QMCLab - REVIEWS · 2020. 10. 10. · genetic modification and returning these modified cells to...

The critical relationship between immune function and cancer was first proposed by Rudolf Virchow 150 years ago when he observed the prevalence of leukocytes in tumours1. A few decades later, William Coley and col-leagues conducted the first experimental immunotherapy treatment of cancer based on the observation that some patients carrying a bacterial infection exhibited more extensive tumour regression compared with uninfected individuals2. For the next 100 years, there was little clini-cal demonstration that the immune system could be mobilized to provide a reproducible benefit for patients with cancer. Recent watershed moments evincing clinical success for cancer immunotherapy include the regulatory approvals of the T cell checkpoint inhibitory antibodies ipilimumab (Yervoy; Bristol-Myers Squibb), pembrolizumab (Keytruda; Merck) and nivolumab (Opdivo; Bristol-Myers Squibb), and the dendritic cell therapy sipuleucel-T (Provenge; Dendreon).

Sipuleucel-T comprises autologous dendritic cells primed with a recombinant protein to enhance the patient’s native immune response in prostate cancer3. Ipilimumab is a fully human IgG1 mAb that blocks ligand engagement and enhances T cell activation by directly binding to the cytotoxic T lymphocyte-associated anti-gen 4 (CTLA4) receptor protein, thus blocking a critical

inhibitory signal for activated T cells4,5. Pembrolizumab and nivolumab are humanized mAbs that block ligand engagement, thus interfering with T cell signalling and cell death6,7. All of these mAbs have demonstrated sig-nificant clinical benefit as single agents in melanoma, and continue to be tested in clinical trials as both single agents and in combinations. Moreover, these mAbs show considerable promise for other solid tumour indi-cations, including lung cancer and renal cell carcinoma. These successful immuno-oncology medicines lever-age distinct immunological mechanisms in relation to cancer growth and treatment, ranging from early tumour-associated antigen responses to T cell activation and relief of tumour-influenced immunosuppression. More direct means of T cell activation are also showing early clinical success and include harnessing patient T cells for ex vivo genetic modification and returning these modified cells to the patient8–13 (TABLE 1).

Current cancer immunotherapy strategies seek to reverse immune tolerance either by modulating T cell co-receptor signals or boosting the recognition of tumour-associated antigens by using native biomol-ecules or mAbs. Although the targeting of protein–protein interactions or their signalling pathways is not currently amenable to small-molecule approaches,

GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, Pennsylvania 19426, USA. Correspondence to A.H. e-mail: [email protected]:10.1038/nrd4596Published online 31 July 2015

Dendritic cellsProfessional antigen-presenting cells that take up and present antigens. In the tumour microenvironment these cells present antigens from dying tumour cells, which are taken up and processed by immature dendritic cells. Upon cell maturation, and as they migrate to the draining lymph node, they display antigen via HLA class I and II molecules to prime effector T lymphocytes.

Big opportunities for small molecules in immuno-oncologyJerry L. Adams, James Smothers, Roopa Srinivasan and Axel Hoos

Abstract | The regulatory approval of ipilimumab (Yervoy) in 2011 ushered in a new era of cancer immunotherapies with durable clinical effects. Most of these breakthrough medicines are monoclonal antibodies that block protein–protein interactions between T cell checkpoint receptors and their cognate ligands. In addition, genetically engineered autologous T cell therapies have also recently demonstrated significant clinical responses in haematological cancers. Conspicuously missing from this class of therapies are traditional small-molecule drugs, which have previously served as the backbone of targeted cancer therapies. Modulating the immune system through a small-molecule approach offers several unique advantages that are complementary to, and potentially synergistic with, biologic modalities. This Review highlights immuno-oncology pathways and mechanisms that can be best or solely targeted by small-molecule medicines. Agents aimed at these mechanisms — modulation of the immune response, trafficking to the tumour microenvironment and cellular infiltration — are poised to significantly extend the scope of immuno-oncology applications and enhance the opportunities for combination with tumour-targeted agents and biologic immunotherapies.

REVIEWSNature Reviews Drug Discovery | AOP, published online 31 July 2015; doi:10.1038/nrd4596

NATURE REVIEWS | DRUG DISCOVERY ADVANCE ONLINE PUBLICATION | 1

© 2015 Macmillan Publishers Limited. All rights reserved

mailto:[email protected]

Effector T cellMediates killing of target cells via cognate antigen recognition on HLA class I molecules.

Regulatory (TReg) cellCD4+CD25+FOXP3+ cells that are suppressive in nature and dampen CD8+ effector T cell responses.

there are many alternative processes and pathways that are open to intervention by traditional small-molecule drugs. Small-molecule drugs offer the following dis-tinct advantages over recombinant protein approaches to medicine design: 1) high feasibility due to a detailed understanding and historical precedent of their clinical application and development; 2) oral bioavailability; 3) greater exposure within the tumour microenviron-ment or across physiological barriers (for example, the blood–brain barrier); 4) access to intracellular disease targets not tractable by protein therapeutic agents; and 5) diverse and well-understood formulation and dosing options to alleviate pharmacokinetic and/or pharmaco-dynamic challenges and enabling titration of drug expo-sure. Another compelling advantage of small-molecule drugs is that patients can more easily access these types of medicines compared to biological immunotherapies. Small-molecule drugs are typically lower in cost due to the goods themselves, the reduced cost of delivering a pill versus an infusion (or injection) and a simpler supply chain that requires no refrigeration. For immunotherapy to reach its full potential in transforming the treatment of cancer, costs must come down to provide access to all those who could benefit. Even when cost is not a barrier to access, the convenience of an oral medication versus an infusion is a powerful point of differentiation from the currently evolving immuno-oncology therapies.

As discussed in the subsequent section, an effective cancer immunotherapy needs to exploit subtle differ-ences in the appearance of tumour cells versus normal cells (both originating from self ), thus enabling an effective immune response to be mounted, as well as

to overcome immunosuppressive mechanisms at the tumour site. Approved T cell checkpoint inhibitors enhance the antitumour immune response through distinctly different mechanisms of action. For example, CTLA4 blockade predominantly enhances T cell activa-tion during the priming phase of the immune response, whereas programmed cell death protein 1 (PD1) block-ade appears to release exhausted but otherwise activated effector T cell populations and reduce regulatory T (TReg) cell function. Because of the various steps required for an effective T cell-mediated antitumour immune response and the associated immunosuppressive mechanisms, a limited immune response or active immune sup-pression can result in immune escape and continued tumour growth. Conversely, facilitating or enhancing any of these steps or controlling immune suppression has the potential to promote a more effective antitumour immune response.

The potency of T cell-based interventions is further illustrated by autologous cell and gene therapy approaches. For these approaches, chimeric antigen receptors or cancer target-specific T cell receptors (TCRs) are transduced and expressed in a patient’s T cells to render them tumour-specific; these T cells are then re-infused into the patient13,14. Another approach is to deliver antigen or broad immune activators directly into the tumour. Studies have shown that intratumoural treatments can lead to the eradication of distal tumours (abscopal effect) and require much lower doses than systemically delivered agents15. Clinical investigators recognizing these multiple points of opportunity for intervention have already initiated combination studies

Table 1 | Therapeutic modalities targeting immune regulation of cancer

Modality General use or utility Limitations Examples Status Refs

Vaccines Prime patient immune response to tumour-specific antigens

Heterogeneous tumour antigen composition and expression; prone to be hampered by mechanisms of immune suppression

Vaccines against targets such as gp100, MUC1, MAGEA3

Various approaches in clinical trials

9

Recombinant cytokines

Agonism or blockade of protein–protein immune pathways

Antigenicity; poor pharmacokinetics; high toxicity

GM-CSF, IL-7, IL-12, IL-15, IL-18, IL-21

IL-2 for metastatic melanoma and renal cell carcinoma, and IFNα for the adjuvant therapy of stage III melanoma are approved

8

mAbs Highly selective agonism or blockade of extracellular protein–protein immune pathways; long half-life; non-immunogenic (human or humanized)

Expensive and time-consuming manufacturing and development costs; challenges in achieving high tumour exposures

mAbs targeted against CTLA4, PD1, PDL1 (T cell checkpoint blockers)

Ipilimumab (CTLA4-specific), nivolumab and pembrolizumab (both PD1-specific) are approved for melanoma; others in clinical development for melanoma, lung cancer, kidney cancer and other diseases

11

Autologous T cells

Tumour-targeted cytotoxicity of extracellular and intracellular tumour-specific antigens

Heterogeneous tumour antigen composition and expression; on-target, off-tumour toxicity

CAR T cells, TCR T cells

None approved, but several in clinical development

10

Small molecules

Uniquely suited for intracellular targets, but also equally applicable to cell surface or extracellular targets

Off-target activities; dose-limiting toxicities; ineffective at blocking protein–protein interactions; require daily dosing

IDO1 and COX2 inhibitors, TLR agonist and chemokine antagonist

Topical imiquimod (a TLR7 and TLR8 agonist) approved for the treatment of basal cell carcinoma; IDO inhibitors in clinical trials

12

CAR, chimeric antigen receptor; COX2, cyclooxygenase 2; CTLA4, cytotoxic T lymphocyte-associated antigen 4; GM-CSF, granulocyte–macrophage-colony-stimulating factor; IDO1, indoleamine 2,3-dioxygenase 1; IFNγ, interferon-γ; IL, interleukin; MAGEA3, melanoma-associated antigen 3; MUC1, mucin 1; PD1, programmed cell death protein 1; PDL1, programmed cell death 1 ligand 1; TCR, T cell receptor; TLR, Toll-like receptor.

