The Emerging Role of TYRO3 as a Therapeutic Target in Cancer

Metformin suppresses the growth of leukemia cells partly through downregulation of AXL receptor tyrosine kinase

Metformin is an anti-diabetic drug known to have anticancer activity by inhibiting mechanistic target of rapamycin (mTOR); however, other molecular mechanisms may also be involved. In this study, we examined the effects of metformin on the activity of receptor tyrosine kinases of the TAM (TYRO3, AXL, and MERTK) family, which have important roles in leukemia cell growth. The results indicated that metformin suppressed the in vitro growth of four leukemia cell lines, OCI/AML2, OCI/AML3, THP-1, and K562, in a dose-dependent manner, which corresponded to the downregulation of the expression and phosphorylation of AXL and inhibition of its downstream targets such as phosphorylation of STAT3. Furthermore, metformin augmented the suppressive effects of a small-molecule AXL inhibitor TP-0903 on the growth of OCI/AML3 and K562 cells and prevented doxorubicin-induced AXL activation in K562 cells, which induces chemoresistance in leukemia cells, thus potentiating doxorubicin anti-proliferative effects. Given that metformin also downregulated expression of TYRO3 and phosphorylation of MERTK, these findings indicate that anti-leukemic effects exerted by metformin could be partly due to the inhibition of TAM kinases. Thus, metformin has a clinical potential for patients with leukemia cells positive for AXL and the other TAM proteins as well as activated mTOR.

Identification and Validation of Immune-Related Gene Prognostic Signature for Hepatocellular Carcinoma

Immune-related genes (IRGs) have been identified as critical drivers of the initiation and progression of hepatocellular carcinoma (HCC). This study is aimed at constructing an IRG signature for HCC and validating its prognostic value in clinical application. The prognostic signature was developed by integrating multiple IRG expression data sets from TCGA and GEO databases. The IRGs were then combined with clinical features to validate the robustness of the prognostic signature through bioinformatics tools. A total of 1039 IRGs were identified in the 657 HCC samples. Subsequently, the IRGs were subjected to univariate Cox regression and LASSO Cox regression analyses in the training set to construct an IRG signature comprising nine immune-related gene pairs (IRGPs). Functional analyses revealed that the nine IRGPs were associated with tumor immune mechanisms, including cell proliferation, cell-mediated immunity, and tumorigenesis signal pathway. Concerning the overall survival rate, the IRGPs distinctly grouped the HCC samples into the high- and low-risk groups. Also, we found that the risk score based on nine IRGPs was related to clinical and pathologic factors and remained a valid independent prognostic signature after adjusting for tumor TNM, grade, and grade in multivariate Cox regression analyses. The prognostic value of the nine IRGPs was further validated by forest and nomogram plots, which revealed that it was superior to the tumor TNM, grade, and stage. Our findings suggest that the nine-IRGP signature can be effective in determining the disease outcomes of HCC patients.

TIMs, TAMs, and PS- antibody targeting: implications for cancer immunotherapy

Immunotherapy for cancer is making impressive strides at improving survival of a subset of cancer patients. To increase the breadth of patients that benefit from immunotherapy, new strategies that combat the immunosuppressive microenvironment of tumors are needed. Phosphatidylserine (PS) signaling is exploited by tumors to enhance tumor immune evasion and thus strategies to inhibit PS-mediated immune suppression have potential to increase the efficacy of immunotherapy. PS is a membrane lipid that flips to the outer surface of the cell membrane during apoptosis and/or cell stress. Externalized PS can drive efferocytosis or engage PS receptors (PSRs) to promote local immune suppression. In the tumor microenvironment (TME) PS-mediated immune suppression is often termed apoptotic mimicry. Monoclonal antibodies (mAbs) targeting PS or PSRs have been developed and are in preclinical and clinical testing. The TIM (T-cell/transmembrane, immunoglobulin, and mucin) and TAM (Tyro3, AXL, and MerTK) family of receptors are PSRs that have been shown to drive PS-mediated immune suppression in tumors. This review will highlight the development of mAbs targeting PS, TIM-3 and the TAM receptors.

