CAR-T cells have yet to show widespread efficacy in solid tumors due in part to their poor persistence and loss of function in the tumor microenvironment. Further, heterogenous expression of most CAR target antigens in solid tumors can lead to escape of antigen-null tumors that resist CAR-T killing. Strategies to cooperatively boost both CAR-T and endogenous anti-tumor immunity could curb tumor escape and may be critical for achieving durable efficacy in cancer patients. NKTR-255 is a polymer-conjugated IL-15 with extended half-life that can boost endogenous T and NK cells, as well as CD19 CAR-T activity in B cell malignancies. However, whether NKTR-255 is sufficient to overcome CAR-T dysfunction in the suppressive solid tumor microenvironment, and how NKTR-255 and CAR-Ts together re-shape endogenous anti-tumor immunity, is not known. Using an autochthonous mouse model of ROR1+ lung adenocarcinoma, we show that NKTR-255 significantly boosted accumulation, reduced exhaustion, and improved function of tumor-infiltrating CAR-T cells. Compared with NKTR-255 or CAR-T treatment alone, combination of NKTR-255 and CAR-T therapy synergistically increased tumor-infiltrating CD11b+ cytotoxic NK cells, activated dendritic cells, and endogenous tumor-specific T cells that preserved a PD-1+Tcf1+ stem-like phenotype. Consequently, NKTR-255 and CAR-T combination therapy induced complete elimination of ROR1+ tumor and significantly improved survival, with enhanced tumor control dependent on activity of both CAR-Ts and endogenous T cells. Altogether, our data suggest that combining NKTR-255 with CAR-T therapy is a promising strategy to enhance both CAR-T and endogenous anti-tumor immunity to promote coordinated control of aggressive tumors.

Conventional monocytes (cMonos) and their progeny, macrophages, support metastasis by priming pre-metastatic niches and aiding tumor colonization. While cMonos recruited to the metastatic lung differentiate into metastasis-associated macrophages (MAMs) that support tumor survival, the molecular mechanisms driving this process within the metastatic tumor microenvironment (mTME) remain unclear. Environmental signals are key to determining cMono lineage fate. My preliminary work shows that cMonos differentiating into MAMs in metastatic lungs exhibit a distinct surface proteomic signature compared to those in naive lungs, highlighting the mTME’s role in their differentiation. To investigate this further, I used high-throughput flow cytometry (Infinity Flow) to profile cMonos at different stages of differentiation, alongside a pilot scRNA-seq assay using Flex from 10x. This revealed heterogeneity within the MAM population, identifying four distinct clusters that suggest multiple differentiation pathways for cMonos. Thus, I hypothesize that the early metastatic niche provides distinct, temporally regulated signals that program cMonos to differentiate into diverse cell fates. To test this, I executed a comprehensive time-course scRNA-seq and CITE-seq assay. Using advanced cell differentiation trajectory analysis tools like Palantir and Mellon, I aim to identify rare transitional states of cMonos and the signaling pathways orchestrating their differentiation into heterogeneous fates. NIH/NCI F31 Award Tumor Immunology: Cellular Responses and Tumor Microevironment (TIME)

Lung development begins in utero and reaches full maturity post birth. Dendritic cells (DC) play a key role in immune regulation in lungs. However, comprehensive exploration of DCs in these immature lungs has not been performed. Here we explored DCs from fetal to newborn mouse lungs phenotypically, ontogenetically, transcriptomically and functionally and found two DC subsets, resembling adult cDC1 and cDC2, but with key differences. Phenotypically, fetal-cDC1 lacks the classical-DC1 (cDC1) marker XCR1, while the fetal-cDC2 express both cDCassociated genes as well as monocyte-derived DC genes. Both DC subsets wane as lungs enter the alveolar stage, giving way to the more familiar adult cDC1 and cDC2. Both fetal-cDC1 and fetal-cDC2 derive from ED14.5 fetal liver Macrophage Dendritic Progenitors, not from monocytes or classic Precursor-cDC (Pre-cDC), indicating a unique ontogeny of first DCs in developing mouse lungs. Together we provide the first in depth exploration of first DCs in developing lungs.

During tumor progression and especially following cytotoxic therapy, cell death of both tumor and stromal cells is widespread. Despite clinical observations that high levels of apoptotic cells correlate with poorer patient outcomes, the physiological effects of dying cells on tumor progression remain incompletely understood. Here, we report that circulating apoptotic cells robustly enhance tumor cell metastasis to the lungs. Using intravenous metastasis models, we observed that the presence of apoptotic cells, but not cells dying by other mechanisms, supports circulating tumor cell (CTC) survival following arrest in the lung vasculature. Apoptotic cells promote CTC survival by recruiting platelets to the forming metastatic niche. Apoptotic cells externalize the phospholipid phosphatidylserine to the outer leaflet of the plasma membrane, which we found increased the activity of the coagulation initiator Tissue Factor, thereby triggering the formation of platelet clots that protect proximal CTCs. Inhibiting the ability of apoptotic cells to induce coagulation by knocking out Tissue Factor, blocking phosphatidylserine, or administering the anticoagulant heparin abrogated the pro-metastatic effect of apoptotic cells. This work demonstrates a previously unappreciated role for apoptotic cells in facilitating metastasis by establishing CTC-supportive emboli, and suggests points of intervention that may reduce the pro-metastatic effect of apoptotic cells. GRAPHICAL ABSTRACT:

