PDX powered humanized mice: Advancing the cutting edge in cancer modelling
Scheme for the generation of humanized patient-derived xenograft models. HSPC: hematopoietic stem and precursor cell. Figure reproduced without modification from Cho, S. Patient-derived xenografts as compatible models for precision oncology. Lab Anim Res36, 14 (2020). https://doi.org/10.1186/s42826-020-00045-1 . Link to article: https://rdcu.be/b8rC5. Use of this figure here is permitted under the Creative Commons Attribution 4.0 International License.
Lakshman Varanasi, PhD, Science Associate October 13, 2020
Humanized mice grafted with human tumours are excellent platforms for modelling the human tumour microenvironment and find utility in a range of oncology studies. These include tumorigenesis, the tumour microenvironment, patient- derived xenografts, metastasis, conventional drug- or drug-candidate efficacy, immuno-oncology therapy (better known as immunotherapy) and its efficacy, novel experimental therapies, side- or adverse effects of therapies, mechanisms of resistance to therapy, and cancer stem cells. Immunotherapy is inclusive of immune checkpoint blockade therapy, adaptive cell therapy, and cancer vaccines. An understanding of the tumour microenvironment and of its interactions with the host has many benefits, not least the design of better therapy.
The tumour microenvironment is the spatial boundary between the tumour and the host tissues. It comprises the extracellular matrix of connective tissue, blood vessels, fibroblasts, immune cells, and signalling molecules, and is characterised by a complex interplay between the cancer cells and the surrounding healthy connective tissue (ECM) and immune cells. Cancer cells and recruited antigen-presenting cells (APCs) bear on their surface ligands that bind to specific immune checkpoint receptors on immune effector cells, and suppress their protective response. They blunt the host’s immune response, to preserve the integrity of the tumour mass, and to allow it the latitude to grow. The immune effector cells in the tumour microenvironment are not rogue per se, as they do not turn against the host, but are no longer true to their mission. The growing tumour’s cells often display in fewer numbers, or not at all, “foreign” peptides (on their cell-surface MHCs) that would normally attract the deadly attention of NK cells and T-lymphocytes. Tumour-associated macrophages (TAMs) and myeloid-derived suppressor cells (MDSCs), both signs of an ascendant tumour, populate the tumour in greater numbers than they do the healthy peripheral tissues. NK cells in the tumour mass are fewer in the tumour, and more enervated than outside it, and the secretion of pro-inflammatory cytokines in the microenvironment is diminished. Tumour-specific factors determine the nature and proportion of the tumour-infiltrating lymphocytes (TILs) in the tumour tissue. Comparative analyses of various established cancer cell-lines grown in humanized mice have been done to improve understanding of immune cell composition and residency in the tumour microenvironment.
Patient-derived xenograft (PDX) mouse models are created by implantation of fresh human tumour tissue, usually less than 2 cubic millimetres in size, subcutaneously or orthotopically. The tumours, either from treatment-naive or resistaont patients, are allowed to grow till they are about 1 cubic centimetre large and then excised, preserved, characterised, and possibly used again in other humanized mice (the last, a valuable feature of mice PDX models). For cancers of the blood, PBMCs or bone marrow samples are engrafted directly in the blood, or implanted in the bone marrow. Engrafted tumours are observed to preserve their morphology and structural and genetic heterogeneity in the latest variants of humanized mice. They grow as well in humanized mice as in non-humanized ones, and respond to currently used cancer drug therapies as patients do; importantly, the human immune cells descended from the haematopoietic (adult) stem cells (HSCs) co-implanted with the PDX, also respond to anti-cancer immuno-oncological therapies. The PDX and the human immune system (HIS) in the murine host affect each other reciprocally, and an immune response towards the tumour, in this model, does not require the tumour and stem-cell donors to be HLA (Human Leukocyte Antigen) matched. PDX-humanized mice models enable the evaluation of B- and T-cell responses to the tumour and also the efficacy of novel compounds in effecting an attack on the tumour by proxy, i.e. through the immune system. The PDX in the humanized mouse appears to offer greater scope for pre-clinical in vitro studies than its peer models.
As the science underlying the interaction between tumour and the host immune system and the tumour has become clearer, new ways of leveraging one to fight the other have also come to light (Ref to MSKCC discovery). Immunotherapy holds promise for some cancers hitherto refractory to treatment. It has brought hope to patients suffering from some cancers, although the therapies haven’t matured and much about the physiology surrounding the therapy’s mechanism remains unknown. Because immunotherapy acts against cancer through a functional HIS, conventional models do not suffice. Immunotherapies used so far, such as the antibody inhibitors against PD-1 AND CTLA4, have been tested in murine systems (inhibitors and the target proteins being of murine origin). The human antibodies were tested in vitro and in cynomolgus monkeys. Inadequate, but something researchers had to make do with, for want of a better alternative. None of these simulations was reliably predictive of outcomes in human systems. But that is a thing of the past now, and newer and more suitable humanized mouse variants are available. For instance, the targeted mutation in the IL2 receptor 𝛾-chain locus in the NOD/SCID mouse strain has yielded a humanized mouse strain, the NSGTM strain, that is deficient in T-cells, B-cells, macrophages, and NK and NKT cells; it has quickly become the strain of choice for studies which require compatibility between the graft and the host. Likewise the BLT model (Bone marrow- Liver-Thymus). Patient-derived xenografts further empower the humanized mouse model and help bypass some limitations of in vitro cancer research. Key features of the tumour, including its histology genetic heterogeneity, are preserved in such models, and data from these have more predictive value than those from non-humanised mice or in vitro studies; this is affirmed by results from drug efficacy studies on PDX humanized mice, which correlate well with clinical outcomes in cancer patients.
PDX humanized mice Hu-mice with PDXs have come in useful in the study of a variety of cancers, including EBV-induced human B-cell lymphoma, rhabdomyosarcoma, breast cancer, head and neck cancer, renal cell carcinoma, epithelial ovarian cancer, melanoma, and leukaemia. Humanized bone constructs too (artificial bone containing bone cells and secreted extracellular matrix) can be now grafted into mice, and are of value for studying bone metastasis in various cancers. PDX humanized mice may be used for metastatic dissemination studies, tumour heterogeneity, cancer stem cell types, and so on. The tumours (and host mice) may be selected on the basis of desired markers or drug sensitivities. Human markers secreted by the tumour into the murine blood can also be used as candidates for diagnosis or prognosis, or for patient stratification.
While the above-described model is cause for optimism to a field critical to human health, it is sobering to keep in mind that the best model is still a model; the application of any findings from it, to the human system, and the subsequent interpretation of results, must be done with care.