Click on an contribution to preview the abstract content.
Click on an contribution to preview the abstract content.
Tumor immune escape via tumor hypoxia and acidosis mediated PD-L1 expression as a prognostic biomarker (#30)
Philipp Knopf1, Natalie Mucha1, Andreas Maurer1, 12, Marilena Poxleitner1, Omelyan Trompak3, Bredi Tako1, Maren Harant1, Lars Zender3, 12, Marieke Fransen4, Irene Gonzalez-Menendez4, Balaji Krishnamachary7, Birgit Fehrenbacher2, Martin Schaller2, Daniela Kramer5, Leticia Quintanilla-Martinez4, 12, Klaus Schulze-Osthoff10, 11, 12, Zaver M. Bhujwalla7, 8, 9, Kamran Ghoreschi2, 6, Bernd Pichler1, 12, Manfred Kneilling1, 2, 12
1 Eberhard Karls University Tübingen , Werner Siemens Imaging Center, Department of Preclinical Imaging and Radiopharmacy, Tübingen, Baden-Württemberg, Germany
Cancer and immune cells, depend on the presence of glucose and oxygen, whereas anaerobic glycolysis leads to lactate production and low pH values. Tumor acidosis inhibits T cell function and together with hypoxia represent a tumor escape mechanism by up-regulation of programmed death ligand-1 (PD-L1)1. Thus, targeting PD-L1/PD-1 axis by monoclonal antibodies (mAbs) is a successful treatment strategy. The aim of our study was to image tumor hypoxia in vivo by optical imaging during tumor pH neutralization and elucidate the underlying mechanisms of acidosis and IFN-γ mediated PD-L1 expression.
We s.c. injected murine 5HREp-ODD-luc transduced MC38 cells (MC38-HRE-ODD-Luc) in mice treated with sodium bicarbonate or normal drinking water to longitudinally monitor the presence of tumor hypoxia non-invasively in vivo with optical imaging. Furthermore, we illuminated the mechanism of acidosis and IFN-γ induced PD-L1 up-regulation in vitro in murine MC38 as well as human HCA-7 tumor cells using siRNA mediated STAT1 knockdowns, fluorescence microscopy, flow cytometry, western blot and qRT-PCR. In vivo extracellular tumor pH neutralization experiments by sodium bicarbonate treatment using tumor models that respond (MC38 and CT26) or not respond (B16-F10 and 4T1) to anti-PD-L1 mAbs therapy were performed and tumors were resected for histopathological analysis and CD3 immunohistochemistry.
In in vivo experiments we determined in sodium bicarbonate-treated MC38-HRE-ODD-Luc tumor bearing mice an elevated luciferase expression most likely as a consequence of an enhanced T cell recruitment and activation. In vitro, IFN-γ and acidic media caused an increased PD-L1 expression in murine MC38 cells. A similar IFN-γ and acidic media induced PD-L1 up-regulation was found in human HCA-7, MCF-7 and U-87 MG cells. In vitro STAT1 siRNA knockdown experiments revealed that IFN-γ and acidosis induced up-regulation of PD-L1 is mediated by pronounced STAT1 expression and activation in a temporally resolved manner. Single sodium bicarbonate as well as combined sodium bicarbonate & PD-L1 mAbs treatment of MC38 or CT26 tumor bearing mice yielded a reduction in tumor volume and a pronounced tumoral homing of T-cell when compared to sham-treated mice. In contrast, single sodium bicarbonate treatment or single anti-PD-L1 mAbs treatment was inefficient in B16-F10 and 4T1 tumor bearing littermates.
IFN-γ together with acidosis increases PD-L1 expression by increased STAT1 expression and activation. Most interestingly, PD-L1 mAbs responsive tumor models are additionally responsive to sodium bicarbonate treatment, while tumors not responsive to PD-L1 mAbs treatment are also not responsive to extracellular tumor pH neutralization treatment. Thus, IFN-γ and acidosis induced up-regulation of PD-L1 might explain a tumor resistance mechanism.
1 Huber, V. et al. Cancer acidity: An ultimate frontier of tumor immune escape and a novel target of immunomodulation. Seminars in cancer biology 43, 74-89, doi:10.1016/j.semcancer.2017.03.001 (2017).
