Redirecting Macrophage Function to Sustain their ‘Defender’ Antitumor Activity

Abstract

Macrophages are multi-functional innate immune cells that occupy normal or pathologic tissues, including cancer. The importance of macrophage ontogeny and the transcriptional networks underlying their functional diversity are underappreciated in immuno-oncology. Here, we discuss the implications of these fundamental characteristics for therapeutically reprogramming macrophages to sustain their tumoricidal activities.

Macrophages are an indispensable component of mammalian tissue that support homeostasis, mediate host defense, and facilitate tissue repair during inflammation, infection, and cancer (Guilliams et al., 2020). We will focus on macrophages in cancer given their abundance in tumor tissue and a longstanding enthusiasm in the field to exploit their biology for therapeutic purposes. Macrophages notably arise from two major sources, embryonic and bone marrow, which collectively contribute to their ample presence within sites of primary tumor growth and metastasis (Cassetta and Pollard, 2018). The cell contact-dependent or -independent cues that macrophages receive within these tumor microenvironments (TMEs) activate signaling and transcriptional networks, which in turn direct their functional diversity comprising a continuum of phenotypes spanning from antitumor ‘defenders’ to pro-tumor ‘remodelers’ (Figure 1A). Understanding mechanisms that direct or redirect macrophage ‘programming’ from one functional state to another is key to uncovering novel therapeutic targets to sustain their defender activity and maximal clinical response.

Figure 1. Strategies for Reprogramming Macrophage Phenotypes toward Defender Activity.

(A) Schematic of macrophage antitumor defender and pro-tumor remodeler phenotypes within the TME. Defender macrophages mediate antitumor activity through both direct (i.e., via phagocytosis or the release cytotoxic molecules) and indirect mechanisms (i.e., via the activation of innate and adaptive immunity). Remodeler macrophages facilitate tumor growth and progression also through direct (i.e., production of tissue invasive and angiogenic factors) and indirect mechanisms (i.e., via the secretion of immune suppressive cytokines or mediators). Reprogramming macrophages from one functional state to another is of substantial therapeutic interest, with the goal to rebalance the overall macrophage response toward a defender phenotype.

(B) Schematic of molecular targets of macrophage reprogramming agents. Efforts to reprogram macrophages toward a defender phenotype have focused on several major fronts, ranging from modifying the extracellular environment to the intracellular nuclear compartment. In addition, other approaches include targeting cell surface receptors and cytoplasmic enzymes, as shown.

(C) List of parameters for designing effective macrophage reprogramming therapies. The ability to optimally reprogram macrophages toward a defender phenotype is complex and reflects several important parameters. Several examples are illustrated, which include consideration of the relevant macrophage population(s) (i.e., TRM vs. BMDM), the platform and method of delivery of the reprogramming agent, and rational combinations with other anticancer therapies.

Created with BioRender.com

Defining the Ideal Macrophage Defender Phenotype for Antitumor Activity

Diverse transcriptional networks direct the different macrophage functional states, which culminate from their communications with other cell types within their extracellular environments. For some time, antitumor macrophages have been coined M1, conventionally activated or pro-inflammatory, while pro-tumor macrophages have been termed M2, alternatively activated or anti-inflammatory with wound healing properties. It is becoming clear that these terminologies do not necessarily inform the function of macrophages in cancer. For example, while the scavenger receptors CD163 and CD206 are considered synonymous with an “M2” phenotype, they are not M2-specific and may be expressed more broadly by macrophages. Moreover, most studies in macrophage-tumor biology loosely characterize macrophages within the TME as tumor-associated macrophages (TAMs), lessening appreciation for potential differences in macrophage ontogeny and function, as well as how these differences may impact tumor outcome. As with an M2 classification, TAM nomenclature is often assigned to pro-tumor behavior and nondescript cell surface receptor expression, which does not provide enough insights into precise macrophage functional status (Guerriero, 2018). Terminology that better characterizes macrophage functionality within the TME may reduce ambiguity associated with these macrophage classifications. Therefore, terminology that is more reflective of macrophage functionality may be more appropriate, such as antitumor defenders or pro-tumor remodelers (Figure 1A).

