBD/flox mice, which had not received tamoxifen, exhibited comparable levels of inflammation (Supplementary Fig. S1J). This reaffirms that the absence of p110α triggers a significant alteration in the inflammatory response and confirms that Pik3ca^WT/flox and Pik3ca^RBD/flox mice do not show any phenotype, as previously described^{11, 24}. Additionally, analysis of the blood sedimentation rate, a reliable indicator of systemic inflammation levels²⁹, showed a lower basal sedimentation ratio in Pik3ca^RBD/- mice under PBS-treated conditions (Supplementary Fig. S1K). Following zymosan injection, both genotypes exhibited an increase in sedimentation rate, but the rise was more pronounced in Pik3ca^RBD/- mice, indicating a heightened systemic inflammatory response. These findings underscore that disruption of RAS-p110α interaction results in an exacerbated inflammatory state, reflected in both localized paw inflammation and systemic inflammatory mediator levels.

To delve deeper into the inflammatory response induced by zymosan, we collected paw samples at two- and five-days post-injection and conducted Haematoxylin & Eosin (H&E) studies. This approach enabled a comprehensive analysis, allowing us not only to scrutinize the early stages of inflammation but also to monitor the subsequent resolution phase of the inflammatory process. Paws obtained from mice injected with zymosan for 2 days displayed extensive areas of damaged connective tissue in both Pik3ca^RBD/- and Pik3ca^WT/- paws (Fig. 1D). Notably, the inflamed region in Pik3ca^RBD/- mice was larger compared to control samples. After 5 days of zymosan injection, there was a significant reduction in the inflamed area, although it remained comparatively larger in the paws from Pik3ca^RBD/- mice (Fig.1D). This observation suggests a prolonged and heightened inflammatory response in mice lacking RAS-p110α interaction, emphasizing the role of this interaction in the regulation of inflammatory processes.

The cellular composition within the inflammatory abscess offers crucial insights into the severity and progression of the disease. Close examination with the pathologist revealed features indicative of an acute inflammatory response, including an inflammatory abscess with elevated numbers of polymorphonuclear cells, primarily neutrophils, and macrophages displaying altered cell shape and increased cell death, accompanied by fibrin and chromatin deposition (Fig. 1E and F). Notably, Pik3ca^RBD/- mice exhibited lower numbers of macrophages, a larger necrotic area with increased chromatin remnants, and reduced fibrin content compared to the paws from Pik3ca^WT/- mice (Fig. 1E and F). By day 5, paws from Pik3ca^WT/- mice showed a significant increase in the number of macrophages, a decrease in polymorphonuclear cells, and a nearly complete resolution of the necrotic area, indicating the initiation of the resolution phase. Moreover, there were abundant activated fibroblasts), suggesting active production of new connective tissue. In contrast, paws from Pik3ca^RBD/- mice exhibited a delayed healing process characterized by a larger central area of polymorphonuclear cells, abundant fibrin deposits, reduced numbers of infiltrating macrophages, and limited fibroblast activity (Fig. 1E and F). These findings collectively indicate an imbalance in the inflammatory response and a slower progression towards resolution in the absence of RAS-p110α interaction, emphasizing the pivotal role of this interaction in orchestrating an effective and timely resolution of inflammation.

To evaluate the presence of macrophages within the inflammatory lesion, we performed specific immunohistochemical (IHC) analysis using the macrophage-specific marker CD68 in paws from day 5, where more cellular preservation and initiation of the healing process had been observed. A notable decrease in the number of CD68-positive cells was observed in the inflamed abscess region of Pik3ca^RBD/- mice (Fig. 1G).

To further confirm the involvement of p110α signalling in the acute inflammatory response, Pik3ca^WT/WT mice were subjected to daily treatment with BYL719 (Alpelisib), a specific inhibitor of p110α isoform. After an initial 48-hour treatment period, the mice received injections of zymosan or PBS into their back-hind paws and were sacrificed 2 days later for analysis. Similar to what we had observed in Pik3ca^RBD/-mice, the inflamed area in BYL719-treated mice exhibited a larger extension than the inflamed are from non-treated mice (Fig. 1H and Supplementary Fig. S1L). The central necrotic region was significantly larger in the BYL719-treated mice, and it contained large amount of apoptotic polymorphonuclear cells, lower number of macrophages, and increased deposits of chromatin (Fig. 1H) resembling the observations from Pik3ca^RBD/- mice. Immunostaining for CD68 revealed a reduction in the number of macrophages present in the inflammatory abscess upon inhibition of p110α signalling with BYL719 treatment (Fig. 1I).

In summary, our findings highlight that both genetic disruption of RAS-p110α interaction and pharmacological inhibition of p110α contribute to an expanded inflamed area and central necrotic region, concurrently reducing macrophage infiltration in a zymosan-induced inflammation model. These results collectively underscore the pivotal role of the p110α isoform of PI3K in orchestrating the resolution of inflammatory responses, emphasizing its significance in modulating the dynamics and outcomes of inflammatory processes.

Disruption of RAS binding to p110α impairs the number of inflammatory monocytes in blood and spleen

Given the decrease in macrophages observed in the inflammatory abscess, we next set out to determine whether disruption of RAS binding to p110α had an effect on the number of monocytes circulating in the blood of adult mice. Pik3ca^WT/flox and Pik3ca^RBD/flox mice were treated with tamoxifen at 12–14 weeks of age and 4 weeks later, blood was collected by cardiac puncture and immune populations were analysed by flow cytometry. We found a decrease in the number of circulating classical (or inflammatory) monocytes (Ly6C^Hi/Ly6G^–/CD11b⁺ cells) in Pik3ca^RBD/- mice (Fig. 2A) and no changes were observed in non-classical (or non-inflammatory) monocytes (Ly6C^Lo/Ly6G^–/CD11b⁺ cells) (Fig. 2B). Together with the decrease in inflammatory monocytes, we observed an increase in the number of neutrophils (Ly6C^–/Ly6G⁺/CD11b⁺ cells) (Supplementary Fig. S2A). We did not detect differences in the numbers of T-cells (CD3⁺, CD8⁺ or CD4⁺) (Supplementary Fig. S2B) or B-cells (CD19⁺) (Supplementary Fig. S2C) in the blood after the disruption of RAS binding to p110α. We also analysed these same cell populations in Pik3ca^WT/flox and Pik3ca^RBD/flox mice and no differences were found (data not shown).

Figure 2.

Since splenic monocytes resemble their blood counterparts³⁰ we next aimed at determining whether splenic monocyte population would also be altered after disruption of RAS p110α interaction. As observed in circulating blood, the number of classical monocytes (Ly6C^Hi/Ly6G^–/CD11b⁺) in the spleen decreased after disruption of RAS-p110α interaction (Fig. 2C) and no differences were found in non-classical activated monocytes (Ly6C^Lo/ Ly6G^–/CD11b⁺) (Fig. 2D). We also checked the levels of resident macrophages present in the spleen and results showed that spleens from the Pik3ca^RBD/- mice had a decrease in the number of differentiated macrophages (F4/80⁺/CD11b⁺/CD45⁺) (Fig. 2E). Analysis of macrophages in the white and red pulp of the spleen indicated that Pik3ca^RBD/- mice had a significant decrease in the macrophage population in the former region (Fig. 2F and 2G). No differences were found in the number of granulocytes (Supplementary Fig. S2D), B-cells (Supplementary Fig. S2E) or T-cells (Supplementary Fig. S2F) present in the spleen between Pik3ca^RBD/- and Pik3ca^WT/- mice. There were no differences in spleen size from Pik3ca^RBD/- and Pik3ca^WT/- mice either (Supplementary Fig. S2G).