Lung-resident type 1 conventional dendritic cells (cDC1s) are critical for initiating immune responses against foreign respiratory viruses and endogenous assaults like cancer cells. Importantly, the cDCs of each tissue display unique phenotypes and functions dictated by the environment they populate. However, our understanding of the factors that regulate cDC1 development and function in the lung is minimal. The cytokines, FMS-like tyrosine kinase 3 ligand (FLT3L), and granulocyte-macrophage stimulating factor (GM-CSF) are essential for the development of tissue-resident cDC1, but in in vitro systems, these factors are insufficient to generate cDC1s with a lung-specific phenotype. RNAseq analysis of lung cDC1 identified a clear enrichment in genes associated with transforming growth factor beta (TGF-β) signaling, a factor also known to be enriched in the airspaces of the lung. TGF-β has been previously reported to maintain DC quiescence in the lung, however, our data suggested it may also play a role in differentiation and lung specification of these cells. Based on this, I hypothesized that adding TGF-β to in vitro bone marrow (BM) cultures alongside established cytokines FLT3L and GM-CSF would generate cDC1s phenotypically and functionally similar to in vivo lung cDC1s. The addition of TGF-β, but not GM-CSF, to optimized FLT3L-BMDC cultures resulted in development of cDC1 which highly resemble lung cDC1. The development of a high throughput in vitro system that differentiates functional lung cDC1s will not only inform the in vivo factors required for development of these cells but also provides a useful platform for interrogation of lung cDC1 function and a stepping-off point to begin models for human lung cDC1 production and study Supported by grants from the University of Washington PREP program NIH (R25) and the Fred Hutchinson Cancer Center Diversity, Equity, and Inclusion Student Fellowship

Pre-metastatic niche formation is a critical step during the metastatic spread of cancer. One way by which primary tumors prime host cells at future metastatic sites is through the shedding of tumor-derived micropar-ticles as a consequence of vascular sheer flow. However, it remains unclear how the uptake of such particles by resident immune cells affects their phenotype and function. Here, we show that ingestion of tumor-derived microparticles by macrophages induces a rapid metabolic and phenotypic switch that is characterized by enhanced mitochondrial mass and function, increased oxidative phosphorylation, and upregulation of adhesion molecules, resulting in reduced motility in the early metastatic lung. This reprogramming event is dependent on signaling through the mTORC1, but not the mTORC2, pathway and is induced by uptake of tumor-derived microparticles. Together, these data support a mechanism by which uptake of tumor-derived microparticles induces reprogramming of macrophages to shape their fate and function in the early metasta-tic lung.

Cancer-associated fibroblasts (CAFs) mediate an immunosuppressive effect, but the underlying mechanism remains incompletely defined. Here we show that increasing prostatic stromal Foxf2 suppresses the growth and progression of both syngeneic and autochthonous mouse prostate cancer models in an immunocompetent context. Mechanistically, Foxf2 moderately attenuates the CAF phenotype and transcriptionally downregulates Cxcl5, which diminish the immunosuppressive myeloid cells and enhance T cell cytotoxicity. Increasing prostatic stromal Foxf2 sensitizes prostate cancer to the immune checkpoint blockade therapies. Augmenting lung stromal Foxf2 also mediates an immunosuppressive milieu and inhibits lung colonization of prostate cancer. FOXF2 is expressed higher in the stroma of human transition zone (TZ) than peripheral zone (PZ) prostate. The stromal FOXF2 expression level in primary prostate cancers inversely correlates with the Gleason grade. Our study establishes Foxf2 as a stromal transcription factor modulating the tumor immune microenvironment and potentially explains why cancers are relatively rare and indolent in the TZ prostate.

This study presents a novel method for low-cost cell surface proteomics using flow cytometry and machine learning.

alpha v beta 8 integrin, a key activator of transforming growth factor beta(TGF-beta), inhibits anti-tumor immunity. We show that a potent blocking monoclonal antibody against alpha v beta 8 (ADWA-11) causes growth suppression or complete regression in syngeneic models of squamous cell carcinoma, mammary cancer, colon cancer, and prostate cancer, especially when combined with other immunomodulators or radiotherapy. alpha v beta 8 is expressed at the highest levels in CD4+CD25+ T cells in tumors, and specific deletion of beta 8 from T cells is as effective as ADWA-11 in suppressing tumor growth. ADWA-11 increases expression of a suite of genes in tumor-infiltrating CD8(+) T cells normally inhibited by TGF-beta and involved in tumor cell killing, including granzyme B and interferon-g. The in vitro cytotoxic effect of tumor CD8 T cells is inhibited by CD4+CD25+ cells, and this suppressive effect is blocked by ADWA-11. These findings solidify alpha v beta 8 integrin as a promising target for cancer immunotherapy.

The lung represents a unique immune environment. The primary function of the lung is to enable gas exchange by facilitating the transfer of oxygen into and carbon dioxide out of the blood. However, as a direct byproduct of this process the lung is also constantly exposed to particles, allergens, and pathogens alongside air itself. Due to this, the pulmonary immune system exists in a fine balance between quiescence and inflammation, deviations from which can lead to a failure in respiratory function. A rich history exists attempting to define the critical features of lung immunity, and most recently advances in intravital microscopy have enabled the visualization of intercellular immune dynamics in both steady-state and a variety of disease conditions. In this review, we will summarize a variety of approaches to intravital lung imaging as well as how its application has advanced our understanding of normal lung function as well as disease states such as pulmonary metastasis, asthma, and lung injury.

The response to the COVID-19 crisis across most research institutions mandated ceasing nonessential research activities in order to minimize the spread of the virus in our communities. With minimal notice, experiments were terminated, cell lines were frozen, mouse colonies were culled, and trainees were prevented from performing bench research. Still, despite the interruption of experimental productivity, the shutdown has proven for many PIs and trainees that doing and thinking science are not activities that are bound to the laboratory. Furthermore, the shutdowns have solidified important emerging trends and forced us to further innovate to get the most out of working remotely. We hope that some of these innovations, hard-gained in this difficult time, will persist and develop into new paradigms-lessons that will improve our science and our relationship to the climate and community beyond the current pandemic.

Modern immunologic research increasingly requires high-dimensional analyses in order to understand the complex milieu of cell-types that comprise the tissue microenvironments of disease. To achieve this, we developed Infinity Flow combining hundreds of overlapping flow cytometry panels using machine learning to enable the simultaneous analysis of the co-expression patterns of 100s of surface-expressed proteins across millions of individual cells. In this study, we demonstrate that this approach allows the comprehensive analysis of the cellular constituency of the steady-state murine lung and to identify novel cellular heterogeneity in the lungs of melanoma metastasis bearing mice. We show that by using supervised machine learning, Infinity Flow enhances the accuracy and depth of clustering or dimensionality reduction algorithms. Infinity Flow is a highly scalable, low-cost and accessible solution to single cell proteomics in complex tissues.