R E V I E W S

2 | ADVANCE ONLINE PUBLICATION www.nature.com/reviews/drugdisc


tooo1高亮

Nature Reviews | Drug Discovery

Dendritic cell

Monocyte

B cell

CD8+ effector T cell TReg cell

M1 macrophage

MDSC

TAMCAF

Fibroblast

Tumour antigen

Tumour cell (live)

Tumour cell (dead)

Antibody

Bonemarrow

Lymph node

Lymphatic vessels

Tumour microenvironment

Blood vessels

NK cell

CD4+ T helper cell

M2 macrophage

Natural killer cellsCells that kill viral and tumour targets via non-MHC restriction.

of checkpoint inhibitors with other agents. The various stages and cell types involved in cancer immunity pro-vide a rich source of potential points of intervention for small-molecule drugs; such potential points include extracellular enzymes and receptors, and intracellular signal transduction pathways. Here, we describe the rationale for using small-molecule drugs as immuno-therapies in cancer, and offer considerations regarding the future promise of their utility both as single agents and in combination with other cancer medicines.

Immunity in the tumour microenvironmentAlthough tumour-associated antigens are self-antigens, they can be immunogenic. These antigens can be mutated and therefore unique (for example, β-catenin in mela-noma), overexpressed (for example, mucin 1 (MUC1) and HER2/neu (also known as receptor tyrosine-protein kinase erbB-2; ERBB2) in colon cancer, breast cancer and lung cancer), or germ line antigens that are silent in normal tissue but expressed in cancers (for example, NY-ESO-1 in melanoma and synovial sarcoma). Tumour-associated antigens can also be differentiation antigens (for example, melanoma antigen recognized by T cells 1 (MART1; also known as Melan-A), gp100 and tyrosinase in melanoma) or viral antigens (for example, hepatitis B virus in liver cancer). Several studies have reported that tumour-associated antigens can induce

T cell responses16 and the production of antibodies17, and cases of spontaneous remission are well documented18, indicating that with the optimum balance of immune effectiveness over immune suppression, the immune system can fight cancer.

The cycle of generating an antitumour response is complex and is exquisitely interdependent on several closely knit factors (FIG. 1). Indeed, there is a precise interplay between the tumour, antigen-presenting cells and T lymphocytes in the appropriate cytokine milieu. Professional antigen-presenting cells, such as immature dendritic cells, engulf antigens from dying tumour cells. As these dendritic cells migrate to the draining lymph node, they process the antigen and mature into den-dritic cells that can present the antigen via their MHC class I and class II molecules on their cell surface. At the draining lymph node, the dendritic cells encounter CD4+ T helper (TH) and CD8

+ effector T cells via engagement through MHC-TCR interactions, leading to T cell prim-ing and activation. As activated T cells emigrate from the lymph node and are engaged in T cell surveillance, they encounter tumour cells that bear the cognate peptide via MHC class I molecules on the cell surface and initiate a programme of events leading to tumour cell death. Natural killer cells, which represent one part of the innate immune system, are also capable of surveillance and tumour lysis in a more rapid but less specific fashion19.

Figure 1 | Immune function and the immune response to cancer. During an immune response to a cancerous lesion, regions of survival or new cell growth (beige cells) coexist with cells undergoing cell death (light grey cells). Dendritic cells invade tumours, engulf and present tumour antigens (red spheres) to a myriad of T cell types (CD4+ helper and CD8+ effector T cells, regulatory T (T

Reg) cells, natural killer (NK) cells)

for their activation in tumour draining lymph nodes. Activated CD4+ and CD8+ effector T cells and NK cells return to the tumour bed through the peripheral blood and contribute to further tumour regression based on

tumour-antigen priming. TReg

cells similarly return to the tumour and suppress effector T cell and NK cell killing efficiency to guard against overt inflammation and neighbouring normal tissue damage. Myeloid-derived suppressor cells (MDSCs) further contribute to this regulation and can indirectly contribute to tumour survival and growth. B cells produce antibodies to tumour antigens and thereby contribute to tumour eradication. Macrophages invade and differentiate into both tumour-promoting (M2) and tumour-suppressive (M1) lineages. TAM, tumour-associated macrophage.

R E V I E W S



Tumour-associated macrophages(TAMs). Arise from anti-inflammatory, pro-tumorigenic M2 macrophages and reside in the tumour stroma, causing inhibition of immune responses, or in blood vessels in the core of the tumour tissue, promoting tumour invasion.

TH1-type responseA CD4+ T cell immune response mediated by pro-inflammatory cytokines (e.g., interferon, interleukin-1 and tumour necrosis factor), which promotes cellular immune responses.

TH2-type responseA CD4+ T cell response driven by interleukin-4, which stimulates antibody production.

Myeloid-derived suppressor cells(MDSCs). Cells that promote immune suppression of multiple cell types and initiate tumour cell evasion.

Cancer-associated fibroblastsComponents of tumour tissue that support and promote tumour growth and immune cell evasion.

Immunogenic cell deathCell death that primes an immune response; characterized by release of ATP, high mobility group box 1 (HMGB1) and the pre-apoptotic display of calreticulin.

CD39Also known as ectonucleoside triphosphate diphosphohydrolase 1 (NTPDase 1), CD39 is a cell surface-bound phosphatase found on lymphocytes and tumours and is responsible for the conversion of extracellular ATP into adenosine.

CD73Also known as 5’-nucleotidase (5’-NT), CD73 is a cell surface-bound phosphatase found on lymphocytes and tumours and is responsible for the conversion of extracellular ATP into adenosine.

Although the process of effector T cell activation followed by their migration to the tumour site, recognition of the cognate antigen and final cytolysis of the tumour appears straightforward, the effectiveness of an antitumour response at the tumour site is achieved only when there is no break in this cycle and when immunosuppressive mechanisms are controlled.

The tumour microenvironment contains a complex mixture of cells, including tumour cells, infiltrating immune cells and stromal cells with vasculature. These diverse immune cell types operate in concert with one another to either eradicate tumours using activated immune effectors or to promote immune evasion that ultimately leads to tumour growth and eventually metas-tasis20. Tissue-resident macrophages, which are the first line of defence against a physiological insult, exhibit a level of plasticity whereby pro-inflammatory but antitumori-genic M1 macrophages convert to anti-inflammatory and pro-tumorigenic M2 macrophages21. M2 macrophages are tumour-associated macrophages (TAMs) that aid in neo-vasculature formation and subsequently invasion and metastasis. M1 macrophages typically produce a TH1-type response in contrast to M2 macrophages, which initiate a TH2-type response. Tissue hypoxic conditions can be a factor for the conversion of M1 to M2 macrophages; the core of the tumour tissue is associated with hypoxia rela-tive to the periphery and is also less accessible by immune effectors22.

The tumour microenvironment contains other cell types that make the environment suppressive to immune effectors. TReg cells and myeloid-derived suppressor cells (MDSCs)23,24 have evolved as immune mechanisms to counter autoimmunity and stress-related inflamma-tion. These two cell types inhibit T cell antigen-specific and nonspecific immune responses at the tumour site. In cancer, their high numbers in the circulation cor-relate with staging of the disease and poor prognosis25. TReg cells are heterogeneous in their phenotypic and cytokine profiles and constitute different subsets based on where they reside26. Peripheral TReg cells, also called natural TReg cells, are characterized by the expression of the transcription factor forkhead box P3 (FOXP3). By contrast, tissue-bound TReg cells are called acquired, adaptive or induced TReg cells or TR1 TReg cells and are CD4+CD25–FOXP3+ (REF. 27). Cancer-associated fibroblasts are found in abundance in the tumour stroma and are distinct from normal fibroblasts. Cancer-associated fibro-blasts are maintained and activated by growth factors, such as transforming growth factor-β (TGFβ) and fibro-blast growth factor, and they promote tumourigenesis. These cells aid in tumour proliferation, vascularization, survival and metastasis28.

Given the multiple immunosuppressive mecha-nisms that are engaged to promote tumour survival and growth, it is believed that targeting these suppres-sive mechanisms will be required to most effectively achieve effector T cell activation and functionality in the tumour microenvironment. Effective interventions would probably cause TReg cell depletion or a reduced function at the tumour site to allow for effector T cell efficacy. The identification of targets on T cells that can

potentiate a robust immune response has paved the way for effective immunotherapies, as noted in the introduction, with examples targeting CTLA4 and PD1 checkpoint receptors.

Targets for small-molecule interventionArmed with an improved mechanistic understanding of the processes that allow tumours to escape immune surveillance, the mechanisms of action of the current stand-ard-of-care anticancer drugs have been re-examined29. These investigations have revealed that tumour cell death promoted by some antitumour agents (for example, doxorubicin and cyclophosphamide) results in effec-tive antigen presentation and priming of the immune response by a process referred to as immunogenic cell death30. Other cytotoxic agents (for example, taxanes) block both tumour proliferation and affect innate immune cell function in the tumour microenvironment (for example, suppression of TReg cells and MDSCs)

28. Similarly, targeted kinase inhibitors that inhibit both the target in the tumour cell (for example, BRAFV600E) and cells of the immune system have a primary effect of blocking tumour proliferation and an off-target effect of immune stimulation31. Based on this new understanding of the immune regulatory properties of these agents, ipili-mumab has been studied in combination with paclitaxel and carboplatin as a treatment for non–small cell lung cancer32 and with vemurafenib (Zelboraf; Plexxikon/Genentech) in melanoma33 to determine whether these small-molecule drugs can positively influence the immune response and achieve clinical synergies. For further details regarding the opportunities afforded by the off-target effects of small-molecule drugs in immuno-oncology, the reader is referred to recent reviews29,34.