Given the decrease in the number of inflammatory monocytes and macrophages observed in blood and spleens of Pik3ca^RBD/- mice, we wondered if disruption of RAS activation of p110α could lead to a decrease in myeloid bone marrow precursors. Myeloid lineage descends from a common myeloid progenitor (CMP) in bone marrow and traverse into blood as mature cells³¹. CMP differentiates into the Granulocyte-Macrophage Progenitor (GMP) and the Megakaryocyte-Erythroid Progenitor (MEP). The CMP can generate all types of myeloid colonies, whereas the GMP or the MEP produces only granulocyte macrophage (GM) or megakaryocyte erythrocyte (MegE) lineage cells, respectively (Supplementary Fig. S2H). No significant differences were noted in the quantity of CMP progenitors within the bone marrow Pik3ca^RBD/- and Pik3ca^WT/- mice (Supplementary Fig. S2I-L). This observation suggests that the variations observed in blood and spleen parameters are not attributable to impairments in progenitor fate.

Finally, bone marrow precursors from Pik3ca^RBD/flox and Pik3ca^WT/flox mice were differentiated into macrophages in vitro, and the number of bone marrow-derived macrophages (BMDM) was determined by flow cytometry analysis. No differences were found in the number of BMDMs obtained from Pik3ca^RBD/- and Pik3ca^WT/- mice (Supplementary Fig. S2M), suggesting that the disruption of RAS-p110α signalling do not interfere with the ability of bone marrow precursors to differentiate to macrophages.

Disruption of RAS binding to p110α impairs monocyte transendothelial extravasation in response to inflammatory cues

The decrease in classical monocytes observed in the inflammatory abscess of paws from Pik3ca^RBD/- mice may indicate a decreased ability to mount an effective immune response. We carried out transwell assays since they are widely used to quantify transendothelial migration. Fibroblasts were seeded in the lower chamber to provide a continuous supply of Ccl2³², as this cytokine is well known for its ability to drive chemotaxis of myeloid cells under inflammatory conditions³³. Pik3ca^RBD/- or Pik3ca^WT/- BMDMs were seeded in the upper chamber of the transwell. Additionally, we stimulated Pik3ca^RBD/- or Pik3ca^WT/- BMDMs with lipopolysaccharide (LPS) and Interferon gamma (IFN-γ) to mimic/recapitulate the proinflammatory phenotype typically associated with bacterial infection that causes macrophage activation³⁴. Results showed that in both unstimulated and LPS+IFN-γ stimulated BMDMs, a lower number of Pik3ca^RBD/- BMDMs were able to go through the trans-well pore membranes compared to Pik3ca^WT/- BMDMs (Fig. 3A). As expected, when BMDMs were stimulated towards a proinflammatory phenotype, a decrease in their migratory ability is observed³⁵.

Figure 3.

Extravasation entails the migration of monocytes through the endothelium^{36, 37}, so we conducted random migration assays of Pik3ca^RBD/- and Pik3ca^WT/- BMDMs growing in matrigel coated plates under unstimulated conditions or activated towards an inflammatory phenotype by addition of LPS and IFN-γ to the media. Analysis of the data revealed that, under pro-inflammatory conditions, Pik3ca^RBD/- BMDMs displayed a decrease in migration speed (Fig. 3B). Migration was also evaluated in Pik3ca^WT/WT BMDMs treated with BYL-719. Results confirmed a decrease in migration speed of macrophages upon treatment with BYL-719 (Fig. 3B) indicating that inhibition of p110α, either genetically or pharmacologically, reduces the ability of macrophages to migrate under inflammatory conditions.

To determine if disruption of RAS binding to p110α impairs monocyte ability to extravasate through the endothelium in vivo in response to an inflammatory stress, we analysed monocyte extravasation through the mesenteric vein in response to intraperitoneal Ccl2 injection. It is well established that loss of p110α function leads to significant impairment in endothelial and lymphatic system^{12, 38}, so in order to determine if monocytes from Pik3ca^RBD/- mice presented an alteration in extravasation, we generated chimeric mice in which only the bone marrow were defective in RAS binding to p110α (Fig. 3C). For this, bone marrow from Pik3ca^RBD/- or Pik3ca^WT/- mice was injected through the tail vein of irradiated LysM-GFP donor mice³⁹. Engraftment of Pik3ca^RBD/- and Pik3ca^WT/- bone marrow could be followed by disappearance of the eGFP signal in donor mice (Fig. 3D and Supplementary Fig. S3A). Neutrophils (but no monocytes) were depleted using an anti-GR1 antibody (Fig. 3E and Supplementary Fig. S3B) to avoid interference with monocyte extravasation. Intraperitoneal injection of Ccl2 was performed in the peritoneum of the chimera mice to induce extravasation of monocytes through the mesenteric vein and rolling, adhesion and extravasation was measured by intravital microscopy. Flow cytometry analysis demonstrated no differences in Ccr2 expression between Pik3ca^RBD/- and Pik3ca^WT/- monocytes (Supplementary Fig. S3C). We observed that blocking RAS-p110α interaction did not induce a decrease in the number of rolling monocytes (Fig. 3F) or in the number of monocytes that adhere to the endothelium (Fig. 3G). However, when extravasation was measured, data showed that disruption of RAS binding to p110α caused a significant decrease in the number of monocytes that were capable of extravasating through the endothelium of the mesenteric vein (Fig. 3H).

Additionally, in order to determine whether the differences observed in the inflammatory response of Pik3ca^RBD/- mice were due to lack of monocyte extravasation, we repeated the zymosan paw swelling assay in the chimera mice. As previously, we observed that disruption of RAS-p110α only in the immune system led to higher inflammatory response and a delay in the initiation of the resolution phase (Fig. 3I).

Disruption of RAS-p110α activation in macrophages induces changes in cytoskeleton reorganization

During transendothelial migration, leukocytes undergo cytoskeletal rearrangements that allow them to squeeze through the tight spaces between endothelial cells and enter the underlying tissue^{40, 41}. Therefore, we next sought to determine whether the differences observed in Pik3ca^RBD/- BMDMs extravasation and migration were attributable to altered cytoskeletal dynamics. To do so, we first aimed at examining cell shape and spread in Pik3ca^RBD/- and Pik3ca^WT/- BMDMs in pro-inflammatory conditions after LPS+IFN-γ treatment. Treatment with LPS+IFNγ induced an increase in cell spread in both Pik3ca^RBD/- and Pik3ca^WT/- BMDMs when compared to their respective unstimulated counterparts (Fig. 4A and 4B). However, cell spread in LPS+IFNγ activated Pik3ca^RBD/- BMDMs was significantly decreased compared to that observed Pik3ca^WT/- BMDMs (Fig. 4A and 4B). Additionally, Pik3ca^RBD/- BMDMs were more elongated and did not acquire the typical rounded shape known to be induced in macrophages after treatment with LPS+IFN-γ⁴² (Fig. 4A and 4C). We also analysed cell height and results showed that Pik3ca^RBD/- BMDMs are higher both in unstimulated conditions and after LPS+IFN-γ stimulation (Fig. 4D).