Modern flow cytometers can quantify around 20 proteins at the single-cell level. Higher-dimensional alternatives exist, such as mass cytometry, or oligonucleotides-tagged antibodies - but they feature lower cellular throughput, higher cost and are not as widely available. To overcome these limitations, we propose a pipeline which allows the quantification of hundreds of proteins across millions of single cells. Experimentally, cells are stained with a customizable 10–20 “backbone” antibody panel which defines the cellular population structure of the sample. This is followed by aliquoting into hundreds (200–300) of individual wells, each containing a unique “exploratory” antibody. Data acquisition is performed using conventional flow cytometry. Computationally, multivariate non-linear regression models are trained, one for each well. These models learn how to impute an exploratory antibody signal from backbone measurements. The models are each applied to the backbone data to obtain hundreds of imputed measurements across millions of single cells. We applied this workflow cell suspensions from the lungs of C57/B6 mice, resulting in 2,660,000 events with imputed expression levels for 266 proteins. Imputed expression and co-expression patterns are accurate and consensual across regression models, allowing near-exhaustive annotation of both immune and non-immune cells, and finer granularity than dimensionality reduction or clustering. By enabling broad characterization of cellular networks at the protein level, this pipeline is well-suited for systems immunology approaches. Given the wealth of data generated by this pipeline and its accessibility, we anticipate it will be widely adopted.

Metastasis, the main cause of cancer-related death, has traditionally been viewed as a late-occurring process during cancer progression. Using the MMTV-PyMT luminal B breast cancer model, we demonstrate that the lung metastatic niche is established early during tumorigenesis. We found that matrix metalloproteinase 9 (MMP9) is an important component of the metastatic niche early in tumorigenesis and promotes circulating tumor cells to colonize the lungs. Blocking active MMP9, using a monoclonal antibody specific to the active form of gelatinases, inhibited endogenous and experimental lung metastases in the MMTV-PyMT model. Mechanistically, inhibiting MMP9 attenuated migration, invasion, and colony formation and promoted CD8+ T cell infiltration and activation. Interestingly, primary tumor burden was unaffected, suggesting that inhibiting active MMP9 is primarily effective during the early metastatic cascade. These findings suggest that the early metastatic circuit can be disrupted by inhibiting active MMP9 and warrant further studies of MMP9-targeted anti-metastatic breast cancer therapy.

The pulmonary immune system defends a huge surface area directly in contact with the contents of the air we breathe. Neutrophils, the most abundant immune cell in the pulmonary vasculature, are critical to immunity but they are also capable of generating life-threatening pathology. Natural Killer cells are the most highly represented lymphocyte subset in the lung, but relatively little is known about their localization, motility or the specific mechanisms by which they contribute to local homeostasis. Here, we used lung-intravital microscopy to directly visualise and quantify neutrophil and natural killer cell dynamics in the pulmonary vasculature of live mice. This approach revealed unexpected sessile behaviour by intravascular natural killer cells. Interactions with natural killer cells made neutrophils scan the endothelium more slowly over larger distances and reduced the number of neutrophils that accumulated in an LPS-triggered inflammatory challenge. This represents a new paradigm by which natural killer cells contribute to lung physiology by diminishing potentially pathogenic neutrophil accumulation.

Group 2 innate lymphoid cells (ILC2s) are enriched in mucosal tissues (e.g., lung) and respond to epithelial cell-derived cytokines initiating type 2 inflammation. During inflammation, ILC2 numbers are increased in the lung. However, the mechanisms controlling ILC2 trafficking and motility within inflamed lungs remain unclear and are crucial for understanding ILC2 function in pulmonary immunity. Using several approaches, including lung intravital microscopy, we demonstrate that pulmonary ILC2s are highly dynamic, exhibit amoeboid-like movement, and aggregate in the lung peribronchial and perivascular spaces. They express distinct chemokine receptors, including CCR8, and actively home to CCL8 deposits located around the airway epithelium. Within lung tissue, ILC2s were particularly motile in extracellular matrix-enriched regions. We show that collagen-I drives ILC2 to markedly change their morphology by remodeling their actin cytoskeleton to promote environmental exploration critical for regulating eosinophilic inflammation. Our study provides previously unappreciated insights into ILC2 migratory patterns during inflammation and highlights the importance of environmental guidance cues in the lung in controlling ILC2 dynamics.