Although much attention in immuno-oncology drug development is given to antibodies for checkpoint mod-ulation, or cell and gene therapy and other T cell-based biologic approaches, in this Review we discuss a parallel track of investigation that has established mechanisms in the adaptive and innate immune system that can be regulated by small molecules to influence the course of events in the tumour microenvironment35−69 (FIGS 2,3; TABLE 2). As such, this Review provides an expansion of the topic covered by Muller and Scherle in 2006 (REF. 12). In many cases, the intracellular mechanisms targeted by these small-molecule drugs cannot be targeted by mAbs and regulate immunosuppressive cell types (for example, MDSCs, dendritic cells and TAMs) that are not directly regulated by checkpoint blockers. There are two particularly well-developed areas of investiga-tion: the regulation of MDSC, dendritic cell and TAM function by indoleamine 2,3-dioxygenase 1 (IDO1), arginase 1 (ARG1), inducible nitric oxide synthase (iNOS) or phosphodiesterase type 5 (PDE5), and the regulation of purinergic signalling in lymphocytes by ATP, CD39 and CD73, adenosine and elevated cyclic AMP. Also discussed are examples of signal transduc-tion inhibitors developed to specifically target dysregu-lated oncogenic signalling, but which are now known to have additional desirable immune regulatory effects in lymphocytes.

R E V I E W S




CD4 T cell

TH1

CTL

MDSC

TAM

TAM

MDSC

a Immunogenic cell death primes DCs b ARG1 and iNOS impair TCR signalling

NLRP3

ARG1

ARG1

NFκB

IFNγ

IL-1β,TNFα

IL-1β

Type IIFN

RR

R

R

RR

R

O O

O OO

TCR

iNOS

PDE5cGMP GMP

CD4 T cell

TCR

ROS

RNS

DC

DC

Tumour cell(dead)

Tumour cell(live)

TRegcell

TT

TK K

KK

TDO2

c IDO and TDO2 causes immune suppression d Chemokines recruit suppressive immune cells

TAM

CD8 T cell

TReg cellT

T

TT

KK

K KK K

KAHR

KAHR

KAHR

Expansion,activation

Suppression,anergy

↓ L-Tryptophanimpairs TCR activation

↑ Kynurenine↓ L-Tryptophan

T T K K KK

IDO1

IDO1

K KK

TT

IDO1

KK

K

5

55

5

5

5

5

5

5

5

5

5

5

5

5

1

11

1

1

1

1

1

12

12

12

12

12

ATP

L-Arginine

L-Ornithine

HMGB1

CCR2

CCR5

CXCR1

CXCR4

P2X7

CCL2

CCL5

CXCL1

CXCL12

TLR4

2

2

2

2

2

2

2

2

2

O

O

O OO

Figure 2 | Small-molecule drug targets to restore cancer immunity in the tumour. These processes are illustrated to be within the tumour microenvironment. a | Tumour cells undergoing intervention-induced immunogenic cell death (for example, after treatment with cyclophosphamide or oxaliplatin) release ATP and high mobility group protein B1 (HMGB1). ATP and HMGB1 bind, respectively, to the purinergic receptor P2X7 and Toll-like receptor 4 (TLR4) present on dendritic cells (DCs), to stimulate a pro-inflammatory response that primes the DCs for effective antigen presentation, resulting in the expansion of CD4+ T helper 1 (T

H1) and CD8+ cytotoxic T lymphocytes

(CTLs) and subsequent tumour killing. However, chronic inflammation resulting in chronic low-level activation of TLR and P2X receptors can lead to tumour promotion and immune suppression (see also FIG. 3a). b | The overexpression of arginase 1 (ARG1) and inducible nitric oxide synthase (iNOS) in myeloid-derived suppressor cells (MDSCs) and tumour-associated macrophages (TAMs) leads to the depletion of both intracellular and extracellular arginine, resulting in impaired T cell receptor (TCR) signalling and the release of reactive oxygen species (ROS) and reactive nitrogen species (RNS) by the same cells types, which further impair the immune response. The expression of ARG1 and iNOS in MDSCs is regulated by cyclic GMP levels, which are in turn controlled

by the activity of phosphodiesterase 5 (PDE5). Hence, agents that can elevate intracellular cGMP levels, such as PDE5 inhibitors, reduce MDSC-mediated immune suppression. c | Upregulation of indoleamine 2,3-dioxygenase 1 (IDO1) in immature DCs results in the depletion of extracellular tryptophan and the production of tryptophan metabolites, such as kynurenine, which signals through the aryl hydrocarbon receptor (AHR; an intracellular receptor) to afford immunosuppressive DCs with defective antigen presentation. The action of IDO1 in DCs and TAMs or IDO and tryptophan 2,3-dioxygenase (TDO2) in the tumour also activates regulatory T (T

Reg) cells and suppresses effector T cell function.

d | The choice between immune stimulation and suppression in the tumour microenvironment is determined by the cells that gain access to this compartment and subsequently by the context of immune signals that they encounter once there. Tumour secreted chemokines (CCL2, CCL5, CXCL1 are CXCL12 are shown) diffuse out of the tumour into the vasculature, producing a chemoattractant gradient that attracts and modulates the activity of immunosuppressive MDSCs, TAMs and T

Reg

cells (through binding to the cell surface receptors: CCR2, CCR5, CXCR1 and CXCR4). TABLE 2 lists the targets indicated in FIG. 2, providing examples of agonists (part a), inhibitors (parts b and c) and antagonists (part d).

R E V I E W S



a Hypoxia elevates ATP and adenosine in the TME b Adenosine suppresses immune function

c PGE2 suppresses immune function d Elevation of intracellular cAMP activates TReg cells

TReg cell

↑ cAMP

ACVII


ATP

P2X7RP2Y11

EP2

EP4

AMP

Adenosine

CD73

CD39

A2B receptor

COX2

COX2 COX2

PGE2

DC

Tumour cell (live)

Tumour cell (dead)

Arachidonic acid

ICD

ICD

↓ Tumour killing

↓ Tumour killing

↑ Tumour growth

↑ MDSC function↓ DC function

↑ Promote survival and proliferation

↑ IL-6, IL-8, IL-10, TGFβ and IDO

↑ Migration ↑ Activation

↑ Promote survival and proliferation

CD4 T cell

↑ FOXP3TAM

↑ IL-10 and immune suppression by myeloid cells

CTL

↓ IFNγ cytotoxicity↑ Anergy

A2Areceptor

NK cell

↓ IFNγcytotoxicity

↓ IFNγ TNFα

TGFβ

Hypoxia activates HIF1α and SP1 responsive genes:• ↑ CD39 • ↑ CD73 • ↑ COX2• ↑ A2A and A2B receptors

Tumourexosome

TReg cell

B cell

MDSC

COX2

COX2

Figure 3 | Purinergic and PGE2 signalling in immune cells. Immunogenic cell death of cancer cells leads to the release of elevated levels of extracellular ATP into the tumour microenvironment. Hypoxia elevates the expression of enzymes that convert ATP into adenosine (CD73 and CD39), arachidonic acid into prostaglandin E

2 (PGE

2)

(cyclooxygenase 2; COX2) and the receptors (A2A

and A2B

) that bind adenosine. a | Lymphocytes (regulatory T (T

Reg) cells and B cells), tumour

cells and exosomes provide a rich source of CD39 and CD73. In contrast to acute elevation of extracellular ATP (FIG. 2a), chronic elevation of extracellular ATP promotes growth in tumours overexpressing the purinergic receptor P2X7. Similarly, sustained extracellular ATP decouples the acute cytotoxic response mediated by P2X7 on myeloid-derived suppressor cells (MDSCs) and instead activates MDSC-mediated immune suppression (increased arginase 1 (ARG1), transforming growth factor-β (TGFβ) and reactive oxygen species (ROS)). Elevated extracellular ATP signalling through P2Y11 promotes the development of dysfunctional semi-mature dendritic cells (DCs). b | The actions of adenosine on the cells of the innate immune system are uniformly immunosuppressive. Examples of the impact of adenosine signalling on these cell types are indicated by coloured text boxes (beige text boxes for A

2A and violet text boxes for A

2B).

When expressed on tumours, adenosine receptors promote growth and proliferation. c | COX2 expression by tumour cells contributes directly to

tumour growth as well as indirectly through the production of PGE2,

which diffuses into the tumour microenvironment. Myeloid cells express COX2 and can serve as an important source of PGE

2. PGE

2 produced by

both the tumour and MDSCs binds to the PGE2 receptors EP

2 and EP

4 on

MDSCs, enhancing their immunosuppressive function by producing TGFβ, which inhibits tumour killing by natural killer (NK) cells. Activation of EP

2 and EP

4 receptors on myeloid cells inhibits the release of tumour

necrosis factor-α (TNFα) and interferon-γ (IFN), and on cytotoxic T lymphocytes (CTLs) inhibits their tumour killing function. d | The processes of adenosine and PGE

2 synthesis and signalling in the hypoxic tumour

microenvironment are illustrated in more detail for a TReg

cell. In the hypoxic tumour microenvironment, CD39+ T

Reg cells start expressing

CD73 and hence aquire the ability to convert ATP to adenosine. TReg

cells in the tumour microenvironment also commonly overexpress COX2 and produce PGE

2. The binding of adenosine to A

2A receptors and PGE

2 to EP

2

receptors present on TReg

cells, both of which are positively coupled to an adenylyl cyclase VII (ACVII), elevates cyclic AMP levels, leading to highly activated immunosuppressive cells. TABLE 2 lists the targets illustrated in FIG. 3, providing examples of antagonists and inhibitors to intervene in these pathways. HIF1α, hypoxia-inducible factor 1α; ICD, immunogenic cell death; IDO, indoleamine 2,3-dioxygenase; IL, interleukin; TAM, tissue-associated macrophage.

R E V I E W S



Amino acid catabolismAt its most basic level, the metabolism of amino acids is an evolutionarily conserved pathway that enables the conservation of nitrogen and the regulation of levels of physiologically required amino acids. Over time and in higher organisms, these processes have taken on addi-tional roles, such as the battle for nutrients between pathogenic organisms and mammalian hosts and the regulation of the innate immune response. Roles for the catabolism of cysteine, glutamine, phenylalanine, tryptophan and arginine in the regulation of immune responses to cancer and, in particular, T cell proliferation and activation, have been elucidated30,70.