Both defender and remodeler macrophage phenotypes would be characterized by distinct effector and regulatory functions. The effector functions of defender macrophages include tumor cell destruction via phagocytosis or the production of toxic mediators such as nitric oxide (NO) and tumor necrosis factor (TNF)-α. The immunoregulatory functions of defender macrophages involve activating innate (NK) and adaptive (T cell) immunity through the production of cytokines, such as interleukin-12 (IL-12), or the presentation of antigen (Ag) in the context of major histocompatibility (MHC) class I or II molecules for immune recognition and activation. In contrast, the effector functions of remodeler macrophages include facilitating tumor growth and metastasis through the production of tissue invasive (e.g., matrix metalloproteinases (MMPs)) and angiogenic (e.g., vascular endothelial growth factor (VEGF)) mediators. The immunoregulatory functions of remodeler macrophages entail inhibition of innate and adaptive immunity through the production of suppressive factors, namely IL-10, transforming growth factor (TGF)-β and prostaglandin E₂ (PGE₂) (Guerriero, 2018). Thus, the defender or remodeler classification, which underscores function, represent less ambiguous ways to characterize macrophage responses within TMEs.

Distinguishing Tissue-resident from Bone Marrow-derived Macrophages to Identify the Most Relevant Defender within the TME

Whether the macrophage response exhibits characteristics of a defender or remodeler phenotype, the outcome may culminate from two major sources of macrophages: (1) those that dwell naturally within organs or tissues known as tissue-resident macrophages (TRMs), which originate from embryonic or hematopoietic progenitors; and (2) those that arise from the bone marrow known as bone marrow-derived macrophages (BMDMs), which may also serve to replenish aging TRMs of particular tissues (Guilliams et al., 2020). Interestingly, in cancer, BMDMs abundantly infiltrate sites of neoplastic disease, often outnumbering TRMs (Cassetta and Pollard, 2018). Furthermore, few reports clearly distinguish the functional contributions of TRMs versus BMDMs within the TME to tumor outcome. One example is the work of DeNardo and colleagues, who demonstrated that TRMs of embryonic origin facilitate pancreatic cancer progression via tissue remodeling activities, including collagen deposition. Within the same tumor model, BMDMs more efficiently present Ag to the immune system compared to TRMs (Zhu et al., 2017). Thus, defining the contributions of BMDMs versus TRMs to tumor outcomes should remain an active area of investigation.

Part of the reason for the inability to effectively distinguish TRMs from BMDMs within TMEs for in-depth functional evaluation is due to an incomplete understanding of unique cell surface markers or gene expression signatures. Recent work by Pollard, Coussens, and colleagues is helping to fill this void in BMDM biology. Their study identified a 37-gene transcriptional signature of BMDMs within human breast tumors associated with more aggressive disease. This analysis further revealed specific upregulation of SIGLEC1 on BMDMs, also associated with worse disease-specific survival. Studies were extended to human endometrial cancer, which indicated that transcriptional networks of BMDMs vary depending on the cancer type and anatomical location (Cassetta et al., 2019). Similarly, in a separate study, the Joyce laboratory identified differential expression of ITGA4 on BMDMs within glioma tumors, associated with high-grade glioma (Bowman et al., 2016). Altogether, such studies underscore the translational impact of a differentially expressed BMDM surface marker and/or gene expression profile on cancer prognosis.