Article Figures and data Abstract eLife digest Introduction Results Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Chimeric antigen receptors (CARs) are synthetic receptors that reprogram T cells to kill cancer. The success of CAR-T cell therapies highlights the promise of programmed immunity and suggests that applying CAR strategies to other immune cell lineages may be beneficial. Here, we engineered a family of Chimeric Antigen Receptors for Phagocytosis (CAR-Ps) that direct macrophages to engulf specific targets, including cancer cells. CAR-Ps consist of an extracellular antibody fragment, which can be modified to direct CAR-P activity towards specific antigens. By screening a panel of engulfment receptor intracellular domains, we found that the cytosolic domains from Megf10 and FcRɣ robustly triggered engulfment independently of their native extracellular domain. We show that CAR-Ps drive specific engulfment of antigen-coated synthetic particles and whole human cancer cells. Addition of a tandem PI3K recruitment domain increased cancer cell engulfment. Finally, we show that CAR-P expressing murine macrophages reduce cancer cell number in co-culture by over 40%. https://doi.org/10.7554/eLife.36688.001 eLife digest Our immune system constantly patrols our body, looking to eliminate cancerous cells and harmful microbes. It can spot these threats because it recognizes certain signals at the surface of dangerous cells. However, cancer cells often find ways to ‘hide’ from our immune system. Chimeric antigen receptors, or CARs, are receptors designed in a laboratory to attach to specific proteins that are found on a cancer cell. These receptors tell immune cells, such as T cells, to attack cancers. T cells that carry CARs are already used to treat people with blood cancers. Yet, these immune cells are not good at penetrating a solid tumor to kill the cells inside, which limits their use. Macrophages are a group of immune cells that can make their way inside tumors and travel to cancers that the rest of the immune system cannot reach. They defend our body by ‘swallowing’ harmful cells. Would it then be possible to use CARs to program macrophages to ‘eat’ cancer cells? Morrissey, Williamson et al. created a new type of CARs, named CAR-P, and introduced it in macrophages. These cells were then able to recognize and attack beads covered in proteins found on cancer cells. The modified macrophages could also limit the growth of live cancer cells in a dish by ‘biting’ and even ‘eating’ them. While these results are promising in the laboratory, the next step is to test whether these reprogrammed macrophages can recognize and fight cancers in living animals. https://doi.org/10.7554/eLife.36688.002 Introduction Chimeric antigen receptors (CARs) are synthetic transmembrane receptors that redirect T cell activity towards clinically relevant targets (reviewed in [Lim et al., 2017; Fesnak et al., 2016]). The CAR-T receptor contains an extracellular single chain antibody fragment (scFv) that recognizes known tumor antigens, and intracellular signaling domains from the T Cell Receptor (TCR) and costimulatory molecules that trigger T cell activation (Fesnak et al., 2016; Kochenderfer et al., 2009). CAR-T cells recognizing CD19, a marker expressed at high levels on the surface of B cells and B cell-derived malignancies, have been used successfully to target hematological malignancies with 70–90% of patients showing measurable improvement (Lim et al., 2017; Engel et al., 1995; Haso et al., 2013). The success of CAR-T suggests that programming immune cells to target cancer might be a broadly applicable approach. Macrophages are critical effectors of the innate immune system, responsible for engulfing debris and pathogens. Harnessing macrophages to combat tumor growth is of longstanding interest (Alvey and Discher, 2017; Lee et al., 2016). Macrophages are uniquely capable of penetrating solid tumors, while other immune cells, like T cells, are physically excluded or inactivated (Lim et al., 2017; Lee et al., 2016). This suggests that engineered macrophages may augment existing T cell-based therapies. Early efforts transferring healthy macrophages into cancer patients failed to inhibit tumor growth, suggesting that macrophages require additional signals to direct their activity towards tumors (Lacerna et al., 1988; Andreesen et al., 1990). Antibody blockade of CD47, a negative regulator of phagocytosis, reduced tumor burden, indicating that shifting the balance in favor of macrophage activation and engulfment is a promising therapeutic avenue (Majeti et al., 2009; Chao et al., 2010; Jaiswal et al., 2009; Tseng et al., 2013). Here, we report a family of chimeric antigen receptors that activate phagocytosis of cancer cells based on recognition of defined cell surface markers, resulting in significantly reduced cancer cell growth. Results To program engulfment towards a target antigen, we created a CAR strategy using the CAR-T design as a guide (Fesnak et al., 2016). We call this new class of synthetic receptors Chimeric Antigen Receptors for Phagocytosis (CAR-Ps). The CAR-P molecules contain the extracellular single-chain antibody variable fragment (scFv) recognizing the B cell antigen CD19 (ɑCD19) and the CD8 transmembrane domain present in the ɑCD19 CAR-T (Fesnak et al., 2016; Kochenderfer et al., 2009). To identify cytoplasmic domains capable of promoting phagocytosis, we screened a library of known murine phagocytic receptors: Megf10 (Figure 1a), the common ɣ subunit of Fc receptors (FcRɣ), Bai1, and MerTK (Penberthy and Ravichandran, 2016). FcR triggers engulfment of antibody-bound particles, while the other receptors recognize apoptotic corpses (Freeman and Grinstein, 2014; Penberthy and Ravichandran, 2016). We also made a receptor containing an extracellular ɑCD19 antibody fragment and a cytoplasmic GFP, but no signaling domain, to test whether adhesion mediated by the ɑCD19 antibody fragment is sufficient to induce engulfment (Figure 1a; CAR-PGFP). Figure 1 with 2 supplements see all Download asset Open asset Identification of intracellular signaling region for CAR-P. (A) Schematics show the structure of CAR-P constructs. An ɑCD19 (purple) or ɑCD22 (blue, center) scFv directs CAR specificity. Intracellular signaling domains from Megf10 or the indicated engulfment receptor (green) activate engulfment. CAR-PGFP contains only GFP and no intracellular signaling domains (right). All constructs include a transmembrane domain from CD8 and a C-terminal GFP. (B) J774A.1 macrophages expressing ɑCD19 CAR-P with the indicated intracellular signaling domain (green) engulf 5 µm silica beads covered with a supported lipid bilayer containing His-tagged CD19 extracellular domain. The beads were visualized with atto390-labeled lipid incorporated into the supported lipid bilayer (magenta). Cells infected with the cell membrane marker, mCherry-CAAX, were used as a control (no CAR, top left). To the right of each image is a histogram depicting the frequency of cells engulfing the indicated number of beads. The average number of beads eaten per cell is quantified in (C). (D) Bone marrow derived macrophages were infected with CAR-PFcRɣ or GFP-CAAX (green, left and center top; grey, center bottom) and incubated with CD19 beads (magenta) for 45 min. Images show an x-y plane