IDO. The IDO family of haem-containing diox-genases converts tryptophan first to N-formylkynurenine and then further to kynurenine and additional metabo-lites71. The IDO family comprises three enzymes: tryptophan 2,3-dioxygenase (TDO2) and two related IDO isozymes (IDO1, IDO2), which are commonly grouped together as IDO (which in most cases refers to IDO1). TDO2 is expressed almost exclusively in the liver where its primary role is in maintaining a correct trypto-phan balance in response to diet. IDO is widely expressed, with the highest expression in antigen-presenting cells (macrophages and dendritic cells). IDO expression is also dysregulated in several tumour types and is correlated with a poor prognosis72. Both tumour- and antigen-presenting cell-derived IDO contribute to the immunosuppressed state of the tumour microenvironment. Multiple immuno-suppressive roles have been ascribed to the action of IDO, including the induction of TReg cell differentiation and hyper-activation, suppression of effector T cell immune responses and decreased dendritic cell function, all of which impair immune recognition and promote tumour growth71.

Several mechanisms have been proposed as the basis for the immunosuppressive functions of IDO and TDO2. The current consensus is that both the depletion of tryptophan and signals generated by its metabolites are important contributors to immunosuppresion73. With respect to the latter mechanism, the binding of kynurenine to the aryl hydrocarbon receptor has been identified as one of the pathways that enhance immune tolerance74. The importance of this pathway for the innate immune response was first demonstrated using 1-methyl tryptophan as an IDO inhibitor to study maternal T cell immunity. 1-Methyl tryptophan was also a key part of the evidence confirming the importance T cell-dependent tumour responses in immunogenic murine tumour models36.

The inhibition of IDO was one of, if not the first, small-molecule drug strategies proposed for the re-establishment of an immunogenic response to cancer71, and the D-enantiomer of 1-methyl trypto-phan (also known as D-1MT/indoximod), which was shown to be the more potent of the two enantiomers, was the first IDO inhibitor to enter clinical trials (see Supplementary information S1,S2 (tables)). Although indoximod clearly does inhibit the activity of IDO, it is a very weak inhibitor of the isolated enzyme and the in vivo mechanism (or mechanisms) of action for

this compound are still being elucidated. Investigators at Incyte were the first to advance an IDO inhibitor into the clinic; the IDO inhibitor was derived from a diversity screening and lead optimization process. The initial efforts provided an inhibitor of sufficient potency, selectivity and oral exposure to delay tumour growth in a mouse melanoma model75. Further development of this series led to the discovery of INCB24360 (REF. 35). When tested in co-cultures of dendritic cells and naive CD4+CD25− T cells, INCB24360 blocked the conver-sion of these T cells to CD4+FOXP3+ TReg cells, and when tested in a syngeneic model (PAN02 pancreatic cells) in immunocompetent mice, orally dosed INCB24360 pro-vided a significant dose-dependent inhibition of tumour growth. The ability of these selective IDO inhibitors to impede CD4+ T cell differentiation into TReg cells, and to preserve and promote dendritic cells capable of acti-vating effector T cells, provides additional confirmation of the importance of the IDO-mediated metabolism of tryptophan in the regulation of immune function. Based on these preclinical studies, INCB24360 entered clinical trials for the treatment of metastatic melanoma, first as a single agent and more recently in combination with checkpoint inhibitors (TABLE 2; see Supplementary infor-mation S1,S2 (tables)). In light of the importance of the catabolism of tryptophan in the maintenance of immune suppression, it is not surprising that overexpression of TDO2 by multiple solid tumours (for example, bladder cancer, liver carcinoma and melanomas) has also been detected38. Similar to the inhibition of IDO1, the selec-tive inhibition of TDO2 is effective in reversing immune resistance in tumours overexpressing TDO2 (REF. 38). These results support TDO2 inhibition and/or dual TDO2 and IDO1 inhibition as a viable therapeutic strategy to improve immune function.

Arginase. The catabolism of l-arginine by ARG1 and iNOS acts to suppress immunity. ARG1 is a cytosolic enzyme expressed in the liver and ARG2, a closely related isoform, is found in the mitochondria of a diverse set of cells. Additionally, ARG1 is highly expressed in, and secreted by, MDSCs and TAMs. Within the micro-environment, the depletion of extracellular arginine results in the depletion of the CD3ζ chain of the TCR, and the subsequent suppression of T cell responses to antigen. This deactivation of the TCR is mediated by the activation of the stress response kinase GCN2 (also known as EIF2AK4), which functions to control tran-scriptional and translational programmes that couple cell growth to amino acid availability, and is a pathway common to both arginine and tryptophan depletion76. Consistent with a prominent role of TAMs and MDSCs in assisting immune-mediated evasion of cancer, inhibi-tion of ARG1 blocked the growth of lung carcinoma in mice77. ARG1 levels in MDSCs are increased in patients with renal cell carcinoma and in patients with breast cancer78,79.

To date, N-hydroxy-nor-l-arginine and other argi-nine analogues have been used as tool compounds to study arginase inhibition, and cardiovascular indications have been the primary interest of those

R E V I E W S



http://www.nature.com/nrd/journal/vaop/ncurrent/full/nrd4596.html#supplementary-information

Table 2 | Immuno-oncology targets amendable to small-molecule medicines*

Target Location Function Compound (MOA)

Company or institution

Model or indication Status‡ Refs

Amino acid catabolism

IDO Macrophages, DCs, upregulated in tumours

Depletion of tryptophan and metabolites promote T

Reg cell differentiation,

suppression of immune response and decreased DC function

INCB24360 (inhibitor)

Incyte Murine syngeneic tumour (PAN02)

Phase II 35

1-Methyl tryptophan (inhibitor)

NewLink Genetics Murine syngeneic tumour model (Lewis lung cancer)

Phase I 36

NLG919 (inhibitor)

NewLink Genetics Murine syngeneic tumour (Pan02)

Phase I 37

TDO Hepatocytes Depletion of tryptophan and metabolites promote T

Reg cell differentiation,

suppression of immune response and decreased DC function

LM10 (inhibitor) Ludwig Institute for Cancer Research

Murine syngeneic tumour (P815B/TDO)

Research 38

ARG1, ARG2

MDSCs, TAMs, vascular endothelium

Depletion of the CD3ζ chain of the TCR suppresses T cell responses to antigen

Compound 9 (inhibitor)

The Institutes for Pharmaceutical Discovery

Reperfusion injury from myocardial ischaemia

Research 39

iNOS, ARG1, ARG2

MDSCs Supports generation of ROS that modify CCL2 levels, disabling T cell chemotaxis

NCX-4016 (dual inhibitor)

NicOx Preventing colorectal carcinoma

Phase II, discontinued

40

AT38 (dual inhibitor)

Istituti di Ricovero e Cura a Carattere Scientifico (IRCCS)

MCA-203 fibrosar-coma-bearing mice

Research 41

PDE5 MDSCs Decreases functional IL-13 receptors

Tadalafil (inhibitor)

Eli Lilly and Company

Investigational for immuno-oncology

Approved for erectile dysfunction and hypertension

42

Signalling of tumour-derived extracellular ATP

P2X7 Broadly expressed on lymphocytes, often upregulated in tumours

Induction of IL-1β release in DCs, enhances tumour-specific CD8 T cell cytotoxicity

ATP (agonist) Istituti di Ricovero e Cura a Carattere Scientifico (IRCCS)

Immuno-stimulant Research 43

Broadly expressed on lymphocytes, often upregulated in tumours

Increases CCL2, ROS, ARG1 and TGFβ levels; activates MDSCs, tumour growth and angiogenesis

AZ10606120 (antagonist)

University of Ferrara, Italy

Murine B16 F10 melanoma

Research 44

P2Y11

ATP derived from tumour binds receptor on DCs

Inhibits synthesis of IL-1, TNFα, IL-6; increases secretion of TSP1, IL-10 and IDO1, resulting in DC semi-maturation

NF340 (antagonist)

University of Duesseldorf, Germany

Immuno-stimulant Research 45

Adenosine signalling

A2A

receptor

TReg

cells, DCs, NK cells, NK T cells, tumours

Elevated cAMP blunts TCR-mediated cytotoxicity; inhibits effector T cells; expands T

Reg cells; enhances NK cell

cytotoxicity

SCH58261 (antagonist)

Peter MacCallum Cancer Centre, Victoria, Australia

B16 melanoma metastasis

Research 46

TReg

cells, DCs, NK cells, NK T cells, tumours

Elevated cAMP blunts TCR-mediated cytotoxicity; inhibits effector T cells; expands T

Reg cells; enhances NK cell

cytotoxicity

SCH420814 (antagonist)

Merck Parkinson disease Phase III, discontinued

47

R E V I E W S



A2B

receptor

Myeloid cells, expression driven by HIF1α

Elevated cAMP increases IL-10 and CCL2 levels; expansion of MDSCs and TAMs

PSB1115 (antagonist)

University of Salerno, Italy


Research 48

Adenosine production

CD39 TReg

cells, B cells, MDSCs, NK cells, tumours, endothelium

Contributes to the production of adenosine, which binds to A

1,A

2A, A

2B

and A3 receptors

ARL 67176 (inhibitor)

OREGA Biotech Murine B16 F10 melanoma

Research 49

CD73 TReg

cells, B cells, MDSCs, NK cells, tumours, endothelium

Contributes to the production of adenosine, which binds to A

1,A

2A, A

2B

and A3 receptors

AMPCP (inhibitor)

Cancer Therapy and Research Center, University of Texas San Antonio, USA


Research 50

Elevation of cyclic AMP

COX2 MDSCs, TAMs, T

Reg cells, tumours

Generates PGE2, which is

immunosuppressive (via EP

2 and EP

4 receptors)