Figure 4.

Actin is a bona fide regulator of cell shape⁴³, so we next aimed at exploring the actin cytoskeleton in Pik3ca^RBD/- and Pik3ca^WT/- BMDMs. Actin fractionation assays were carried out in unstimulated or LPS+IFNY activated Pik3ca^RBD/- and Pik3ca^WT/- BMDMs. Our results revealed that disruption of RAS-p110α binding caused a decrease in the F-actin pool in unstimulated BMDMs (Fig. 4E), with no changes observed in the G-actin pool when compared to matched controls (Fig. 4F). Analysis of actin dynamics after activation with LPS+IFNγ showed that, in proinflammatory conditions, the F-actin pool is increased, as expected⁴⁴ (Fig. 4E). However, actin polymerization was significantly active in Pik3ca^RBD/- BMDMs, leading to a striking increase in F-actin when compared to that observed in controls (Fig. 4E). In parallel, a decrease in the G-actin pool in Pik3ca^RBD/- BMDMs was observed, suggesting an increase in the stabilization of F-actin after disruption of RAS binding to p110α (Fig. 4F).

We next sought to investigate whether the enhanced F-actin stabilization corresponded to increased cellular rigidity. Consequently, we measured the elastic modulus of Pik3ca^RBD/- BMDMs, revealing a significant increase in cell stiffness following RAS-p110α disruption under both basal and proinflammatory conditions (Fig. 4G). To validate that the heightened cell stiffness was attributed to p110α, we treated Pik3ca^WT/WT BMDMs with BYL-719 and observed an amplified elastic modulus in these cells when stimulated with LPS+IFN-γ (Fig. 4G), confirming that the inhibition of p110α indeed results in augmented cellular rigidity.

These observations led us to hypothesize that phagocytosis, which heavily relies on actin rearrangement, might also be affected in Pik3ca^RBD/- BMDMs. To test this hypothesis, we assessed the ability of Pik3ca^RBD/- and Pik3ca^WT/- BMDMs to phagocytose fluorescent microspheres, Borrelia burgdorferi, and apoptotic cells. The results revealed varying outcomes depending on the target. When phagocytosing 1 μm non-opsonized green-fluorescent beads, Pik3ca^RBD/- BMDMs were less efficient at engulfing the microspheres compared to their WT counterparts (Supplementary Fig. S4A). Similar results were obtained when Pik3ca^WT/WT BMDMs were treated with BYL-719 (Supplementary Fig. S4A). In contrast, no significant differences were observed between Pik3ca^RBD/- and Pik3ca^WT/- BMDMs in their ability to phagocytose Borrelia burgdorferi (Supplementary Fig. S4B).

Lastly, we evaluated the phagocytosis of apoptotic cells by Pik3ca^RBD/- BMDMs over extended periods. There were no differences in the initial uptake of apoptotic cells between Pik3ca^RBD/-and Pik3ca^WT/- BMDMs. However, at later time points, Pik3ca^RBD/- BMDMs showed an accumulation of phagocytosed particles. Similar results were obtained when Pik3ca^WT/WT BMDMs were treated with BYL-719. These data suggest a delay in processing and degradation, indicating a potential role for RAS-p110α in the later stages of phagocytosis (Fig. 4H).

In summary, our findings suggest that loss of RAS-p110α interaction leads to increased actin polymerization, resulting in stiffer and less deformable cells, which impairs phagocytic efficiency for certain targets. Specifically, RAS-p110α is important for the effective phagocytosis of non-biological particles and may also play a role in the proper processing and degradation of phagocytosed apoptotic cells. The differential effects on various phagocytic targets highlight the complexity of RAS-p110α’s role in macrophage biology and underscore the importance of cytoskeletal flexibility in efficient phagocytosis.

Disruption of RAS-p110α activation impacts the secretome of macrophages

Given our previous observations, we hypothesized that the cytoskeletal alterations observed after disruption of RAS-p110α interaction, may significantly impact the secretory functions of macrophages. The accumulation of apoptotic material in Pik3ca^RBD/- BMDMs suggests a disruption in normal phagocytic processing, which could influence the release of cytokines and other inflammatory mediators. Thus, we analysed the secretome of Pik3ca^RBD/- and Pik3ca^WT/- BMDMs both under steady state condition and during phagocytosis. We utilized apoptotic LKR10 cells, a murine lung cancer cell line, as the substrate in our phagocytosis assay. LKR10 cells were exposed to cisplatin for 16 h and apoptosis was confirmed by Western Blot (Supplementary Fig. S5A). Our goal was to create a more physiologically relevant experimental setting that more closely mimics the complex nature of the inflammatory response. For the secretome analysis, macrophages were incubated with or without apoptotic cells for 16 h. Culture supernatants were collected, clarified, and subjected to label-free quantitative proteomics analysis.

A total of 127 peptides corresponding to 105 proteins (Supplementary Table 1) showed differential expression between Pik3ca^RBD/- and Pik3ca^WT/- BMDM secretomes at steady state conditions. Additionally, 359 peptides corresponding to 210 proteins (Supplementary Table 2) were present a significantly different levels in the secretomes of Pik3ca^RBD/- and Pik3ca^WT/- BMDM during phagocytosis of apoptotic cells. Next, we compared these peptides with the list of secreted proteins available at The Human Protein Atlas, and removed those that do not correspond to secreted proteins. After this step, 18 proteins were found to be differentially secreted by Pik3ca^RBD/- and Pik3ca^WT/- BMDM in steady state conditions (Fig. 5A), and 38 by Pik3ca^RBD/- and Pik3ca^WT/- BMDM during phagocytosis (Fig. 5B and Supplementary Fig. S5B and S5C). Surprisingly, most proteins were secreted at lower levels in Pik3ca^RBD/- BMDMs, independently of the conditions under study. 12 proteins were differentially secreted in both experimental conditions under study (Fig. 5C).

Figure 5.

Functional analysis of differentially secreted proteins in unstimulated BMDMs showed no significant pathways related to these group of proteins (Fig. 5D). However, differentially secreted proteins in phagocytosing conditions were involved in two main biological processes: complement activation (C1qa, C1qb and C1qc connected through physical interaction and C3, C9 and Cfb through coexpression) and lysosome function, mainly cathepsins (Ctsd, Ctsb, Ctsz, Cst3, Psap, Anxa1 and Gsn) (Fig. 5E). Together, the complement cascade and lysosome function work in concert to provide an effective defence against pathogens and promote overall maintenance of cellular homeostasis^{22, 45–47} and data from the secretome analysis of Pik3ca^RBD/- and Pik3ca^WT/- BMDMs suggests that RAS activation of p110α may play a crucial role in the regulation of both response pathways. Moreover, these findings align with previous observations showing an accumulation of phagocytosed material in Pik3ca^RBD/- BMDMs, indicating a potential disruption in the processing and degradation of engulfed targets.