As noted earlier, both TRMs and BMDMs have the capacity to exert defender roles, as the expression of cytotoxic molecules and immune activation networks are evident in TRMs, such as microglia, and BMDMs (Bowman et al., 2016). Ag presentation pathways are more commonly expressed by BMDMs (Cassetta et al., 2019; Zhu et al., 2017). The discovery of population- or subset-defining functional markers, including transcription factors, may identify macrophages within the TME that express defender activities. Macrophages co-expressing MHC class II (HLA-DR) and cytotoxic molecules (TNF-α or inducible nitric oxide synthase (iNOS), the enzyme that leads to NO production) were associated with increased survival in lung cancer patients (Ohri et al., 2009). Our laboratory revealed that high expression of interferon regulatory factor-8 (IRF8), a transcriptional regulator of immune activation and anti-microbial activities, in macrophages was associated with improved outcomes in patients with renal cell carcinoma (Muhitch et al., 2019). Such studies underscore the implications of the identification of both functional traits and transcriptional network profiles of defender macrophages to prognostication. Whether such knowledge of defender macrophage status may be utilized to augment response to antitumor therapy remains to be investigated. Unfortunately, such TRM- or BMDM-mediated defender properties evolve in cancer and are outweighed by remodeler activities that enhance tumor cell invasion and angiogenesis (Cassetta et al., 2019; Zhu et al., 2017). Through reversible mechanisms of functional programming, however, remodeler macrophage populations may be therapeutically targeted via strategies to induce or maintain defender phenotypes (Figure 1A).

Reprogramming Macrophages to Achieve Defender Activity

Therapies that inhibit macrophage recruitment or cause macrophage depletion may exert toxicities that increase susceptibility to infection and impair homeostatic macrophage functions. Alternatively, reprogramming macrophages toward defender phenotypes may be a more beneficial strategy to sustain immune surveillance against cancer. Current reprogramming strategies target receptors, enzymes, or transcription factors (Cassetta and Pollard, 2018) (Figure 1B). Engaging pathogen recognition receptors (PRRs) or CD40 costimulatory molecules may induce macrophage tumoricidal activities (Rodell et al., 2018). The phagocytic capacity of macrophages may be enhanced by relieving SIRPα activation through blocking its cognate receptor, CD47 expressed by tumor cells. The Weissman laboratory also identified PD-1 expression by BMDMs in colon cancer models and showed that PD-1 blockade restores the ability of macrophages to mediate tumor cell phagocytosis (Gordon et al., 2017). Inhibiting the expression of certain kinases in macrophages, namely PI3Kγ or RIP1, may reduce the production of immune suppressive cytokines (IL-10 or TGF-β) while enhancing MHC class II levels and antitumor T cell responses (Kaneda et al., 2016; Wang et al., 2018). Interestingly, nutrient supplementation has been shown to mediate macrophage reprogramming effects. Intratumoral administration of nanoparticles containing iron oxide enhanced the expression of macrophage costimulatory molecules (CD86) or cytotoxic mediators (TNF-α) (Zanganeh et al., 2016). Macrophage activity may also be modulated by certain chemotherapeutic regimens; for example, in lung cancer models, the combination of oxaliplatin and cyclophosphamide activated macrophages to enhance the mobilization of adoptively transferred antitumor T cells to the TME through the production of T cell-recruiting chemokines (Srivastava et al., 2021). Attention should be considered with the use of reprogramming strategies, however, that act in non-specific cellular manners to minimize adverse global effects.

Transcription factors are emerging as novel targets of macrophage reprogramming to achieve macrophage defender roles directly at the DNA level, rather than targeting cell surface or cytoplasmic compartments (Figure 1B). Nuclear targeting offers a unique approach to maximize activation of gene expression programs required for multiple macrophage defender activities. Recently, the Stephan laboratory developed a nanoparticle-based technology to deliver IRF5, an immune stimulatory transcriptional regulator analogous to IRF8, to macrophages within ovarian tumors. Administration of IRF5 to macrophages using this delivery platform increases MHC class II expression, IL-12 and IFN-γ production and cytotoxic molecule (TNF-α) secretion, while doubling overall survival in mouse tumor models (Zhang et al., 2019). Nanocarrier delivery strategies will likely be most efficacious when targeting specific receptors or mechanisms differentially expressed by TRMs or BMDMs, rather than pan-macrophage cell surface markers. The route of nanoparticle administration may be adapted for local intratumoral delivery or systemic delivery to inhibit primary or metastatic tumor growth. Nanoparticle delivery of transcriptional regulators is thus a new and provocative avenue to safely and effectively augment macrophage defender activity.

