Background: Targeting of MDSCs is a major clinical challenge in the era of immunotherapy. Antibodies which deplete MDSCs in murine models can reactivate T cell responses. In humans such approaches have not developed due to difficulties in identifying targets amenable to clinical translation. Methods: RNA-sequencing of M-MDSCs and G-MDSCs from cancer patients was undertaken. Flow cytometry and immunohistochemistry of blood and tumours determined MDSC CD33 expression. MDSCs were treated with Gemtuzumab ozogamicin and internalisation kinetics, and cell death mechanisms determined by flow cytometry, confocal microscopy and electron microscopy. Effects on T cell proliferation and CART cell anti-tumour cytotoxicity were identified in the presence of Gemtuzumab ozogamicin. Findings: RNA-sequencing of human M-MDSCs and G-MDSCs identified transcriptomic differences, but that CD33 is a common surface marker. Flow cytometry indicated CD33 expression is higher on M-MDSCs, and CD33+ MDSCs are found in the blood and tumours regardless of cancer subtype. Treatment of human MDSCs leads to Gemtuzumab ozogamicin internalisation, increased p-ATM, and cell death; restoring T cell proliferation. Anti-GD2-/mesothelin-/EGFRvIII-CART cell activity is enhanced in combination with the anti-MDSC effects of Gemtuzumab ozogamicin. Interpretation: The study identifies that M-MDSCs and G-MDSCs are transcriptomically different but CD33 is a therapeutic target on peripheral and infiltrating MDSCs across cancer subtypes. The immunotoxin Gemtuzumab ozogamicin can deplete MDSCs providing a translational approach to reactivate T cell and CART cell responses against multiple cancers. In the rare conditions of HLH/MAS gemtuzumab ozogamicin provides a novel antimyeloid strategy.

Therapeutic approaches which aim to target Acute Myeloid Leukaemia through enhancement of patients’ immune responses have demonstrated limited efficacy to date, despite encouraging preclinical data. Examination of AML patients treated with azacitidine (AZA) and vorinostat (VOR) in a Phase II trial, demonstrated an increase in the expression of Cancer‐Testis Antigens (MAGE, RAGE, LAGE, SSX2 and TRAG3) on blasts and that these can be recognised by circulating antigen‐specific T cells. Although the T cells have the potential to be activated by these unmasked antigens, the low arginine microenvironment created by AML blast Arginase II activity acts a metabolic brake leading to T cell exhaustion. T cells exhibit impaired proliferation, reduced IFN‐γ release and PD‐1 up‐regulation in response to antigen stimulation under low arginine conditions. Inhibition of arginine metabolism enhanced the proliferation and cytotoxicity of anti‐NY‐ESO T cells against AZA/VOR treated AML blasts, and can boost anti‐CD33 Chimeric Antigen Receptor‐T cell cytotoxicity. Therefore, measurement of plasma arginine concentrations in combination with therapeutic targeting of arginase activity in AML blasts could be a key adjunct to immunotherapy.

Haematological and solid cancers catabolise the semi-essential amino acid arginine to drive cell proliferation. However, the resulting low arginine microenvironment also impairs chimeric antigen receptor T cells (CAR-T) cell proliferation, limiting their efficacy in clinical trials against haematological and solid malignancies. T cells are susceptible to the low arginine microenvironment due to the low expression of the arginine re-synthesis enzymes argininosuccinate synthase (ASS) and ornithine transcarbamylase (OTC). We demonstrate T cells can be re-engineered to express functional ASS or OTC enzymes, in concert with different chimeric antigen receptors. Enzyme modifications increase CAR-T cell proliferation, with no loss of CAR cytotoxicity or increased exhaustion. In vivo, enzyme-modified CAR-T cells lead to enhanced clearance of leukaemia or solid tumour burden, providing the first metabolic modification to enhance CAR-T cell therapies.

Renewed interest in the use of therapeutic enzymes combined with an improved knowledge of cancer cell metabolism, has led to the translation of several arginine depletion strategies into early phase clinical trials. Arginine auxotrophic tumors are reliant on extracellular arginine, due to the downregulation of arginosuccinate synthetase or ornithine transcarbamylase-key enzymes for intracellular arginine recycling. Engineered arginine catabolic enzymes such as recombinant human arginase (rhArg1-PEG) and arginine deiminase (ADI-PEG) have demonstrated cytotoxicity against arginine auxotrophic tumors. In this review, we discuss the molecular events triggered by extracellular arginine depletion that contribute to tumor cell death.

Neuroblastoma is the most common childhood solid tumor, yet the prognosis for high-risk disease remains poor. We demonstrate here that arginase 2 (ARG2) drives neuroblastoma cell proliferation via regulation of arginine metabolism. Targeting arginine metabolism, either by blocking cationic amino acid transporter 1 (CAT-1)-dependent arginine uptake in vitro or therapeutic depletion of arginine by pegylated-recombinant arginase BCT-100, significantly delayed tumor development and prolonged murine survival. Tumor cells polarized infiltrating-monocytes to a M1-macrophage phenotype, which released IL-1β and TNF-α in a RAC-alpha serine/threonine-protein kinase (AKT)-dependent manner. IL-1β and TNF-α established a feedback loop to upregulate ARG2 expression via p38 and extracellular regulated kinases 1/2 (ERK1/2) signalling in neuroblastoma and neural crest-derived cells. Proteomic analysis revealed that enrichment of IL-1β and TNF-α in stage IV human tumor microenvironments was associated with a worse prognosis. These data thus describe an immune-metabolic regulatory loop between tumor cells and infiltrating myeloid cells regulating ARG2, which can be clinically exploited.