Celecoxib (inhibitor)

Pfizer Rheumatoid arthritis, osteoarthritis, pain

Approved 51

EP2

receptorMDSCs, NK cells, T

Reg cells, tumours

TReg

cell activation; tumour proliferation and angiogenesis

PF-04418948 (antagonist)

Pfizer None indicated Phase I, discontinued

52

EP4

receptorMDSCs, NK cells, T

Reg cells, tumours

Activates suppressor cell function of MDSCs and TAMs

RQ-15986 (antagonist)

RaQualia Pharma Murine mammary 66.1 tumour metastasis

Preclinical 53

Chemokines and chemokine receptors

CXCR1, CXCR2

PMNCs, monocytes, endothelium, mast cells

Migration of CXCR2 expressing MDSCs into the TME; directs effects on tumour proliferation

CXCR2-specific mAb§ (antagonist)

Pediatric Oncology Branch, National Cancer Institute, National Institutes of Health, USA

Murine rhabdomyosarcoma

Research 54

CXCR4 T cells, B cells, monocytes, PMNCs, immature DCs, tumours

Ligand expression in stroma mediates metastasis by tumour-specific and T cell-based mechanisms

Plerixafor (also known as AMD3100) (antagonist)

Sanofi-Aventis, Cancer Research UK

Pancreactic ductal adenocarcinoma

Approved for stem cell mobilization

55

CCR2 Monocytes, PMNCs, immature DCs, T cells, NK cells

Drives TAM and monocytic MDSC infiltration into the TME

PF-4136309 (antagonist)

Pfizer, Washington University School of Medicine, National Cancer Institute, USA

Murine pancreatic model supportive of clinical study

Phase IB 56

CCR5 TH1 cells,

CD8+ T cells, monocytes, macrophages

TReg

cell infiltration and infiltration of precursors to generate TAMs and MDSCs

Maraviroc (antagonist)

National Center for Tumour Diseases, Germany

Blockade of metastatic colorectal cancer

Phase I 57

Recognition of foreign organisms to activate the immune response

TLR4 Monocytes, macrophages, DCs

Bacterial host defence; activation results in cytokine burst (IL-1, TNFα and type I IFNs)

OM-174 (agonist)

Centre Hospitalier Universitaire, France

Rat colon cancer, solid tumours

Phase I 58

TLR7, TLR8

DCs, plasmacytoid DCs, macrophages

Binds to viral ssRNA and bacterial DNA; induces secretion of inflammatory cytokines and type I IFN, which promotes a T

H1-directed activation

of DCs and NK cells to directly kill tumour cells and suppress T

Reg cells

Imiquimod (agonist)

Graceway Pharmaceuticals

Basal cell carcinoma Approved 59

Table 2 (cont.) | Immuno-oncology targets amendable to small-molecule medicines*




R E V I E W S



TLR7 DCs, plasmacytoid DCs, macrophages

Host defence recognizing viral ssRNA and bacterial DNA; inflammatory cytokines and type I IFN secretion promoting a T



Reg cells

852A (agonist) Pfizer Solid and haematological malignancies

Phase I/II 60





Reg cells

VTX-2337 (agonist)

VentiRx Pharmaceuticals

Solid and haematological malignancies

Phase I/II 61





Reg cells

IMO-2055 (agonist)

Hybridon, Idera Pharmaceuticals

Advanced solid malignancies

Phase I/II 62

Signal transduction: kinase inhibitors

ALK5 Downstream of TGFβ, which is often overexpressed by tumours

Attenuation of TGFβ signalling causes activation of CD8+ cells, generation of CTLs, and stimulation of NK cells

LY2157299 Eli Lilly and Company


Phase I/II 63

EW-7197 Ewha Womens University, Seoul, Korea


Phase I 63

BRAFV600E Tumours V600E-driven IL-1 expression promotes immunosuppressive TAF and MDSC function

Vemurafenib Plexxikon, Genentech, GlaxoSmithKline, MD Anderson Cancer Center, USA

Patients with melanoma

Approved for metastatic melanoma

64

Dabrafenib 65

RON Expressed on myeloid cells. Tumours secrete its ligand MSP

Decreases IL-12, IFNγ and TNF, and increases IL-10; favours M2 phenotype

BMS-777607 Bristol-Myers Squibb, Huntsman Cancer Institute, Utah, USA

Inhibits metathesis in MMTV-PyMT transgenic mice

Phase I/II 66

CSF1 Glioma cells and TAMs express CSF ligand

M1 to M2 polarization, which promotes tumour growth and survival

BLZ945 Memorial Sloan-Kettering Cancer Center, New York, USA

Murine glioblastoma Research 67

PI3Kδ B cells, T cells, myeloid lineage cells

Inhibition preferentially suppresses T

Reg cell

function, resulting in effector T cell activation

PI-3065 Piramed Pharma, University College London Cancer Institute, UK

4T1 breast cancer and other solid tumours

Research 68

PI3Kγ Haematopoietic cells, primarily myeloid lineage

Required for α4β1-dependent myeloid cell infiltration into tumours

TG100-115 University of California San Diego, Moores Cancer Center, USA

Lewis lung carcinoma and PyMT spontaneous breast carcinomas

Research 69

AMPCP, adenosine 5ʹ-(α,β methylene)diphosphate; ARG, arginase; COX2, cyclooxygenase 2; CSF, colony stimulating factor; CTL, cytotoxic T lymphocyte; DC, dendritic cell; HIF1α, hypoxia-inducible factor 1α; IDO, indoleamine 2,3-dioxygenase; IFN, interferon; IL, interleukin; iNOS, inducible nitric oxide synthase; MDSC, myeloid-derived suppressor cell; MOA, mechanism of action; MSP, macrophage-stimulating protein; NK, natural killer; PDE5, phosphodiesterase type 5; PGE

2, prostaglandin E

2; PMNC, peripheral mononuclear cell; ROS, reactive oxygen species; TAF, tumour-associated fibroblasts; TAM, tumour-associated

macrophage; TCR, T cell receptor; TDO, tryptophan 2,3-dioxygenase; TH, T helper; TGFβ, transforming growth factor-β; TLR, Toll-like receptor; TME, tumour

microenvironment; TNF, tumour necrosis factor; TReg

, regulatory T; TSP1, thrombospondin 1.*Listed are small-molecule drug targets that have been proposed for cancer immunotherapy. ‡For some examples, the clinical development status provided is for a non-immuno-oncology indication. In these cases the cited literature supports clinical consideration in light of its impact on innate immune function. §While the cited reference illustrates CXCR2 antagonism using a mAb, several small-molecule CXCR1 and CXCR2 antagonists have reached clinical trials and in principle could show similar efficacy.

Table 2 (cont.) | Immuno-oncology targets amendable to small-molecule medicines*




R E V I E W S



Hypoxia-adenosinergic axisA programme of gene regulation in response to hypoxia whereby elevated extracellular adenosine mediates a broadly suppressive immune response.

seeking to develop compounds that could potentially be progressed into the clinic. The most effective com-pounds have half-maximal inhibitory concentration values of ~0.5 μM against both ARG1 and ARG2 and moderate bioavailability39.

Also present at high levels in MDSCs is iNOS, which converts arginine to nitric oxide. Nitric oxide primarily suppresses T cell function either by S-nitrosylation of critical cysteine residues of proteins or through its regu-lation of guanylyl cyclase and cyclic GMP-dependent kinases80. The expression of ARG1 and iNOS is usually mutually exclusive. This is not the case in MDSCs and TAMs, thus setting up a dynamic competition between the two enzymes for the same substrate, l-arginine. When arginine is depleted by the action of ARG1, iNOS switches to superoxide production. Subsequent reactions generate reactive oxygen species and reactive nitrogen species, which are additional mediators of the immuno-suppressive activity of iNOS. Given the central role of arginine metabolism, it was anticipated and subse-quently demonstrated that the inhibition of both ARG1 and iNOS would be optimal for restoring immune func-tion. NCX-4016, a nitroaspirin that inhibits cyclooxy-genase and releases nitric oxide, was the first reported small-molecule drug to inhibit both ARG1 and iNOS81. The more potent nitroaspirin analogue AT38 was identi-fied in a screening protocol using a co-culture of T cells with MDSCs (TABLE 2)41. In vivo, AT38 produced a large influx of T lymphocytes into mouse tumours, and com-bining AT38 with an adoptive cell transfer of cells recog-nizing an engineered antigen (EG7-OVA) produced a high number of durable responses41.

PDE5 inhibitors, such as tadalafil, inhibit the degra-dation of cGMP in MDSCs, resulting in decreased immunosuppressive activity42. Although the underlying mechanism of this effect is not fully elucidated, PDE5 inhi-bition decreases the amount of functional interleukin-13 (IL-13) receptors by decreasing IL-4 receptor-α (subunit of IL-13R) on MDSCs82. Reduction of IL-13 signalling in turn reduces the levels of phosphorylated signal trans-ducer and activator of transcription 6 (STAT6), leading to decreased expression of ARG1 and iNOS. The impact of tadalafil on human immune function has been char-acterized in two clinical trials of tadalafil in head and neck squamous cell carcinoma (HNSCC), a tumour characterized by an evasion of immune surveillance and suppression of systemic and specific immunity (see Supplementary information S2 (table)). In both studies, tadalafil improved general and tumour-specific immu-nity, with a lowering of MDSC and TReg cell numbers, and a reduction in their suppressor function83,84.