Considerations to potentially heighten therapeutic efficacy of macrophage reprogramming strategies include: (1) defining pathways or mechanisms for selective targeting of TRMs and/or BMDMs; (2) constructing effective delivery platforms to target those specific populations; (3) determining the most relevant route of agent administration that targets those populations within particular TMEs; (4) identifying the amount of time required to induce macrophage reprogramming effects; and (5) designing rational combinations of macrophage reprogramming approaches with other modes of antitumor therapy (Figure 1C). Precision targeting of TRMs or BMDMs, based on distinguishing cell surface marker or effector characteristics, will enhance reprogramming efficacy of the most relevant macrophage population. In addition, specifically targeting the reprogramming agents to the most relevant macrophage population will likely minimize dosing requirements to achieve greater therapeutic effects. Constructing an ideal delivery platform will facilitate effective macrophage targeting while reducing off-target effects or toxicities observed with pan-macrophage approaches. Nanocarriers are a novel platform to selectively target macrophages for delivery of encapsulated reprogramming agents. Reprogramming agents may be administered locally or systemically to target macrophages within the primary or metastatic TME. Finally, macrophage reprogramming agents may be combined with other forms of antitumor therapy to potentiate macrophage immune stimulatory and antitumor properties (Kaneda et al., 2016; Rodell et al., 2018; Srivastava et al., 2021; Zhang et al., 2019). However, timing the administration of each agent may be critical for maximizing therapeutic response. Ideally, macrophage reprogramming agents may be administered first to induce defender macrophage properties that may synergize with subsequent immunotherapy or standard-of-care therapy. Sustaining the defender macrophage phenotype, which may be achieved through a persistent delivery of the reprogramming agent, may maximize response to combination therapies, prevent tumor recurrence or metastasis, and prolong overall survival.

Next Steps

It is now clear that the macrophage response in cancer is highly complex and does not simply reflect a single functional state, a single population, nor a single TME. Understanding the precise nature of each of these characteristics is critical to improving not only an understanding of macrophage-cancer biology, but also ways to better exploit the clinical use of macrophages in prognosis and therapy. Macrophage functional states vary from defenders to remodelers; macrophage populations vary from TRMs to BMDMs; and relevant TMEs vary from primary to distal sites. Of these major determinants, the one most amenable to experimental or clinical exploitation is functional state. Indeed, the macrophage defender phenotype is critical for tumor control. Therefore, intensive efforts should continue to identify the most relevant cell surface markers, effector mechanisms, and unique transcriptional profiles of defender macrophages to utilize for prognostication and to design effective, durable therapeutic interventions. While both TRMs and BMDMs may exhibit defender or remodeler phenotypes, more detailed studies are needed to fully appreciate the relative contributions of macrophage ontogeny to tumor outcome. Also, although select BMDM markers are now emerging to bear significant prognostic value, the discovery of unique TRM-specific surface markers is lagging and their impact on prognosis remains less clear (Bowman et al., 2016; Cassetta et al., 2019). TRMs are noteworthy as they already reside in potential sites of neoplastic growth and thus may be better exploited not only for therapeutic purposes, but also preventative measures. Furthermore, while progress is being made in the design and application of strategies to reprogram macrophages for therapy, most target and impact individual genes, pathways, or mechanisms. Alternatively, rather than targeting a single or limited set of biologic events or processes, utilizing innovative strategies to deliver transcription factors, which regulate the expression of many genes, may represent the next generation of clinical approaches. This reprogramming paradigm is thus likely to impact the expression of a broader repertoire of genes that direct the complex biology of the optimal macrophage defender.