Disruption of RAS-p110α signalling leads to altered lysosomal function

We next aimed to functionally validate the proteomics data suggesting that RAS-p110α activation regulates lysosomal function in macrophages. First, we performed immunofluorescence analysis of lysosomal-associated membrane protein 1 (LAMP1) to assess lysosomal biogenesis and function in Pik3ca^RBD/- and Pik3ca^WT/- BMDMs. Analysis of LAMP1 in Pik3ca^RBD/- and Pik3ca^WT/- BMDMs showed a decrease in LAMP1 expression in steady state conditions (Fig. 6A) suggesting a decrease in the number of lysosomes after disruption of RAS-p110α interaction. However, Pik3ca^RBD/- showed increased levels of Lamp1 expression when subjected to phagocytosis of apoptotic cells, suggesting an increase in the number of phagolysosomes present in these cells.

Figure 6.

We hypothesised that the increase in Lamp1 expression in BMDMs lacking RAS-p110α interaction during phagocytosis could be attributed to aberrant lysosomal function and phagolysosome retention. Thus, we next evaluated lysosomal activity by using the lysosomotropic dye lysotracker red coupled with epifluorescence analysis, since it is specifically taken up by acidic organelles. As such, its accumulation is proportional to the number of acidic vesicles. Data analysis showed that disruption of RAS-PI3K in BMDMs leads to a significant decrease in lysotracker uptake, both in unstimulated conditions and also after activation with LPS+IFN-γ (Fig. 6B). Similar results were obtained with control BMDMs treated with BYL719, the p110α specific inhibitor. Data analysis showed a decrease in the uptake of lysotracker after p110α inhibition (Fig. 6B), further suggesting that loss of p110α function in BMDMs results in altered lysosomal pH. To confirm that lysosomal pH of Pik3ca^RBD/- BMDMs was less acidic, we next stained Pik3ca^RBD/- and Pik3ca^WT/- BMDMs with Green lysosensor, a pH-sensitive dye that exhibit a pH-dependent increase in fluorescent intensity upon lysosomal acidification. As shown in Fig. 6C, Pik3ca^RBD/- BMDMs presented attenuated fluorescent intensity both in unstimulated and activated when compared with that in the control group, indicating that their lysosomal content is less acidic than in Pik3ca^WT/- BMDMs.

Considering that Pik3ca^RBD/- lysosomes were less acidic, we next investigated the expression level and activation of some of the cathepsins identified in the secretome analysis (Fig. 5) by western blotting. Cathepsins play a critical role in lysosomal protein degradation. They are initially synthesized as inactive precursors and are activated through proteolysis in the lysosome at a low pH⁴⁸. We found a reduction in the expression and activation of cathepsin D and cathepsin B in Pik3ca^RBD/- BMDMs upon stimulation with LPS+IFN-γ (Figure 6D). This observation suggests that the impaired lysosomal pH observed in Pik3ca^RBD/- BMDMs could potentially account for the observed decrease in cathepsin activity.

Acidification of lysosomes and cathepsin activation are critical steps for activation of the resolutive stage of the inflammatory response, so we next evaluated the ability of Pik3ca^RBD/- BMDM’s lysosomal compartment to degrade internalised particles. We set up another phagocytosis assay in which Pik3ca^RBD/- and Pik3ca^WT/- BMDMs would phagocytose apoptotic cells that were transduced with GFP (which is pH sensitive⁴⁹) and also labelled with Celltracker Red CMTPX (pH insensitive). Pik3ca^RBD/- and Pik3ca^WT/- BMDMs were allowed to engulf these apoptotic cells for 16 h and after this time, apoptotic cells were eliminated by washes and BMDMs were analysed by flow cytometry at different time points to measure GFP signal. Our results showed that GFP signal is lost significantly faster in control BMDMs than in Pik3ca^RBD/- BMDMs (Fig. 6E). As expected, we did not observe any decay in the red tracker signal. Collectively, these data evidence that Pik3ca^RBD/- BMDMs exhibit a significant impairment in removing engulfed particles, that can be attributed to the absence of lysosome acidification.

We have shown a delayed clearance of apoptotic cells in Pik3ca^RBD/- mice after zymosan injection (Fig. 1). This, together with previous data took us to analyse the levels of Cathepsin D in the inflamed abscess from the paws of Pik3ca^RBD/- and Pik3ca^WT/- mice. Results showed a significant decrease in the levels of Cathepsin D in the inflamed area of Pik3ca^RBD/- mice (Fig. 6F) and BYL719-treated mice (Fig. 6G).

Lysosomal digestion plays a pivotal role in activation of resolutive programs by mediating PPAR activation ^{50, 51}. Analysis of PPARδ and PPARγ expression in phagocyting macrophages showed a significant decrease in the expression of PPARδ, but no PPARγ in Pik3ca^RBD/- BMDMs (Supplementary Fig. S6A), suggesting that resolutive programs might not be activated effectively.

In summary, our findings underscore the crucial role of RAS-p110α signalling axis in maintaining the balance of the inflammatory response and promoting timely resolution, providing further evidence for the significant involvement of RAS-p110α signalling in the response to inflammatory stimuli.

Discussion

In this study, we provide compelling evidence that disruption of RAS-p110α signalling or chemical inhibition of p110α impairs response to inflammatory stresses due to defects in both monocyte extravasation during the early stages of the inflammatory response and decreased lysosomal function during the later stages.

Monocyte extravasation during the inflammatory response is a critical step that allows immune cells to reach the site of infection or injury¹⁹. Impairment on this process causes slower or inadequate immune response, as macrophages are crucial for detecting and engulfing pathogens or cellular debris at the site of inflammation, as well as delayed resolution of inflammation, resulting in prolonged inflammation and potential tissue damage ^{52, 53}. Our data shows that RAS binding to p110α is involved in macrophage extravasation by modulating actin dynamics. During monocyte extravasation, actin filaments form actin-rich protrusions that are essential for monocyte migration across the endothelium^{41, 54}. Our data show that, during extravasation, RAS-p110α signalling regulates actin dynamics so monocytes are able to squeeze through endothelial cells^{41, 43}.

The disruption of RAS-p110α in macrophages emerges as a pivotal determinant in cellular mechanics, elucidating a cascade of events resulting in increased F-actin levels. The increase in cytoskeletal components, particularly F-actin, causes a notable increase in cell stiffness and a concurrent decrease in cell deformability. The orchestration of these changes underscores the intricate balance in cytoskeletal dynamics. Notably, the regulatory role of Rho-GTPases comes into focus as potential mediators of this phenomenon. Rho-GTPases are well-known mediators of actin cytoskeletal rearrangements⁵⁵, and their dysregulation is known to impact cell mechanics^{56, 57}. The observed increase in cell stiffness and deformability, may thus be governed by the modulation of Rho-GTPase activity. This intricate interplay provides a compelling avenue for further investigation into the molecular mechanisms through which RAS-p110α disruption influences Rho-GTPases, ultimately shaping the biomechanical properties of macrophages and may open avenues for therapeutic interventions targeting cytoskeletal dynamics in macrophages.