Purinergic signalling in immune cellsLocal levels of extracellular ATP are acutely elevated as a consequence of infection, tissue injury, ischaemia or intervention-induced tumour cell death. Elevated extracellular ATP is recognized by the immune system as a danger signal to initiate multiple pro-inflammatory events, including the recruitment of macrophages and dendritic cells and the activation of the NLRP3 (NOD-, LRR- and pyrin domain-containing 3) inflammasome

in dendritic cells, leading to the release of IL-1β43. Successive processing of extracellular ATP by the ecto-nucleotidases CD39 and CD73 lowers extracellular ATP levels and can rapidly elevate extracellular adeno-sine85 from a low homeostatic level (20–200 nM) to as much as 1,000–10,000 nM47. These elevated adenosine concentrations engage the immunosuppressive actions of adenosine A2A and A2B receptors on the infiltrating lymphocytes, shielding cells from an excessive inflam-matory response and thereby providing a self-limiting mechanism to resolve the immune response. Within the context of a solid tumour, hypoxia has been shown to increase adenosine levels by 10–20-fold compared with normal levels86. It has been proposed that adenosine elevation is sufficient to maintain a chronic suppression of the innate immune response, resulting in immune tolerance and, subsequently, uncontrolled malignant growth85.

Extracellular adenosine signalling can be terminated either by facilitated uptake by the widely expressed equilibrative nucleoside transporter 1 (ENT1) and ENT2 or by adenosine deaminase, which deaminates adeno-sine to produce inosine47. Identification of a hypoxia-inducible factor 1α (HIF1α)-responsive element in the promoter of the gene encoding CD73 and a transcription factor SP1-dependent regulatory pathway for CD39 pro-vides a direct mechanistic link between hypoxia and ele-vated adenosine levels. A HIF1α-responsive element in the promoter of the A2B receptor gene (ADORA2B) pro-vides further coordination of a transcriptional response to hypoxia87. A unified view detailing how the hypoxia–adenosinergic axis enforces TReg cell-mediated immune tolerance has been presented88, whereby hypoxia drives the expression of enzymes that elevate the levels of aden-osine and the A2A and A2B receptors, leading to increased cAMP levels in lymphocytes. The expression of both CD73 and CD39 on TReg cells enables the conversion of pro-inflammatory extracellular ATP to immunosuppres-sive adenosine, which can then bind to A2A receptors on effector T cells to silence the antitumour immune response89. TReg cells also express A2A receptors, which creates an autocrine loop to enhance other immunosup-pressive mechanisms of TReg cells, such as the expression of CTLA4 and other checkpoint exhaustion markers90. This axis provides a powerful mechanism to enhance the proportion of TReg cells relative to effector T cells in the tumour microenvironment, which further drives the production of additional adenosine and immuno-suppressive cytokines. In this regard, adenosine acts in much the same way as TCR co-receptor engagement on TReg cells (for example, via the checkpoints CTLA4 and PD1), thus providing a potent stop signal that is then transmitted throughout the tumour microenvironment.

ATP and ADP receptors: the P2 family. There are seven subtypes of P2X ion channel receptors (P2X1-7) and eight subtypes of G protein-coupled P2Y receptors, many of which are found on immune cells and on the endothe-lium of the tumour microenvironment. Extracellular ATP is the sole agonist for the P2X receptors, whereas P2Y receptors bind both ADP and ATP, or UTP and UDP

R E V I E W S



B cellsLymphocytes that produce antibodies in response to antigen presentation by antigen-presenting cells. B cells may also function as antigen-presenting cells in some instances.

(P2Y4 and P2Y6). High concentrations of extracellular ATP activate the P2 family of receptors on dendritic cells to initiate an innate immune response through the induc-tion of IL-1β release, thereby enhancing specific CD8+ T cell cytotoxicity43. Extracellular ATP also promotes antitumour immunity by skewing TH cell commitment towards TH1 and type 17 (TH17)

91,92, and limiting immu-nosuppressive activity by inducing TReg cell apoptosis

93,94. Many of these effects appear to be mediated by the acti-vation of P2X7, which suggests that a P2X7 agonist or a compound inhibiting the degradation of ATP would boost the antitumour immune response. Reinforcing this targeting approach, high extracellular ATP has been reported to induce apoptosis in tumour cells95. However, these observations appear paradoxical to the observed overexpression of P2X7 in human tumours, whereby its expression is better correlated with promoting growth and survival and not tumour apoptosis96. These para-doxical effects may be the result of a chronic elevation of extracellular ATP levels, which leads to an uncoupling from the normally observed acute apoptotic response. Instead, tumour-derived extracellular ATP acts through P2X7, thereby driving tumour growth and enhancing MDSC immune suppression97.

Recently, the requirement of P2Y11 for thrombospon-din 1 and IL-8 secretion was confirmed in human den-dritic cells using a P2Y11-selective non-nucleoside agonist (NF546) and antagonist (NF340)45, which provides sup-port for advancing the development of a P2Y11-selective antagonist to prevent the production of dysfunctional semi-mature dendritic cells. The intrinsic complexity of P2 receptor subtype expression and function has discouraged the investigation of P2X and P2Y receptor modulation as an area of research for immuno-oncology. However, there has been interest in P2X antagonists for anti-inflammatory and autoimmune indications98, and P2Y12 antagonists for blocking ADP-induced platelet aggregation have advanced to the clinic. The continued interest in this class of recep-tors has led to the identification of selective agonists and antagonists for several P2 receptor subtypes. These com-pounds, and those for additional P2 family members, should provide the tools required to determine the poten-tial utility of these receptors for modulating the innate immune response to cancer.

Adenosine A2A receptor. A2A receptors are widely expressed on leukocytes, including lymphocytes, dendritic cells, natural killer cells and T cells, and binding of adenosine suppresses the effector functions of these cells85. Of par-ticular interest to immune suppression are the regulatory effects of A2A receptor activation on T cells, in which adenosine significantly blunts TCR-mediated cytotoxic-ity and cytokine production, inhibits T cell proliferation89, and induces the expansion of TReg cells that have stronger immunosuppressive activity due to an increased expres-sion of CD39, CD73 and CTLA4 (REF. 90). In models of metastatic disease (B16F10 melanoma and 4T1.2 breast cancer), a selective A2A receptor antagonist (SCH58261) was shown to decrease metastatic spread46. Furthermore, the combination of an A2A antagonist (SCH58261 or ZM241365) with a T cell checkpoint blocker (an anti-PD1

or anti-CTLA4 agent) was more effective than the individual agents in reducing metastasis in the B16F10 melanoma model (see Supplementary information S1 (table)). The additional efficacy of an adenosine antago-nist with T cell checkpoint blockade supports a non-redundant role for adenosine in the immune response in cancer. Two A2A receptor antagonists, istradefylline (also known as KW-6002) and preladenant (also known as SCH420814), have been studied in Phase II and Phase III trials of Parkinson disease. Istradefylline demonstrated modest efficacy, which was sufficient for it to gain regis-tration approval in Japan (marketed as Nouriast by Kyowa Hakko Kirin)47. The preclinical data (see Supplementary information S1 (table)) for A2A receptor antagonists in combination with checkpoint antagonists provides a rationale for testing this combination in the clinic.

Adenosine A2B receptor. A2B receptors are expressed at low levels across multiple cell types, but their expression is sig-nificantly upregulated under hypoxic conditions. The A2B receptor is the least sensitive of the four adenosine recep-tors, requiring micromolar adenosine concentrations that are only achieved under conditions of physiologi-cal stress. On classically activated macrophages, which release pro-inflammatory cytokines and are central ele-ments of the cellular immune response, activation of the A2B receptor has a prominent role in M1 to M2 switching, thereby suppressing the antitumour T cell response and stimulating tumour angiogenesis99. Elevated adenosine concentrations derived from the activity of CD73 also activates A2B receptors on myeloid precursor cells to drive the expansion of MDSCs and thereby suppress effector T cell activity100. The first in vivo demonstration of anti-tumour activity of a selective A2B receptor antagonist was achieved using intratumoural injections of ATL801 to reduce the growth of bladder cancer and of breast cancer in mice101. Intratumoural injection of a different A2B receptor-selective antagonist (PBS1115) provided in vivo support for an A2B receptor-dependent role in the accumulation of tumour-infiltrating MDSCs47. These results support the prediction that high adenosine levels within the tumour microenvironment are sufficient to trigger engagement of low-affinity A2B receptors, which drive MDSC-mediated suppression of tumour immunity and lead to the differentiation of naive T cells into a TReg cell suppressive lineage. Several groups are pursuing the clinical development of A2B antagonists for anti-inflam-matory indications (inflammatory bowel disease and chronic obstructive pulmonary disease), and one selec-tive antagonist, CVT-6883, demonstrated tolerability in Phase I safety trials102.

CD39. CD39 is expressed on the surface of endothelial cells and subsets of leukocytes, particularly on B cells. CD39 converts extracellular ATP first to ADP and then, without release, ADP is converted in a second step to AMP. With respect to tumour immune tolerance, there are mul-tiple lines of evidence linking elevated CD39 expression to immune suppression and increased tumour susceptibility. First, the immunosuppressive potential of both mouse and human TReg cells requires CD39 expression

89,103. Second,

R E V I E W S



the level of CD39 expression on TReg cells in patients with HNSCC104, chronic lymphocytic leukemia105 or follicular lymphoma106 is elevated above that of normal healthy subjects, and expression levels are positively correlated with disease severity. Third, in a model of hepatic meta-static cancer, metastatic growth was strongly inhibited in mice with CD39-null vasculature and in wild-type mice with circulating CD39-null bone marrow–derived cells91. Fourth, in a MCA205 mouse tumour model, overexpres-sion of CD39 abolished the ability of chemotherapy to reduce the rate of tumour growth, a result attributed to the reduced immunogenicity of chemotherapy-mediated tumour cell death107. With respect to the first three points, the requirement for CD39 was dependent on the produc-tion of elevated adenosine levels, whereas in the fourth point, the prevention of immunogenic cell death was dependent on the ability of CD39 to remove extracellular ATP from the tumour microenvironment. For several of the examples cited above, small-molecule inhibitors of CD39 have been used as proof that CD39 is involved in the production of adenosine. However, these tool com-pounds have many shortcomings108. ARL 67176 and 8-butylthio-ATP are nucleotide analogues with marginal selectivity and potency, and although they have been used for in vivo experiments, the pharmacokinetic behaviour of these compounds is unknown109,110. Polyoxotungstate (POM-1), which is a high-molecular-mass metal complex, raises similar concerns.