through the center of the engulfed beads (left), or a cross section (center) of the z plane indicated in the inset panel (white line). The histogram depicts the number of cells engulfing the indicated number of beads. The scale bar indicates 5 µm and n = 78–163 cells per condition, collected during three separate experiments. Error bars denote 95% confidence intervals and *** indicates p<0.0001 compared to mCherry-CAAX control by Kruskal-Wallis test with Dunn’s multiple comparison correction. https://doi.org/10.7554/eLife.36688.003 To assay our library of CAR-Ps, we introduced each CAR-P into J774A.1 murine macrophages by lentiviral infection. As an engulfment target, we used 5 µm diameter silica beads coated with a supported lipid bilayer. A His8-tagged extracellular domain of CD19 was bound to a NiNTA-lipid incorporated into the supported lipid bilayers. Macrophages expressing a CAR-P with the Megf10 (CAR-PMegf10) or FcRɣ (CAR-PFcRɣ) intracellular domain promoted significant engulfment of CD19 beads compared to macrophages with no CAR (Figure 1b,c, Figure 1—video 1). Macrophages expressing CAR-PBai1, CAR-PMerTK, and the adhesion-only CAR-PGFP did not bind the CD19 beads even though these CAR-Ps are present at the cell surface (Figure 1b,c, Figure 1—figure supplement 1). To confirm that the CAR-P was a viable strategy for redirecting primary macrophages, we expressed the CAR-PFcRɣ in primary murine bone marrow derived macrophages and found that these transfected primary cells also were able to trigger engulfment of CD19 beads (Figure 1d). Next we asked if the CAR-P strategy could target a different antigen. Because CAR-PMegf10 performed well in our initial screen (Figure 1a), we developed ɑCD22 CAR-PMegf10 using a previously developed ɑCD22 antibody fragment (Xiao et al., 2009; Haso et al., 2013). Consistent with our results using ɑCD19-based CARs, ɑCD22 CAR-PMegf10 promoted engulfment of CD22 beads (Figure 2a). To confirm antigen specificity of CAR-P, we incubated ɑCD19 CAR-PMegf10 macrophages with CD22 beads, and ɑCD22 CAR-PMegf10 macrophages with CD19 beads. CD19 beads were not eaten by ɑCD22 CAR-PMegf10 macrophages, and CD22 beads were not eaten by ɑCD19 CAR-PMegf10 macrophages (Figure 2a). These data indicate that CAR-PMegf10 specifically triggers engulfment in response to the target ligand and that the CAR-P strategy is able to target multiple cancer antigens. Figure 2 Download asset Open asset CAR-P expression drives specific engulfment of diverse beads. (A) Macrophages infected with the ɑCD19 (purple) or ɑCD22 (blue) CAR-PMegf10 or mCherry-CAAX control were fed 5 µm beads ligated with either CD19 (left) or CD22 (right). Engulfment is quantified as the mean beads eaten per cell. The fraction of phagocytic cells is as follows: 31/144 GFP-CAAX cells engulfed CD19 beads, 87/149 ɑCD19 CAR-PMegf10 engulfed CD19 beads, 20/142 ɑCD22 CAR-PMegf10 engulfed CD19 beads, 28/140 GFP-CAAX cells engulfed CD22 beads, 18/151 ɑCD19 CAR-PMegf10 engulfed CD22 beads, 103/148 ɑCD22 CAR-PMegf10 engulfed CD22 beads (pooled data was collected during three separate experiments). Error bars denote 95% confidence intervals and *** indicates p<0.0001 compared to mCherry-CAAX control by Kruskal-Wallis test with Dunn’s multiple comparison correction. (B) J774A.1 macrophages expressing the ɑCD19 CAR-PMegf10 (green) were fed beads of various sizes (magenta, diameter of bead indicated below image). The beads were covered in a supported lipid bilayer ligated to His-tagged CD19 extracellular domain and the number of beads engulfed per cell is reported below each image (magenta bars indicate CAR-PMegf10 macrophages and pink bars indicate CAR-PGFP). The ɑCD19 CAR-PMegf10 macrophages were also incubated with 10 µm beads coated in phosphatidylserine (PS) and ICAM-1 (blue bar in graph, 51/390 cells engulfed a bead). The fraction of cells engulfing a CD19 bead is as follows: 135/169 CAR-PMegf10 and 134/187 CAR-PGFP cells engulfed 2.5 um bead, 126/395 CAR-PMegf10 and 112/499 CAR-PGFP cells engulfed a 5 um bead, 48/377 CAR-PMegf10 and 21/378 CAR-PGFP cells engulfed a 10 um bead, 120/706 CAR-PMegf10 and 45/675 CAR-PGFP cells engulfed a 15 um bead, 194/760 CAR-PMegf10 and 23/587 CAR-PGFP cells engulfed a 20 um bead (data is pooled from at least three separate experiments). Error bars denote 95% confidence intervals of the mean. *** indicates p<0.0001 respectively by Mann-Whitney test. All scale bars represent 5 µm. https://doi.org/10.7554/eLife.36688.006 To further define the capabilities of the CAR-P, we assessed the capacity of CAR-P-expressing macrophages to engulf variably sized targets. We found that CAR-PMegf10 was able to trigger specific engulfment of beads ranging from 2.5 µm to 20 µm in diameter, with higher specificity above background engulfment being demonstrated for the larger beads (Figure 2b). The high background in this assay is due to heterogeneity in the bilayers on beads purchased from a different manufacturer (Corpuscular) than previous assays. For the 10 µm bead condition, we also tested the phagocytic efficiency of beads containing the endogenous Megf10 ligand, phosphatidylserine. We found that CAR-PMegf10 macrophages engulfed CD19 beads and beads containing 10% phosphatidylserine and the adhesion molecule ICAM-1 at a similar frequency (Figure 2b). This indicates that the CAR-P is comparably efficient to the endogenous system. To determine if the CAR-PMegf10 initiates active signaling at the synapse between the macrophage and target, we stained for phosphotyrosine. Macrophages expressing CAR-PMegf10 exhibited an increase in phosphotyrosine at the synapse, while macrophages expressing CAR-PGFP did not show this enrichment (Figure 3a). Consistent with previous reports, we found that F-actin also was enriched at the cell bead synapse (Figure 3—figure supplement 1). This result suggests that CAR-PMegf10 initiates engulfment through a localized signaling cascade involving tyrosine phosphorylation. Figure 3 with 1 supplement see all Download asset Open asset A phosphorylated ITAM at the cell-target synapse drives engulfment. (A) Macrophages expressing ɑCD19 CAR-PMegf10 (green, top) or ɑCD19 CAR-PGFP were incubated with CD19-ligated beads (position indicated with dotted line), fixed and stained for phosphotyrosine (magenta, top; greyscale, bottom). The fold enrichment of phosphotyrosine at the cell-bead synapse compared to the cell cortex is graphed on the right (n ≥ 11; each dot represents one cell-bead synapse; lines represent the mean ±one standard deviation). (B) Schematic shows the structure of CAR-P constructs in the plot at right. An αCD19 (purple) scFv directs CAR specificity. The intracellular signaling domains from CD3ζ activate engulfment. On the right is a histogram depicting the fraction of macrophages engulfing the indicated number of CD19-coated beads. (C) Comparison showing the average number of beads eaten per cell in J774A.1 macrophages expressing αCD19 CAR-Ps with the indicated intracellular signaling domain. 