Monocyte extravasation shares many features with monocyte egression from the bone marrow. During monocyte egress, monocytes also undergo changes in cytoskeleton dynamics to detach from the sinusoidal endothelial cells and to migrate through the endothelial fenestrae and basement membrane into the bone marrow sinusoids⁵⁸. Thus, the decrease in the number of classical monocytes in the Pik3ca^RBD/- mice in blood may be due, at least in part, to a defect in the extravasation process. This result may also explain, at least partially, the lack of macrophage recruitment to lung tumours found in previous studies using the RBD mouse model^{11, 12}.

Our findings also revealed a crucial role of RAS-p110α activation in the acute phase of the inflammatory response by regulating effective lysosomal degradation of phagocytosed and engulfed material. Phagocytosis constitutes a vital step in the inflammatory response, whereby phagosomes bind to lysosomes to form phagolysosomes in which pathogens are eradicated to facilitate an appropriate host response. This response encompasses antigen presentation to engage T-cell responses, secretion of inflammatory mediators that guide the adaptive immune response, and initiation of tissue repair mechanisms ^{22, 47}. Our data provide evidence that the activation of RAS-p110α signalling pathway is involved in the critical process of lysosomal acidification, which is essential for the efficient degradation of internalized particles and the activation of proteolytic enzymes, ultimately resulting in the formation of fully functional lysosomes. Consequently, lack of lysosome acidification impairs the expression and activation of important proteases such as cathepsin B and cathepsin D. The proper functioning of lysosomes is essential for mounting a robust response to inflammatory stress. Lysosomes play a central role in the breakdown of pathogens and dead cells by providing the necessary degradative enzymes and maintaining an acidic environment that facilitates the degradation of engulfed particles^{20, 21}. When lysosomal function is compromised in macrophages, the degradation of phagocytosed material becomes impaired, leading to the accumulation of toxic debris. This accumulation subsequently triggers inflammation and causes damage to the surrounding tissues, leading to chronic inflammation and appearance of lysosomal storage diseases (LSD)^{59, 60}.

Our data showed that the disruption of Ras-p110α in macrophages presents a multifaceted impact in inflammatory response, notably manifesting as a reduction in NF-κB activation and cathepsin expression coupled with diminished lysosomal function. NF-κB, a pivotal regulator of inflammatory responses^{61, 62}, is intricately linked to lysosomal dynamics, with lysosomes playing a crucial role in modulating NF-κB signalling through the degradation of relevant molecules^{63, 64}. Additionally, the observed decrease in NF-κB activation aligns with the decline in lysosomal function, which has previously been suggested in previous reports^{65, 66}. This bidirectional influence, where perturbations in lysosomal function coincide with alterations in NF-κB activity, underscores the intricate and context-dependent nature of this regulatory network, suggesting a plausible mechanistic interdependence. Our data suggest that disruption in Ras-p110α signalling cascade could trigger downstream effects impacting both NF-κB activity and lysosomal integrity. This intricate nexus between Ras-p110α signalling, NF-κB activation, cathepsin expression, and lysosomal function underscores the complexity of cellular regulatory networks. Further analysis of the molecular pathways connecting these observations holds the potential to reveal novel insights into the coordinated regulation of inflammatory and lysosomal processes in macrophages. Thus, further investigations are necessary to decipher the causal relationships and unravel the broader implications of Ras-p110α disruption in shaping these intricate cellular responses during the inflammatory response.

The possible functional link between the increased actin polymerization and lysosomal dysfunction observed in Pik3ca^RBD/- mice remains an unanswered question. Lysosome acidification and actin polymerization are tightly interconnected processes that have pivotal roles in numerous cellular functions^67–69. Studies have demonstrated that lysosome acidification can influence actin polymerization dynamics through the activation of specific actin-binding proteins and the modulation of actin-regulatory proteins^{67, 70}. Conversely, disruption in lysosome acidification, such as impaired proton pump activity or lysosomal storage disorders, have been associated with changes in actin polymerization and the organization of the cytoskeleton. Notably, it has been reported that depolimeryzation of F-actin plays a crucial role in assembling the macromolecular components of the acidification machinery in nascent endosomes⁷¹. Therefore, the intricate relationship between lysosome acidification and actin polymerization suggests a potential reciprocal influence, where perturbations in one process could impact the other. Further investigation is required to determine the precise regulatory mechanisms by which p110α influences both lysosome acidification and actin polymerization, whether it occurs in a linear manner or through separate pathways.

This study provides valuable insights into the mechanisms that govern the immune response to inflammation, particularly emphasizing the essential role of RAS-p110α signaling. Our results underscore the pivotal role of RAS-p110α signaling in both the initiation and resolution of inflammation. The impaired ability of RAS-PI3K-deficient monocytes and macrophages to effectively migrate, initiate a robust inflammatory response, and resolve inflammation highlights the critical importance of this pathway in maintaining immune homeostasis. The persistent inflammation observed in these models may contribute to the development and perpetuation of chronic inflammatory conditions. Given the centrality of p110α in both initiating and resolving inflammation, its dysfunction could be a key factor in the chronic, dysregulated inflammation characteristic of diseases such as rheumatoid arthritis, inflammatory bowel disease, psoriasis, or systemic lupus erythematosus.

The potential of p110α as a therapeutic target in these conditions is especially intriguing. The recent identification of a p110α small molecule activator⁷² offers a promising tool to explore this avenue. By transiently enhancing p110α function, it may be possible to promote the resolution of inflammation and thereby alleviate the symptoms and progression of these inflammatory diseases. Further research into this therapeutic strategy could open new pathways for treating chronic inflammatory disorders, providing much-needed relief for patients affected by these debilitating conditions.

Additional Declarations

The authors declare no competing interests.

Authors’ Disclosures

The authors declare no conflict of interest.

Authors’ Contributions

Conception and design: E. Castellano. Development of methodology: A. Rosell, M. Alcón Pérez, M.B. Voisin, J. de Paz, V. Rajeeve, C. Cuesta, O. Swinyard, E. Gabandé-Rodriguez, H. Fraile. Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Rosell, M. Alcón Pérez, A. Krywoska, M.B. Voisin, J. Downward, P. Cutillas, C. García-Macías, J. Alcaraz, E. Castellano. Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Rosell, M. Alcón Pérez, A. Krywoska, M.B. Voisin, V. Rajeeve, A. Berral-Gonzalez, J. De Las Rivas, P. Cutillas, E. Castellano. Writing, review, and/or revision of the manuscript: A. Rosell, M. Alcón Pérez, M.B. Voisin, J. De Las Rivas, P. Cutillas, E. Castellano.