CD73. CD73 is a glycophosphatidylinositol-anchored di-Zn2+ metallo-phosphatase specific for the dephos-phorylation of purine and pyrimidine ribo- and deoxy-ribonucleoside monophosphates to the corresponding nucleoside. AMP is the preferred substrate of CD73. Although other extracellular phosphatases, such as tissue-nonspecific alkaline phosphatase, can convert AMP to adenosine, the catalytic properties of these enzymes make them unlikely contributors for this con-version in the tumour microenvironment. Consistent with the non-redundant role of CD73 in the conversion of ATP to adenosine, CD73 expression is directly regu-lated by HIF1α111. CD73 expression is also upregulated by type I interferons (IFNs), protein kinase C activation, IL-1β, tumour necrosis factor-α and prostaglandin E2 (PGE2). IFNγ and IL-4 are negative regulators of CD73 expression112. CD73 is overexpressed in multiple solid tumour types and leukaemias, including aggressive and difficult to treat tumours (glioblastoma and ovarian tumours)112. In patients with HNSCC, TReg cells (both cir-culating and tumour associated) express both CD39 and CD73, thus providing a mechanism for the conversion of ATP to adenosine that only depends on TReg cells

113. However, as both CD73 and CD39 are membrane-anchored enzymes processing extracellular nucleotides, it may not be required that both enzymes are expressed on the same cell. Tumour cells overexpress CD39 and CD73 and secrete these enzymes in exosomes114, and B cells, which are abundant in the tumour microenvi-ronment, also provide a rich source of these enzymes115. Similar to CD39, there are multiple reports using siRNA, transgenic knockouts and overexpression models to

confirm the involvement of CD73 in the generation of adenosine and promotion of immune tolerance112. Adenosine 5ʹ-(α,β-methylene)diphosphate (AMPCP), in contrast to all other small-molecule CD73 inhibitors that have been described in the literature so far, is both potent and selective and has been widely used in vitro to eluci-date the role of CD73 catalytic activity in suppressing immune responses to cancer113,116,117. AMPCP has also been used in animal models by several research groups to demonstrate reduced tumour growth rates, cytokine production and metastasis in B16 melanoma116,118. Given the availability of active recombinant soluble CD73 and of inhibitor-bound X-ray crystal structures119, further progress in this area may soon be forthcoming.

COX2 and PGE2 receptorsIn contrast to the observation by Coley and colleagues2 that acute activation of the immune system can initiate an immune response to cancer, there is abundant litera-ture linking chronic inflammation to tumour promotion and progression. In colorectal cancer, the expression of the inducible form of cyclooxygenase (COX2) is asso-ciated with lower survival rates120. COX2 is also over-expressed in colonic epithelial cells of patients with inflammatory bowel disease. Treating inflammatory bowel disease and ulcerative colitis with either selective or non-selective cyclooxygenase inhibitors was associ-ated with a decrease in disease progression51. Moreover, long-term preventive treatment of patients with familial adenomatous polyposis with either non-selective COX1 and COX2 inhibitors (non-steroidal anti-inflammatory drugs) or COX2-selective inhibitors led to the regression of adenomas51. In addition to stimulating tumour prolif-eration, survival and migration, the metabolites of COX2, primarily PGE2, have direct effects on immune-mediated tumour escape. PGE2 is the primary mediator of the clas-sical signs of inflammation: oedema, redness, swelling and pain. Acutely, PGE2 favours a pro-inflammatory (TH1-type) immune response, but sustained levels, such as those induced by the upregulation of COX2 in the hypoxic tumour microenvironment121, enhance the activity of multiple immunosuppressor cells, including TAMs122, TReg cells

123 and MDSCs124.PGE2 signals through a set of four G protein-coupled

receptors, referred to as EP receptors (EP1, EP2, EP3 and EP4), which are coupled to different signal transduction pathways. In lymphocytes, PGE2 activation of the EP2 and EP4 receptors is positively coupled to adenylyl cyclase, leading to elevated cAMP levels and the subsequent activation of protein kinase A125. In human TReg cells, it was shown that either a non-selective COX inhibitor (indomethacin) or an EP2 receptor antagonist (AH6809) decreased TReg cell-mediated suppression, but a selective EP4 receptor antagonist was without effect

113. As dem-onstrated by the use of selective EP receptor antagonists, the production of immunosuppressive TGFβ by MDSCs requires the EP2 receptor or a combination of EP2 and EP4 receptors

53,126. In murine splenic natural killer cells, which express all four PGE2 receptors, a selective EP4 ago-nist inhibited IFNγ production, whereas selective EP1-3 agonists did not127. All of these experiments implicate

R E V I E W S



that elevated cAMP levels are the primary signal leading to immunosuppression, irrespective of which EP recep-tor is predominant (EP2 or EP4). Selective PGE2 receptor antagonists have recently been studied clinically for vari-ous non-oncology indications; among these agents are selective EP2 (PF-04418948)

52 and EP4 (BGC20-1531)128

antagonists. To date, the studies on direct effects of EP antagonists on tumour proliferation and angiogenesis have been restricted to preclinical investigations, but with the recent successes in immuno-oncology and a growing body of literature highlighting the role of EP signalling in immune tolerance, these studies may soon extend into clinical investigations53.

cAMP is the common second messenger for adeno-sine receptors (A2A and A2B) and PGE2 receptors (EP2 and EP4) in lymphocytes. As discussed previously, in TReg cells, both adenosine and PGE2 elevate cAMP levels, leading to enhanced suppressor activity. Their effect on suppressor cell activity can be blocked by selective antagonists (A2A and EP2) and when used together, the effect of these antag-onists is additive113. Of the nine transmembrane-bound forms of adenylyl cyclase, isoform VII is highly expressed in lymphocytes and macrophages; in these cell types, this adenylyl cyclase isoform exclusively couples to agonists that elevate cAMP levels129. The convergence of the adeno-sine and PGE2 pathways on an isoform of adenylyl cyclase that is preferentially associated with the haematopoietic linage provides an opportunity to block, with a single small-molecule drug, the input from multiple immuno-suppressive agonists across multiple immunosuppressive cell types130. Although selective adenylyl cyclase VII inhi-bition is highly attractive in principle, there are currently no known isotype-selective inhibitors for any of the adenylyl cyclase isoforms. In the near term, combinations of COX2 inhibitors or selective PGE2 receptor antagonists with modulators of adenosine signalling could provide a useful approach to regulate the immunosuppressive effects of cAMP on immune function.

Toll-like receptorsToll-like receptors (TLRs) are expressed on antigen-presenting cells such as monocytes and macrophages, B cells, neutrophils and dendritic cells, as well as on tissues exposed to the external environment, such as the gastrointestinal tract and lungs. All TLRs are type 1 transmembrane proteins and are localized in the cyto-plasm, with the exception of TLR3, TLR7, TLR8 and TLR9, which are localized in the endosomal compart-ment. TLR1 and TLR2 are expressed by macrophages and neutrophils, and recognize molecular patterns on Gram-positive bacteria and fungi. Both TLR2 homo-dimers and heterodimers of TLR1–TLR2 have been proposed to exist, but only the heterodimers have been repeatedly confirmed by observation and experimental practice. TLR3 recognizes double-stranded RNA and polynosinic:polycytidylic acid (polyI:C) moieties from viruses, whereas TLR4 recognizes lipopolysaccaride from bacterial cell walls. TLR5 binds to bacterial flagellin. Both TLR6 and TLR7 recognize single-stranded RNA and small synthetic molecules such as imidazoquinolines and nucleoside analogues. TLR9 binds to unmethylated

CpG in bacterial and viral DNA. When engaged, these TLRs trigger a strong pro-inflammatory cytokine response that stimulates immune surveillance.

Until recently, clinical trials of TLR agonists for cancer indications have utilized the agonists as either vaccine adjuvants or as monotherapy. The majority of trials have examined agonists of endosomal TLRs (TLR3, TLR7, TLR8 and TLR9). Small-molecule heterocyclics, such as the imidazoquinolines, are recognized as nucleoside agonists by TLR7 and TLR8, which distinguishes these from TLR3 and TLR9 for which only oligonucleotide agonists have been discovered131. The antitumour effects of TLR7 and TLR8 agonists are primarily mediated by the activation of dendritic cells and natural killer cells to directly kill tumour cells, and by the suppression of TReg cells132. Imiquimod (Aldara; Graceway Pharmaceuticals), a TLR7 and TLR8 agonist, has been approved as a topical monotherapy for the treatment of basal cell carci-noma59. Small-molecule agonists of TLR7 (852A)60 and TLR8 (VTX-2337)61 suitable for systemic administra-tion are being evaluated as single agents in both solid and haematological malignancies. TLR9 agonists, such as IMO-2055, CPG 7909 and MGN1703, induce type l IFN secretion in dendritic cells, promoting a TH1-type response of cytotoxic dendritic cells, natural killer cells and T cells as well as promoting M2 to M1 switch-ing133; all these effects contribute to tumour-specific immune responses and the reversal of immune suppres-sion. Although there are several trials underway (see Supplementary information S2 (table)), to date, there has been no clear demonstration of clinical benefit when using these agonists as single agents.