5 μm silica beads covered with a supported lipid bilayer containing His-tagged CD19 extracellular domain were used as an engulfment target (n = 156–167 cells per condition collected during three separate experiments). Error bars denote 95% confidence intervals and *** indicates p<0.0001 compared to CAR-PGFP control by Kruskal-Wallis test with Dunn’s multiple comparison in correction. (D) Model of the liposome-based fluorescence quenching assay used to determine affinity between the Syk tSH2 domains and the receptor tails of CD3ζ and FcRɣ, two intracellular signaling domains that promote engulfment. Binding between the Syk tSH2 reporter (Syk tSH2), green, and a receptor tail, purple, was detected by rhodamine quenching of BG505 dye on the reporter (see Materials and methods). Kd was determined by assessing mean fluorescence quenching for the last 20 timepoints collected ~45 min after ATP addition over a receptor titration from 0 to 500 nM. Each point represents the mean ± SD from three independent experiments. Kd ± SE was calculated by nonlinear fit assuming one site specific binding. https://doi.org/10.7554/eLife.36688.007 Both successful CAR-P intracellular domains (from FcRɣ and Megf10) have cytosolic Immunoreceptor Tyrosine-based Activation Motifs (ITAMs) that are phosphorylated by Src family kinases. Based on this observation, we hypothesized that the expression of an alternate ITAM-containing receptor might initiate phagocytosis when expressed in macrophages. The CD3ζ subunit of the T cell receptor contains three ITAM motifs. To test if the CD3ζ chain was able to activate phagocytic signaling, we transduced macrophages with the first generation CAR-T (Figure 3b). The CAR-T was able to trigger engulfment of CD19 beads to a comparable extent as CAR-PMegf10 (Figure 3c). In T cells, phosphorylated ITAMs in CD3ζ bind to tandem SH2 domains (tSH2) in the kinase ZAP70. Zap70 is not expressed in macrophages, but Syk, a phagocytic signaling effector and tSH2 domain containing protein, is expressed at high levels (Andreu et al., 2017). Previous work suggested that Syk kinase can also bind to the CD3ζ ITAMs (Bu et al., 1995), indicating that the CAR-T may promote engulfment through a similar mechanism as CAR-PFcRɣ. To quantitatively compare the interaction between SyktSH2 and CD3ζ or FcRɣ in a membrane proximal system recapitulating physiological geometry, we a used liposome-based assay (Figure 3d [Hui and Vale, 2014]). In this system, His10-CD3ζ and His10-Lck (the kinase that phosphorylates CD3ζ) are bound to a liposome via NiNTA-lipids and the binding of labeled tandem SH2 domains to phosphorylated CD3ζ was measured using fluorescence quenching. Our results show that Syk-tSH2 binds the CD3ζ and FcRɣ with comparable affinity (~15 nM and ~30 nM respectively, Figure 3d). Collectively, these results demonstrate that the TCR CD3ζ chain can promote phagocytosis in a CAR-P, likely through the recruitment of Syk kinase. We next sought to program engulfment towards a cellular target. We incubated the CAR-PMegf10 and CAR-PFcRɣ macrophages with cancerous Raji B cells that express high levels of endogenous CD19. Strikingly, the majority of CAR-P-expressing macrophages internalized bites of the target cell (Figure 4a, Figure 4—video 1, 78% of CAR-PMegf10 and 85% of CAR-PFcRɣ macrophages internalized bites within 90 min). The biting phenotype resembles trogocytosis, or nibbling of live cells, which has been reported previously in immune cells (Joly and Hudrisier, 2003). This process was dependent on the ITAM-bearing intracellular signaling domain, as removing the signaling domain (CAR-PGFP) dramatically reduced trogocytosis (Figure 4a). Enrichment of phosphotyrosine at the cell-cell synapse further supports active signaling initiating trogocytosis (Figure 4—figure supplement 1). The CAR-P module also was able to induce trogocytosis in non-professional phagocytes, human NIH 3T3 fibroblast cells (Figure 4—figure supplement 2). This suggests that the CAR-P can promote cancer antigen-dependent engulfment by both professional and non-professional phagocytes. Figure 4 with 7 supplements see all Download asset Open asset CAR-P promotes trogocytosis and whole cell eating. (A) J774A.1 macrophages expressing the ɑCD19 CAR-PMegf10 (top panel, green in merge, left; greyscale, center) engulf pieces of CD19 +Raji B cells (labeled with mCherry-CAAX; magenta in merge, left; greyscale, right). The corresponding control ɑCD19 CAR-PGFP-infected cells are shown below. Arrows point to pieces of ingested Raji B cell. The proportion of CAR-P expressing macrophages internalizing one or more bite within 90 min is quantified on the right. Bites are defined as a fully internalized mCherry-positive vesicle >1 µm in diameter; n = 46 CAR-PMegf10 macrophages, n = 39 CAR-PFcRɣ macrophages and 102 CAR-PGFP macrophages acquired during three separate experiments. (B) Time course of a J774A.1 macrophage expressing CAR-PFcRɣ (green) internalizing a whole Raji B cell labeled with mCherry-CAAX (magenta). These images correspond to frames from Figure 4—video 2. (C) Schematic shows the structure of CAR-Ptandem construct, combining the intracellular signaling domain from FcRɣ and the p85 recruitment domain from CD19. (D) Time course of a J774A.1 macrophage expressing CAR-Ptandem (green) internalizing a whole Raji B cell labeled with mCherry-CAAX (magenta). These images correspond to frames from Figure 4—video 3. (E) Macrophages and Raji B cells were incubated together at a 1:2 macrophage:Raji ratio, and the number of whole Raji B cells eaten per 100 macrophages during 4–8 hr of imaging is graphed. Graph depicts pooled data from four independent experiments; n = 921 CAR-PGFP, n = 762 CAR-PFcRɣ, n = 638 CAR-PPI3K, n = 555 CAR-Ptandem cells. Sample sizes were selected for their ability to detect a 5% difference between samples with 95% confidence. (F) 10,000 macrophages and 20,000 Raji B cells were incubated together for 44 hr. The number of Rajis was then quantified by FACS. 2–3 technical replicates were acquired each day on three separate days. The number of Rajis in each replicate was normalized to the average number present in the GFP-CAAX macrophage wells on that day. * indicates p<0.01, *** indicates p<0.0001 by two-tailed Fisher Exact Test (a and e) or by Ordinary one way ANOVA with Dunnet’s correction for multiple comparisons (f); error bars denote 95% confidence intervals. https://doi.org/10.7554/eLife.36688.009 We next focused on engineering strategies to engulf whole human cancer cells. We observed that macrophages expressing the CAR-PMegf10 or CAR-PFcRɣ were capable of engulfing whole Raji B cells (2 cancer cells eaten per 100 macrophages in a 4–8 hr window for both CAR-PMegf10 or CAR-PFcRɣ, Figure 4b,e, Figure 4—video 2). Whole cell engulfment was infrequent but trogocytosis was robust, suggesting that productive macrophage target interactions were frequently insufficient to trigger whole cell engulfment. To determine if whole cell eating could be enhanced by further opsonization of CD19, we opsonized Raji B cells with a mouse IgG2a anti-CD19 antibody. While addition of this antibody did not trigger additional whole cell internalization, blockade of the ‘don’t eat me’ signal CD47 using the mouse IgG1 anti-human B6H12 clone resulted in a 2.5 fold increase of whole cell eating of opsonized Raji B cells (Figure 4—figure supplement 3). Both endogenous FcR recognition of the anti-CD47 antibody and blockade of CD47 signaling may contribute to this effect. To develop a receptor to enhance whole cell eating, we hypothesized that combining signaling motifs in a tandem array might increase the frequency of whole cell engulfment by