Acknowledgements

This work was supported by grants from the Spanish Ministry of Science and Innovation (RTI2018- 099161-A-I00), Programa JAE-Intro ICU from CSIC (JAEICU-21-IBMCC-6), JCyL (CSI185-20), Marie Curie Initial Training Network on Tumour Infiltrating Myeloid Cell Compartment (PF7 MCA-ITN317445) and CRUK-Barts Cancer Centre Development Fund. This research was co-financed by FEDER funds. The CIC is supported by the Programa de Apoyo a Planes Estratégicos de Investigación de Estructuras de Investigación de Excelencia of Castilla y León autonomous government (CLC-2017-01) and AECC Excellence program Stop Ras Cancers (EPAEC222641CICS). The authors wish to thank the Pathology Unit, the Mouse Model Experimentation Unit, and the Advanced Cellular Analysis Unit at CIC for their assistance in carrying out this work.

Methods

Animal studies

For removal of Pik3ca-floxed allele, Pik3ca^RBD/lox and Pik3ca^WT/flox mice⁶⁴ were given 3.2 mg tamoxifen (Sigma) dissolved in 80 μl of corn oil by oral gavage once per day during 3 consecutive days. Efficiency of tamoxifen treatment was routinely performed by genotyping for the presence of the floxed allele.

For paw oedema studies Pik3ca^RBD/- and Pik3ca^WT/- mice⁶⁴ were divided into groups of four two weeks before oedema induction. Before inducing the paw oedema, the mice were anesthetized with 4% isofluorane. To induce the oedema, mice received ipsilateral i.pl. injection (30 μl) of either zymosan (10 μg/μl, Sigma Aldrich) or PBS into the back-hind paw. Injection of 0.1mg/kg Buprenorphine (NOAH, Vetergesic) was given for pain prevention. Paw thickness was measured using a calliper every hour during the first 6 hours after injection and then at 8, 10 hours and twice per day afterwards. Buprenorphine was injected twice per day during the length of the experiment. Mice were kept, managed, and sacrificed in the NUCLEUS animal facility of the University of Salamanca according to current European (2007/526/CE) and Spanish (RD 1201/2005 and RD53/2013) legislation. All experiments were approved by the Bioethics Committee of the Cancer Research Center.

Isolation, culture and treatments of BMDM

Bone marrow cells from tibias and femurs of 12–14-week-old Pik3ca^RBD/Lox mice and Pik3ca^WT/Lox littermates were cultured with DMEM supplemented with 10% FBS, 100 units/ml penicillin, 100 μg/mL streptomycin, 2 mM L-Glutamine and 20 ng/mL M-CSF for 7 days. 4-hydroxytamoxifen (Sigma Aldrich) (100 nM) was added to culture media on day 3 to eliminate Pik3ca-Lox allele. The differentiated BMDM were then detached using cell dissociation buffer (C5914-100, Merck) and cultured in DMEM supplemented with 10% FBS, 100 μg/mL streptomycin, 2 mM L-Glutamine, 20 ng/mL M-CSF for unstimulated BMDMs. For macrophage polarization towards an inflammatory phenotype 20 ng/ml IFNγ (Peprotech) and 100 ng/ml LPS (Sigma Aldrich) was added to culture media.

Generation of chimeric animals

Chimeric mice exhibiting WT or RBD deficient leukocytes were generated by lethal irradiation with 5.5 Gy twice, 4 h apart of LysM GFP recipient animals (mice exhibiting endogenously GFP fluorescent monocytes and neutrophils) followed by an injection of bone marrow cells (1.5 × 10⁶ cells/recipient i.v.) from C57BL/6 WT or RBD donor mice. Chimerism was then assessed 4 weeks later by flow cytometry from blood samples (reconstitution of 99.8 ± 0.2 % and 99.8 ± 0.1% for WT and RBD deficient donor cells, respectively; n =5 mice per group).

Neutrophil depletion

Neutrophil depletion of chimeric mice was induced by intraperitoneal injection of anti-GR1 25μg/mouse/day for 3 days). Numbers of blood circulating monocytes and neutrophils were quantified by flow cytometry pre- and post-depletion. Neutrophils were found to be reduced by 99.5%, whilst this anti GR1-depleting protocol had no effect on blood monocyte proportion (n=5 mice/group).

Brightfield intravital confocal microscopy

Mesenteric inflammation was induced following intraperitoneal injection of mouse recombinant CCL2 (500ng/mouse in 500uL of PBS). Six hours later, anesthetized chimeric mice (150 mg/kg ketamine, 7.5 mg/kg xylazine, i.p.) were placed in supine position on a heating pad (37 °C) for maintenance of body temperature. The mesenteric vascular bed was exteriorized, placed on a purpose-built stage of an upright brightfield microscope (Zeiss Axioskop). Mesenteries were superfused with warmed (37 °C) Tyrode’s solution (Sigma). After a 5-min equilibration period, analysis of leukocyte-endothelium interactions was made in at least 9 (and up to 16) randomly selected segments (100 μm in length) of post-capillary venules (20–40 μm in diameter) for each mouse. Leukocyte rolling was quantified by counting the number of rolling cells passing a fixed transversal line in the middle of the vessel segment for 5 min. Leukocyte adhesion (stationary position of the cell for 30 s or longer) was quantified along a 100 μm vessel length and data were normalized as the number of cells per 500um vessel segments. Leukocyte extravasation response was quantified within 50 μm on either side of the 100 μm vessel segment in the perivenular tissue; and data were normalized as the number of extravasated leukocytes per mm2 of extravascular tissue. At the end of the analysis period, mice were humanely killed by cervical dislocation.

BMDM elastic modulus

The elastic modulus of BMDMs was quantitatively assessed utilizing Atomic Force Microscopy (AFM). Specifically, cells were cultured on cover glass slides, which were subsequently positioned in a specialized BIO-AFM setup integrated with an inverted optical microscope (Nikon TE2000). The BIO-AFM was equipped with a V-shaped, four-sided pyramidal silicon nitride tip (Bruker AFM Probes) to facilitate accurate force measurements. To avoid localized variations and ensure data representativeness, no more than three cells were selected from a single field of view for mechanical characterization, and these cells were never contiguous. Subsequently, the stiffness of each individual cell was characterized through measurements obtained at three perinuclear points. For each point, force-displacement (F vs. z) curves were captured (10 μm amplitude at a speed of 5 μm/s). The determination of the elastic modulus was performed based on the analysis of force-displacement curves. A total of five curves were recorded for each perinuclear point at an indentation depth of 300 nm. Data were analysed by fitting the curves to the Hertz model as previously described^{73, 74}. Finally, the elastic modulus for each cellular population under distinct experimental conditions was calculated based on a minimum of fifteen measurements per independent cell, with each condition comprising at least ten independent cells (n ≥ 10).