Although not a pure TLR agonist, the tuberculosis vaccine Bacillus Calmette-Guerin (BCG) has been used as an effective treatment for bladder cancer, acting pri-marily through the stimulation of TLR2 and TLR4. In the context of chronic inflammation TLRs are associated with elevated pro-inflammatory cytokine production, leading to tumour promotion132. However, these receptors are also present on CD4+ and CD8+ T cells and can promote immune stimulation. TLR4 was discovered as the primary target mediating the potent inflammatory response to bacterial endotoxin (lipopolysaccharide) and is the most studied member of the TLR family. TLR4 is also essential in providing a ‘danger signal’ that breaks tumour antigen tolerance134. Lipopolysaccharide binding to TLR4 acti-vates both a pro-inflammatory response mediated by myeloid differentiation primary response 88 (MYD88) signalling and a TRIF-dependent pathway, leading to type I IFN production. Semi-synthetic analogues of lipid A, which is the membrane anchor of lipopolysaccharide that engages TLR4, have been pursued to favour the TRIF over the MYD88 pathway and thereby reduce toxicity and also improve its pharmaceutical properties135. OM-174, in which three of the six lipid tails of lipid A have been removed, induced the complete regression of tumours and haemorrhagic ascites in 90% of the tumour-bearing rats in a colon cancer metastasis model58. Other ana-logues such as monophosphoryl lipid A and aminoalkyl glucosamine phosphates have been used as vaccine adju-vants, including a therapeutic cancer vaccine based on the

R E V I E W S



tumour-specific protein melanoma-associated antigen 3 (MAGEA3)136. Going forward, important considerations related to pro- and antitumour roles attributed to TLR signalling are the strength of the immune stimulation and the duration of the signal. Chronic infection and inflam-mation resulting in continuous low-level stimulation of TLRs promotes cytokine signals that lead to TReg cell stim-ulation and impaired effector T cells. Conversely, there is now compelling preclinical data to suggest that robust acute activation of TLRs can restore antigen-presenting cell function, leading to amplification and activation of tumour-specific effector T cells137.

ChemokinesThe choice between immune stimulation and suppres-sion in the tumour microenvironment is determined by the cells that gain access to this compartment and sub-sequently by the context of immune signals that they encounter there. Chemokines and their receptors have critical roles in this immune signalling cascade and in immune responses to diseases, including cancer.

Interruption of chemokine signalling has been pur-sued with mAbs; however, chemokine receptors belong to the class of G protein-coupled receptors that is also druggable using a traditional small-molecule approach. The biology for this class of ligands and receptors was first described as “chemo-attractive for leukocytes” (REF. 138) but has since been established as “essential for coordinating homing, retention and activation of T cell and other immune infiltrates to tumours” (REF. 138). Moreover, chemokines and their receptors signal to ensure effective biodistribution and functioning of immune infiltrate in tumours139.

Recent reviews of the entire chemokine receptor family and its ligands lists more than 50 members classified in four distinct groups that include the CXC, CC, XC and CX3C subfamilies, which are distinguished from one another based on structure, position and placement of key cysteine residues, with ‘L’ or ‘R’ in the nomenclature denoting ligand or receptor respectively140. The entirety of the chemokine receptor and ligand family has been implicated in various aspects of immune biology that may have important effects in cancer139,141,142. A number of these drug targets have been considered for cancer intervention based largely on their role in trafficking, retention and activation of immune infiltrate and, in some cases, preclinical and even clinical cancer investi-gations have been initiated139 (see Supplementary infor-mation S2 (table)). However, although the biology is compelling, a cancer-relevant clinical demonstration of enhanced tumour immunity for a chemokine antagonist remains to be seen. Given the wealth of recent reviews on this subject, this manuscript will only exemplify a few recent examples from this class of small-molecule inhibitors to illustrate the value of such opportunities

CXCR family. CXCR4 is expressed in a wide range of solid and haematopoietic malignancies, and expression of its ligand CXCL12 in stroma is thought to help mediate metastasis via both tumour-specific and T cell-based mechanisms143. Recruitment of TReg cells and increased

levels of CXCR4 expression were correlated with poor prognosis in basal-like breast cancers144. In a model of pancreatic ductal carcinoma, a CXCR4 antagonist improved response to anti-PD1 treatment, which was correlated with increases in the numbers of both effector T and TReg cells

55 (see Supplementary information S1 (table)). Plerixafor (Mozobil; AnorMED/Genzyme), a CXCR4 antagonist, is an approved immunostimulant used to mobilize haematopoietic stem cells in patients with cancer, and pharmacodynamic studies in non-Hodgkin lymphoma evince that neutrophil, lympho-cyte and monocyte populations all increase following its systemic administration145. Other CXCR4 inhibi-tors currently being evaluated in cancer clinical trials include TG-0054 (also known as burixafor), MSX-122, CTCE-9908, POL6326 (REF. 143) and BKT140 (REF. 146).

The CXCR2 (IL-8β receptor)-dependent accumula-tion of MDSCs in the tumour microenvironment has recently been demonstrated in both colitis-associated147 and rhabdomyosarcoma tumour models54.

CCR family. Macrophages can produce pro-tumorigenic factors, including growth-promoting and angiogenic factors that can accelerate cancer growth if recruited to the tumour bed. CCL2 was the first chemokine ligand to be validated as a chemoattractant for macrophage migration into the tumour bed, and blockade of this axis has therefore been pursued as a means to limit macrophage infiltration and pro-tumorigenic activities148. A plethora of experimental medicines have entered clinical trials that inhibit the CCL2–CCR2 pathway, including AZD-2423, BMS-741672, BMS-813160, CCX-140, cenicriviroc, JNJ-17166864, MK-0812, MLN-1202, PF-04634817 and PF-4136309. PF4136309 has been evaluated in a human pancreatic cancer model, which indicated that CCR2 blockade depletes inflammatory monocytes and mac-rophages from the primary tumor and premetastatic liver, resulting in enhanced antitumor immunity, decreased tumor growth, and reduced metastases56 (TABLE 2).

CCR5 on TH1 cells, CD8+ T cells, monocytes and

macrophages supports the infiltration of TReg cells as well as progenitor cells to generate TAMs and MDSCs in the tumour microenvironment 57. Maraviroc (Selzentry/Celsentri; Pfizer) is a small-molecule CCR5 antagonist that has been studied extensively in clinical trials and was approved in 2007 for treating HIV. Maraviroc has been partially evaluated in a colorectal cancer Phase I study that completed in October 2014 (ClinicalTrials.gov identifier: NCT01736813).

Signal transductionIntracellular signal transduction pathways have proved to be successful points of intervention for the development of tumour-directed small-molecule therapeutics. This sec-tion primarily highlights protein kinases, which have been demonstrated to be the most amendable intracellular tar-get class for small-molecule drug intervention. The obser-vations described below provide hope that as the pathways of signal transduction involved in immune tolerance are more fully understood, additional intracellular targets for small-molecule drug intervention will be discovered.

R E V I E W S



http://www.clinicaltrials.gov

Cytokine signalling. The Janus kinase (JAK)–STAT pathway, which mediates the signalling of multiple cytokine receptors (for example, IL-2, IL-6 and IFNγ), has been an area of intense interest for the development of drugs to treat autoimmune disorders and cancer. TGFβ, which has been characterized as one of the most potent immunosuppressive cytokines in the tumour microenvironment, binds to TGFβ receptor type 2 and signals via ALK5 (also known as TGFβ receptor type 1)- mediated phosphorylation of SMAD2 and SAMD3. Both the JAK–STAT and ALK5–SMAD2/3 pathways were discussed in the 2006 review on regulators of immune tolerance by Muller and Scherle12. Despite the long-standing interest in these pathways, progress in the development of small-molecule drugs targeting these pathways has been slow. One problem has been specificity, as the JAK–STAT pathway is widely used in cytokine signalling and TGFβ can both positively and negatively regulate the immune response. A second problem has been achieving selective kinase inhibition; there are three JAKs and six ALKs, and inhibitors often block other non-family member kinases. Early on in the development of ALK5 inhibitors, there were multiple reports of myocardial toxicity, which appeared to be due to ALK5 inhibition149. Recently, an ALK5 inhibitor (LY-2157299) was reported to have an improved pre-clinical safety profile in a number of solid tumours, and a Phase I trial is planned to monitor effects on immune status150. In a murine melanoma model, the activa-tion of cytotoxic T lymphocytes by EW-7197, a potent ALK5 inhibitor, led to inhibition of tumour growth63. EW-7197, similar to LY-2157299, has an improved safety profile and is being progressed into the clinic as an immune activator151. Although JAK inhibitors have been developed for the treatment of myelofibro-sis and rheumatoid arthritis152, the potential utility of JAK inhibitors in immuno-oncology remains unknown. Targeting of the downstream effectors of the JAKs, specifically STAT3, has been pursued as a more selective and effective alternative mechanism to activate immune function. Unfortunately, as transcription factors remain a largely undruggable target class, no small-molecule drug candidates have yet emerged for STAT3 (REF. 153). Interestingly, the immunosuppressive actions of both TGFβ and STAT3 in TH17 cells has been positively linked to an enhanced expression of CD39 and CD73, resulting in increased extracellular adenosine produc-tion, thus providing an additional mechanism linking the adenosine pathway with immune suppression in the tumour microenvironment92.

Tumours often overexpress cytokines, such as macro-phage colony stimulating factor 1 (CSF1), to provide a more favourable environment for growth. The CSF1 receptor, which is exclusively expressed on cells of the myeloid lineage, is activated by autophosphorylation through its tyrosine kinase domain. Activation of the CSF1 receptor promotes myeloid proliferation, and within the tumour microenvironment promotes M1 to M2 polarization and TAM accumulation. These effects can be blocked by small-molecule CSF1 inhibitors; for example, BLZ945 in a glioblastoma67 and PLX3397 in

a breast cancer model154. In

QMCLab - REVIEWS · 2020. 10. 10. · genetic modification and returning these modified cells to...

Documents

Transcript of QMCLab - REVIEWS · 2020. 10. 10. · genetic modification and returning these modified cells to...