specifically recruiting effectors required for the engulfment of large targets. Previous work demonstrated that PI3K signaling is important for internalization of large targets (Schlam et al., 2015). To increase PI3K recruitment to the CAR-P, we fused the portion of the CD19 cytoplasmic domain (amino acids 500 to 534) that recruits the p85 subunit of PI3K to the CAR-PFcRɣ creating a ‘tandem’ CAR (CAR-Ptandem, Figure 4c) (Tuveson et al., 1993; Brooks et al., 2004). A CAR-P containing the p85 recruitment motif alone (CAR-PPI3K) was able to induce some whole cell engulfment, comparable to the CAR-PFcRɣ (Figure 4e). Expression of CAR-Ptandem tripled the ability of macrophages to ingest whole cells compared to CAR-PGFP (6 cancer cells eaten per 100 macrophages, Figure 4d,e, Figure 4—video 3). These data indicate that assembling an array of motifs designed to recruit distinct phagocytic effectors can increase CAR-P activity towards whole cells. To determine if the combination of whole cell eating and trogocytosis was sufficient to drive a noticeable reduction in cancer cell number, we incubated CAR-P macrophages with Raji B cells for two days. After 44 hr of co-culture, we found that CAR-P macrophages significantly reduced the number of Raji cells (Figure 4f). Although the CAR-Ptandem was much more efficient at whole cell eating, the CAR-PFcRɣ performed nearly as well at eliminating Rajis. Importantly, our assay does not distinguish between whole cell engulfment or death following trogocytosis, so it is possible both CAR-P activities are contributing to Raji death rates. Overall, these data suggest that the CAR-P is a successful strategy for directing macrophages towards cancer targets, and can initiate whole cell eating and trogocytosis leading to cancer cell elimination. In summary, we engineered phagocytes that recognize and ingest targets through specific antibody-mediated interactions. This strategy can be directed towards multiple extracellular ligands (CD19 and CD22) and can be used with several intracellular signaling domains that contain ITAM motifs (Megf10, FcRɣ, and CD3ζ). Previous work has suggested that spatial segregation between Src-family kinases and an inhibitory phosphatase, driven by receptor ligation, is sufficient to trigger signaling by the T cell receptor (Davis and van der Merwe, 2006; James and Vale, 2012) and FcR (Freeman et al., 2016). The CAR-Ps that we have developed may similarly convert receptor-ligand binding into receptor phosphorylation of ITAM domains through partitioning of kinases and phosphatases at the membrane-membrane interface. Further development of CAR-Ps could be useful on several therapeutic fronts. Targeting of tumor cells by macrophages has been suggested to cause tumor cell killing (Jaiswal et al., 2009; Majeti et al., 2009; Chao et al., 2010; Jadus et al., 1996), either through directly engulfing cancer cells or by stimulating antigen presentation and a T cell-mediated response (Liu et al., 2015; Tseng et al., 2013). Inhibition of the CD47-SIRPA ‘Don’t eat me’ signaling pathway has also been shown to result in engulfment of cancer cells (Chen et al., 2017; Gardai et al., 2005; Jaiswal et al., 2009; Majeti et al., 2009; Chao et al., 2010). A recent study suggests that CD47 inhibition is most effective when combined with a positive signal to promote target engulfment, which raises the possibility of combining CAR-P expression with CD47 or SIRPA inhibition for an additive effect (Alvey et al., 2017). Although we were able to increase whole cell engulfment by recruiting the activating subunit of PI3K to the phagocytic synapse, the engulfment of larger 20 micron beads was more frequent than the engulfment of whole cells. We hypothesize that this is due to differing physical properties of the engulfment target. Specifically, increased target stiffness has been shown to promote engulfment, suggesting that manipulating the physical properties of the engulfment target could also be a potential strategy for increasing CAR-P efficiency (Beningo and Wang, 2002; Cross et al., 2007). While the CAR-P can engulf whole, viable cancer cells, the ingestion of a piece of the target cell is more common. Trogocytosis, or nibbling of living cells, has also been described in immune cells (Harshyne et al., 2003; Harshyne et al., 2001; Kao et al., 2006; Joly and Hudrisier, 2003; Batista et al., 2001) and brain-eating amoebae (Ralston et al., 2014). In vivo studies also have shown that endogenous dendritic cell populations ingest bites of live tumor cells, contributing to presentation of cancer neo-antigen (Harshyne et al., 2001; Harshyne et al., 2003). Although we were able to use the CAR-P to induce trogocytosis in dendritic cells, we were not able to detect robust cross presentation of the model antigen ovalbumin (Figure 4—figure supplement 4). Thus, although using CAR-Ps to enhance cross presentation of cancer antigen is an intriguing future avenue, such a strategy would likely require more optimization of the dendritic cell subset employed or the CAR-P receptor itself. Overall, our study demonstrates that the CAR approach is transferrable to biological processes beyond T cell activation and that the expression of an engineered receptor in phagocytic cells is sufficient to promote specific engulfment and elimination of cancer cells. Materials and methods Key resources table Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional informationCell line (Mus musculus)J774A.1 MacrophagesUCSF Cell Culture FacilityCell line (Homo sapiens)Raji B CellsOtherObtained from M. McManus, UCSFCell line (Mus musculus)3t3 FibroblastsUCSF Cell Culture FacilityCell line (Mus musculus)C57BL/6JPMID: 21356739Bone Marrow Derived Macrophages (BMDM)Cell line (Mus musculus)C57BL/6JPMID: 7489412Bone Marrow derived Dendritic Cells (BMDC)Cell line (Homo sapiens)HEK293T cellsUCSF Cell Culture FacilityLentivirus productionGenetic Reagent (Mus musculus)OTIPMID: 8287475E. Roberts/M. Krummel Lab UCSFRecombinant DNA reagentCD19-mMegf10 CARthis paperSignal peptide: aa 1–21 CD8 (Uniprot Q96QR6_HUMAN) Extracellular antibody sequence: V-L chain: aa 23–130 anti-CD19 CAR (Genbank AMZ04819) -- GS linker: ggtggcggtggctcgggcggtggtgggtcgggt ggcggcggatct -- V-H chain: aa 148–267 anti-CD19 CAR (Genbank AMZ04819) Stalk/Transmembrane: aa 138–206 CD8 (Uniprot Q96QR6_HUMAN) Cytosolic sequence: aa 879–1147 Mouse Megf10 (Uniprot Q6DIB5 (MEG10_MOUSE)) Fluorophore: mGFPRecombinant DNA reagentCD19-FcGamma CARthis paperSignal peptide: aa 1–21 CD8 (Uniprot Q96QR6_HUMAN) Extracellular antibody sequence: V-L chain: aa 23–130 anti-CD19 CAR (Genbank AMZ04819) -- GS linker: ggtggcggtggctcgggcggtggtgggtcgg gtggcggcggatct -- V-H chain: aa 148–267 anti-CD19 CAR (Genbank AMZ04819) Stalk/Transmembrane: aa 138–206 CD8 (Uniprot Q96QR6_HUMAN) Cytosolic sequence: aa 19–86 Mouse Fc ERG precursor (Uniprot P20491 (FCERG_MOUSE)) Fluorophore: mGFPRecombinant DNA reagentCD19-empty CARthis paperSignal peptide: aa 1–21 CD8 (Uniprot Q96QR6_HUMAN) Extracellular antibody sequence: V-L chain: aa 23–130 anti-CD19 CAR (Genbank AMZ04819) -- GS linker: ggtggcggtggctcgggcggtggtgggtcggg tggcggcggatct -- V-H chain: aa 148–267 anti-CD19 CAR (Genbank AMZ04819) Stalk/Transmembrane: aa 138–206 CD8 (Uniprot Q96QR6_HUMAN) Cytosolic sequence: basic linker NHRNRRR (nucleotide AACCACAGG AACCGAAGACGT) Fluorophore: mGFPRec