Cytokine arrays

Bone marrow-derived macrophages (BMDMs) were cultured in 6-well plates at a density of 1×10⁶ cells per well. Cytokine production was analyzed in unstimulated and M1 macrophages using commercial mouse cytokine array (R&D Systems, #ARY006). Cells were lysed at 4°C for 30 minutes using a lysis buffer containing 1% Igepal CA-630, 20mM Tris-HCl (pH 8.0), 137mM NaCl, 10% glycerol, 2mM EDTA, 10μg/ml Aprotinin, 10μg/ml Leupeptin, and 10μg/ml Pepstatin, following the manufacturer’s protocol. Lysates were then centrifuged at 16,100 RCF for 15 minutes at 4°C. The supernatant was collected, and protein concentration was determined using a Qbit 2.0 Fluorometer. A total of 1400μg of protein was diluted to a final volume of 1ml with Array Buffer 6, followed by the addition of 0.5ml Array Buffer 4 and 15μl of the reconstituted Mouse Cytokine Array Panel A Detection Antibody Cocktail. The samples were mixed and incubated for one hour on a shaker at room temperature. Nitrocellulose membranes from the mouse cytokine array were blocked with Array Buffer 6 for one hour at room temperature on a shaker. The blocking buffer was then removed, and the sample mixture was added to the arrays, which were incubated overnight at 4°C on a shaker. Membranes were washed with Wash Buffer provided by the manufacturer. Membranes were then incubated with 1× Streptavidin-HRP in Array Buffer 6 for 30 minutes at room temperature. After further washing, the Chemi Reagent Mix was applied, and signals were detected using an Amersham Imager 600 (GE Healthcare, #29-0981-07 AC).

Immunofluorescence

BMDM were fixed using 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 and blocked for 1 h with 3% BSA in PBS before incubation with the primary antibodies, used at a 1:100 dilution: Lamp1 (#553792, BdPharmigen), Deoxyribonuclease I-Alexa Fluor™ 488 Conjugate (#D12371, Invitrogen, 1:2000). To stain actin cytoskeleton, Alexa Fluor™ 647 Phalloidin (Invitrogen, 1:10 000) was directly added to the primary antibody mixture. Alexa Fluor 488- or Alexa Fluor 555-conjugated secondary antibodies (Invitrogen) were used to detect the indicated proteins at a 1:1000 dilution. Cells were counterstained with DAPI on the mounting solution (ProLong Gold Antifade Reagent with DAPI, Invitrogen). Images were taken using a Zeiss LSM510 confocal microscope or Leica DM6 B THUNDER Imager 3D Tissue.

Transwell migration assay

Transwell migration assays were carried out using the 6.5 mm Transwell® with 8.0 μm Pore Polyester Membrane Insert (Corning). 9×10⁴ MEFs from wild type mice were used as a chemo-attractant to encourage macrophage migration. 8×10⁵ BMDM were seeded in the transwell. Transwells were performed following the manufacturer instructions.

Random migration assay

For random migration assays, BMDMs were seeded in 24-well plates coated with matrigel (0.5mg/ml) and labelled using CellTracker™ Red CMTPX Dye 1 μM (ThermoFisher) for 30 minutes. 24 hours later LPS + IFN-γ was added when necessary. Triplicates of each condition and genotype were prepared. Time-lapse imaging was carried out for 24 h. One image was taken every 10 min within the same well using a Nikon microscope driven by Metamorph (Molecular Devices, Chicago, IL, USA). A total of 80–100 cells per condition were tracked using the Fiji plugin Trackmate. Pre-processing was done using Mexican hat filter 3.0 radius to increase particle detection. Images were segmented using the fluorescence channel with the Laplacian of the Gaussian detector with a 30 μm estimated particle diameter, a 10.0 threshold and median filter option selected. Segmented objects were linked from frame to frame with a Linear Assigment Problem (LAP) tracker with 45 μm frame-to-frame linking distance and 2 frame gap closure. Criteria for track acceptance were track duration at least the 90% of the video. Tracks were visually inspected for completeness and accuracy of the tracking.

Secretome mass spectrometry

Samples for secretome analysis were prepared as previously described¹³⁰. In brief, 100ug of proteins were digested into peptides using trypsin and peptides were desalted using Oasis HLB extraction cartridges (Waters UK Ltd)) and eluted with 50% acetonitrile (ACN) in 0.1% Trifluoroacetic acid (TFA).

Dried peptides were dissolved in 0.1% TFA and analysed by nano ACQUITY liquid chromatography (Waters Corp., Milford, MA, USA) coupled on-line to a tandem LTQ Orbitrap XL, mass spectrometer (Thermo Fisher Scientific)¹³¹. Gradient elution was from 5% to 25% buffer B in 180 min at a flow rate 300nL/min with buffer A being used to balance the mobile phase (buffer A was 0.1% formic acid in water and B was 0.1% formic acid in ACN). The mass spectrometer was controlled by Xcalibur software and operated in the positive mode. The spray voltage was 1.95 kV and the capillary temperature was set to 200 °C. The LTQ Orbitrap XL was operated in data dependent mode with one survey MS scan followed by 5 MS/MS scans. Label-free quantitative proteomics analysis was performed using three independent biological samples per group. Additionally, each sample was analysed in technical duplicates. To ensure robust quantitative analysis, we utilized LTQ Orbitrap XL tandem mass spectrometry (MS/MS) to generate six distinct mass spectral profiles from each group.

MS raw files were converted into Mascot Generic Format using Mascot Distiller (version 2.3.0) and searched against the SwissProt database (release December 2015) restricted to human entries using the Mascot search daemon (version 2.3.1). Allowed mass windows were 10 ppm and 600 mmu for parent and fragment mass to charge values, respectively. Variable modifications included in searches were oxidation of methionine, pyro-glu (N-term) and phosphorylation of serine, threonine and tyrosine.

Spectral counting quantification method relies on the number of times peptides are identified by tandem mass spectrometry (with expectancy value <0.05) from a given protein. Spectral counts were obtained from Mascot result (DAT) files using a python script written in house in the Mascot Parser Toolkit environment (version 2.4.x).

Proteomic data analysis

The proteomic data obtained consisted of 6,844 peptides and 30 samples: Pik3ca^WT/- and Pik3ca^RBD/- BMDMs in steady state conditions (labelled as cell samples: WT/- and RBD/-), phagocytosing apoptotic cells (labelled as cell samples: WT/-Phag and RBD/-Phag), and the apoptotic LKR10 cells alone (labelled as LKR). For each of these samples, proteomic experiments were performed with 6 replicates (3 biological replicates × 2 technical replicates), yielding a data set of 30 samples.

The first analytical step was to remove all peptides for which there was no information contained in the proteomic raw data matrix and the peptides for which 85% or more of the signal values were missing. All these peptides were specific of mouse proteins and in many cases were unique. The corresponding proteins were annotated and labelled together with each measured peptide. Next, low-quality samples were also removed, testing the overall signal per sample to identify if there were clear outliers with very low signal or with a very different signal distribution. Comparison of the overall signal distributions of the 30 samples (comparing boxplots) and identified 3 samples that were very different were obtained and discarded (WT/-Phag_s3r2 (sample 3, replicate 2), RBD/-_s2r1 and LKRc_s1r1). These 3 samples showed a median signal in their distributions that deviated >20% from the median signal of the distributions of all other samples.

Differential expression analysis for each peptide of each protein were next performed. The algorithm used to carry out this analysis was limmaVoom within EdgeR R package^{75, 76}. Prior to this analysis, a Bartlett test was performed to see the homogeneity of variances, verifying that for this data we cannot consider equality of variances and this factor was included in the differential expression algorithm. With this algorithm, normalization factors to use a-posteriori were calculated and, transformation and calculation of the variance weights was performed. The model to fit before using Voom as specified since it uses the variances of the model’s residuals (observed - fitted). Finally, an estimation of the contrast for each feature tested (i.e. each peptide) was carried out using the Empirical Bayes approach in limma as previously described⁷⁷. Peptides were ordered by the p.value of the limma test considering significant peptides changed only with a p.value below 0.05 and with a log2(Fold-Change) >|2|. All these analyses were performed using the statistical computing language R and packages or libraries obtained from R-cran (cran.r-project.org) or Bioconductor (www.bioconductor.org).