PK was supported by a grant from the German Academic Exchange Service (DAAD PPP USA 2018, Project-ID 57387312)
Keywords: PD-L1, Hypoxia, Acidosis, Checkpoint Inhibitor
A multimodal in vivo imaging approach to longitudinally assess and quantify infection and immune cell infiltration in preclinical models of fungal infections. (#3)
Shweta Saini1, Hannelie Korf3, Rein Verbeke5, James Dooley2, 4, Greetje Vande Velde1, Stefaan Soenen1, Conny Gysemans7, Katrien Lagrou6, Stefaan De Smedt5, Ine Lentacker5, Adrian Liston2, 4, Uwe Himmelreich1
1 University of Leuven, Biomedical MRI/ Dept Imaging and Pathology, Leuven, Belgium
Fungal infections caused by Aspergillus fumigatus commonly cause invasive pulmonary aspergillosis (IPA) in immuno-compromised patients. The development of the infection in immunodeficient patients demonstrates the importance of the host immune response in controlling aspergillosis. However, investigations of host-microbe interactions has been hampered by the lack of tools for their non-invasive assessment. We report on a multimodal approach to study the evolution of the infection (CT, BLI and MRI) and the response of the host’s immune system (19F MRI) simultaneously and longitudinally in vivo.
Four groups of Balb/c mice were studied: (1) non-infected mice receiving perfluorocarbon-particles (PFCE) ; (2-4) mice being infected with A. fumigatus (fLuc+, intranasal 106 spores), receiving PFCEs and either (2) no immune-suppression, (3) hydrocortisone acetate (HCA, 9mg/mouse) or (4) cyclophosphamide (CY, 200mg/mouse) . Pulmonary infection was followed up by BLI (IVIS Spectrum, Perkin Elmer), CT (Skyscan 1278, Bruker) and MRI (ultrashort-echo MRI using a Bruker Biospec 94/20) according to [2, 3]. 19F MRI was performed with a purpose-built, double-tuned 1H/19F coil, commencing 4h post infection/ 1h post injection of Zonyl® FSP (Z-PFCE) particles(d0/d1) . 19F and 1H MRI was performed daily. In vivo CT was performed on d1/d3. Ex vivo BLI was performed on d3 on the excised lungs.
Macrophages, identified by their characteristic high surface expression of CD11b and F4/80, phagocytosed ZPFCE-NPs without toxic effects.
In vivo data for the aspergillosis models (CY and HCA) and for non-infected mice are shown in Fig.1. A rapid influx of macrophages into the lung was seen in immunocompetent mice. Inflammation was quickly resolved with return to baseline levels after 24 hrs. In both models of immunosuppression, the immediate innate response to infection was reduced. In HCA-treated mice, the exacerbated intrusive recruitment of immune cells resulted not only in the labeling of tissue-resident macrophages but also dendritic cells, as indicated by elevated ZPFCE-NPs levels in cervical lymph nodes. Inflammation was less pronounced in the CY-treated mice.
Immune reaction was correlated with infection using BLI, CT and UTE MRI (Fig.2) No signs of infection were seen in immunocompetent mice. Mild lesion formation was seen in HCA- and strong lesion formation in CY-treated mice.
Our results demonstrate that CY-immunosuppression can lead to lethal infections (IPA) with a mild initial inflammation. HCA-based immunosuppression leads to less lesion formation but triggers acute inflammation leading to possibly lethal tissue destruction. For the first time, the dynamic profile of infection and inflammation was monitored in vivo and simultaneously in these aspergillosis models.
We are grateful for the financial support by the following funding agencies: the European Commission Marie Curie (ITN) BetaTrain (289932), the European Horizon 2020 ‘PANA’ project under grant agreement n° 686009, the Research Foundation Flanders (FWO, G.0A75.14, G.0B28.14 and G.069115N), the Agentschap voor Innovatie door Wetenschap en Technologie for the SBO NanoCoMIT (IWT SBO 140061) and the European ERA-NET project ‘CryptoView’ (3rd call of the FP7 programme Infect-ERA).
Assessment of differential local immune response to infection by A. fumigatus by 19F-MRI
(A) 19F MRI and 1H MRI were obtained from hydrocortisone acetate (HCA), cyclophosphamide (CY) and infected-immunocompetent (I-IC) mice as well as non-infected control mice (N-IC) after systemic injection of ZPFCE-NPs on day 0 (4h) and day 1.
(B) Quantification of 19F MRI from the lung region by comparison with a reference (R in panel A) containing 30mM ZPFCE-NPs.
(C) 19F MRI signal was also observed from the lymph node region for the HCA group on day 1. Mean 19F MR signal intensity was quantified with respect to the 30mM reference placed next to each animal.
Assessment of lung infection.
(A) In vivo follow-up of A. fumigatus infected mice using computed tomography (CT) and 1H MR imaging. Mice were scanned on day 1 and day 3, post infection for the confirmation of pulmonary fungal infection progression using CT imaging where lesions in HCA (hydrocortisone acetate and CY (Cyclophosphamide) mice can be observed as compared to I-IC and non-infected immunocompetent (N-IC) group.
(B) Quantitative estimation of lung lesion volume based on UTE-MR images (mean±SEM).
Keywords: 19F MRI, Bioluminescence, CT, fungal infection, immune response