Chimeric antigen receptors (CARs) are synthetic receptors that reprogram T cells to kill cancer. The success of CAR-T cell therapies highlights the promise of programmed immunity and suggests that applying CAR strategies to other immune cell lineages may be beneficial. Here, we engineered a family of Chimeric Antigen Receptors for Phagocytosis (CAR-Ps) that direct macrophages to engulf specific targets, including cancer cells. CAR-Ps consist of an extracellular antibody fragment, which can be modified to direct CAR-P activity towards specific antigens. By screening a panel of engulfment receptor intracellular domains, we found that the cytosolic domains from Megf10 and FcRɣ robustly triggered engulfment independently of their native extracellular domain. We show that CAR-Ps drive specific engulfment of antigen-coated synthetic particles and whole human cancer cells. Addition of a tandem PI3K recruitment domain increased cancer cell engulfment. Finally, we show that CAR-P expressing murine macrophages reduce cancer cell number in co-culture by over 40%.

Metastasis is the leading cause of cancer-related deaths. Recent developments in cancer immunotherapy have shown exciting therapeutic promise for metastatic patients. While most therapies target T cells, other immune cells, such as monocytes, hold great promise for therapeutic intervention. In our study, we provide primary evidence of direct engagement between human monocytes and tumor cells in a 3D vascularized microfluidic model. We first characterize the novel application of our model to investigate and visualize at high resolution the evolution of monocytes as they migrate from the intravascular to the extravascular micro-environment. We also demonstrate their differentiation into macrophages in our all-human model. Our model replicates physiological differences between different monocyte subsets. In particular, we report that inflammatory, but not patrolling, monocytes rely on actomyosin based motility. Finally, we exploit this platform to study the effect of monocytes, at different stages of their life cycle, on cancer cell extravasation. Our data demonstrates that monocytes can directly reduce cancer cell extravasation in a non-contact dependent manner. In contrast, we see little effect of monocytes on cancer cell extravasation once monocytes transmigrate through the vasculature and are macrophage-like. Taken together, our study brings novel insight into the role of monocytes in cancer cell extravasation, which is an important step in the metastatic cascade. These findings establish our microfluidic platform as a powerful tool to investigate the characteristics and function of monocytes and monocyte-derived macrophages in normal and diseased states. We propose that monocyte-cancer cell interactions could be targeted to potentiate the anti-metastatic effect we observe in vitro, possibly expanding the milieu of immunotherapies available to tame metastasis.