Cytoscape software (v3.9)⁷⁸ including GeneMania app⁷⁹ was then used to generate and visualize protein-protein networks of the significantly altered proteins selected in secretome analysis of unstimulated and phagocyting BMDMs. This tool provides information on protein-protein associations based in co-expression studies and also based in physical interaction studies.

Western blot analysis

Immunoblot was performed per a general western-blot protocol (Abcam). Total protein was extracted using Cell Lysis Buffer (Cell Signaling Technology) supplemented with c0mplete mini protease inhibitor cocktail (Roche), 50 mM sodium fluoride and 1 mM of PMSF. Protein was quantified using Bradford Method (Bio-Rad). 20 μg of protein was separated by SDS-PAGE and transferred to 0.2 um pore-size PVDF membranes (Sigma-Aldrich). Blots were probed using the following antibodies, at a concentration 1:1000 unless otherwise stated: cathepsin B (12216-1-AP, Proteintech), cathepsin D (21327-1-AP, Proteintech), α-tubulin (ab15246, Abcam; concentration 1:5000). Horseradish peroxidase-conjugated secondary antibodies (Amersham) were used (1:5000) and detected using an enhanced chemiluminescent substrate (Amersham). Signal was detected using an iBright 1500 System (Invitrogen).

Flow cytometry analysis

Single-cell suspensions from cultured cell, spleen or blood monocytes were generated from mice, washed twice in staining buffer and incubated with 1:100 Fc-block (BD Biosciences, #553142) diluted in FACS buffer. Cells were subjected to surface antibody staining with labelled antibodies diluted in staining buffer for 30 min at 4 °C: CD3-PE-Cy7 (#100328, Biolegend), CD4-BV605 (#100548, Biolegend), CD8-APC (#100712, Biolegend), CD19-PerCP-Cy5.5 (#115534, Biolegend), CD45-BV785 (#103149, Biolegend), Ly6C-PerCP-Cy5.5 (#128012, Biolegend), Ly6C-E450 (#48-5932-82, eBioscience), Ly6G-AF700 (#56-5931-82, eBioscience), CD11b-BV650 (#101239, Biolegend), F4/80-PE-Cy7 (#123114, Biolegend), F4/80-PE (#123110, Biolegend), CCR2 (CD192)-PE-Vio ® 770 (#130-108-724, Miltenyi Biotec). After incubation, cells were washed in staining buffer and analysed immediately. For all staining, isotype controls were used.

The gating strategy for analysing distinct cellular populations began by isolating cells positive for the CD45 marker. Subsequently, B cells, identified by their positivity for CD19 and negativity for CD11b, were selected. To determine the frequency of T cells, the CD45+ cell population, devoid of both CD19 and CD11b, underwent examination for CD3 expression. CD3+ cells were further categorized based on CD4 and CD8 expression, distinguishing CD4+, CD8+, and double-negative (DN) T cells. Following the application of these gating criteria across all samples, the percentages of CD45-positive cells expressing CD19 and T cells (CD3+) within the CD45+ population were calculated. Additionally, myeloid populations were analysed by quantifying the proportion of CD11b+ cells. To distinguish between circulating monocytes and granulocytes, we assessed their expression of Ly6G (granulocytes) and Ly6C proteins (monocytes). Classical monocytes were characterized by negativity for Ly6G and positivity for Ly6C, while alternative monocytes lacked both markers, and neutrophils/granulocytes displayed weak positivity for Ly6C and positivity for Ly6G.

Samples were acquired on a BD LSR FORTESA FACS or FACS Aria III machine that uses FACS DIVA software (BD Biosciences). Compensation was performed using 1 drop of Ultracomp ebeads (eBioscience) in 300αl of FACS buffer. 1μl of each antibody used in the pool was mixed with 100μl of compensation beads solution and acquired. A total of 50000 cells per mouse were analysed. Analysis was performed with FlowJo software (FlowJo V10.4). Once the different pools were compensated samples were acquired.

Phagocytosis assay

For phagocytosis assay, BMDMs in suspension in DMEM without FBS were labelled with 1:200 red cell tracker (Molecular Probes) for 30 minutes at 37°C. Cells were then centrifuged for 5 minutes at 300g at 4°C, supernatant was removed and labelled BMDMs were washed twice with 5ml of PBS and plated overnight in complete DMEM containing 20 ng/ml M-CSF. On the following day, 100 ng/ml LPS (Sigma Aldrich) was added to BMDMs overnight.

LKR10 cells (murine lung cancer cell line) were stained with red cell tracker (Molecular Probes) as previously described for macrophages. Stained cells were plated in complete DMEM medium and, after 12 hours, 50 μM Cisplatin (MCE MedChemExpress) was added to the media and left overnight. BMDMs were then incubated with apoptotic cancer cells at a 1:2 ratio and cultured at 37 °C for different time points in DMEM supplemented with 10% FBS.

Flow cytometry data were acquired using a BD LSR FORTESSA FACS instrument with FACS DIVA software (BD Biosciences) and analysed using FlowJo V10.4 software. A minimum of 2×10⁵ events were acquired and analysed. Data analysis and interpretation was done using FlowJo software (FlowJo V10.4).

Image analysis

Confocal images were post-processed and analysed using Fiji distribution of ImageJ version 1.53q. Cell shape descriptors such as “aspect ratio” (AR), “circularity” (C) and “cell area” were measured using Fiji. Specifically, aspect ratio is calculated as (major axisxminor axis-1) therefore representing solely the degree of elongation, whereas circularity is calculated as [4π*(area × perimeter-2)], thus representing the degree of similarity to a circumference with a value ranging from 0 to 1 (perfect circle).

Histology

Tissue was fixed using 4% formaldehyde for 48h, dehydrated and paraffin-embedded. Sections (3 μm) were cut and stained using hematoxylin-eosin. For immunodetection, citrate pH 6 buffer was used for antigen retrieval. Staining was used using the following primary antibodies: CD68 (ab125212, Abcam, 1:200), cathepsin B (12216-1-AP, Proteintech, 1:100), cathepsin D (21327-1-AP, Proteintech, 1:400). Dako EnVision+ System HRP labelled Polymer secondary antibodies (Dako) were used, and DAB+ Substrate Chromogen System (Dako) was used for color development.

For the quantification of loose chromatin remnants in paw edema images, the pathologist assigned a score ranging from 1 to 3 based on the chromatin content within the inflammatory abscess. A score of 1 indicates low chromatin content, occupying less than 30% of the abscess area, while a score of 3 represents high chromatin content, covering over 60% of the abscess area.

Supplementary Data

Supplementary Figure S1.

Supplementary Figure S2.

Supplementary Figure S3.

Supplementary Figure S4.

Supplementary Figure S5.

Supplementary Figure S6.