TYRO3 facilitates cell growth and metastasis via activation of the Wnt/β-catenin signaling pathway in human gastric cancer cells

It has become increasingly important to identify valuable therapeutic targets to improve the prognosis of cancer patients. Although emerging evidence has suggested TYRO3 as a potential therapeutic target in various types of cancers, less is known about its role in gastric cancer (GC) development. Herein, we investigated the functional and molecular mechanisms by which TYRO3 influenced GC. TYRO3 mRNA and protein were evaluated by quantitative real-time PCR (qRT-PCR), western blotting, and immunohistochemistry. Other methods including stable transfection of TYRO3 into GC cells, wound healing, Transwell assays, CCK-8 assays, colony formation assays, immunocytochemistry in vitro, and tumorigenesis in vivo were also conducted. Our results indicated that high levels of TYRO3 significantly correlated with clinical metastasis and poor prognoses in patients with GC. In addition, TYRO3 silencing distinctively suppressed GC cell growth, invasion, and metastasis both in vitro and in vivo. Conversely, TYRO3 overexpression led to the opposite effects. Mechanistic analyses revealed that the Wnt/β-catenin signaling pathway might be involved in TYRO3-facilitated GC cell behavior. Collectively, we demonstrated that elevated TYRO3 expression contributed to GC cell growth and metastasis via the Wnt/β-catenin pathway, suggesting a novel therapeutic target for GC.

Tumour-Secreted Protein S (ProS1) Activates a Tyro3-Erk Signalling Axis and Protects Cancer Cells from Apoptosis

The TAM subfamily (Tyro3, Axl, MerTK) of receptor tyrosine kinases are implicated in several cancers, where they have been shown to support primary tumorigenesis as well as secondary resistance to cancer therapies. Relatively little is known about the oncogenic role of Tyro3, including its ligand selectivity and signalling in cancer cells. Tyro3 showed widespread protein and mRNA expression in a variety of human cancer cell lines. In SCC-25 head and neck cancer cells expressing both Tyro3 and Axl, Western blotting showed that both natural TAM ligands ProS1 and Gas6 rapidly stimulated Tyro3 and Erk kinase phosphorylation, with ProS1 eliciting a greater effect. In contrast, Gas6 was the sole stimulator of Axl and Akt kinase phosphorylation. In MGH-U3 bladder cancer cells, which express Tyro3 alone, ProS1 was again the stronger stimulator of Tyro3 and Erk stimulation but additionally stimulated Akt phosphorylation. Conditioned medium from ProS1-secreting 786-0 kidney cancer cells replicated the kinase activation effects of recombinant ProS1 in SCC-25 cells, with specificity confirmed by ProS1 ligand traps and warfarin. In addition, ProS1 protected cancer cells from acute apoptosis induced by staurosporine, as well as additionally, long-term serum starvation-induced apoptosis in MGH-U3 cells (Tyro3 only), which reflects its additional coupling to Akt signalling in these cells. In conclusion, we have shown that ProS1 is a tumour-derived functional ligand for Tyro3 that supports cancer cell survival. Furthermore, the ProS1-Tyro3 interaction is primarily coupled to Erk signalling although it displays signalling diversity dependent upon its representative expression as a TAM receptor in tumour cells.

Phosphatidylserine receptor-targeting therapies for the treatment of cancer

Asymmetric distribution of phospholipids across the plasma membrane is a unique characteristic of eukaryotic cells. Phosphatidylcholine and sphingomyelin are exposed in the outer leaflet, and phosphatidylserine (PS) is predominantly located in the inner leaflet. Redistribution of PS to the cell surface can be observed in several physiological conditions, such as apoptosis and platelet activation, or in pathological conditions, such as the release of microvesicles/exosomes from tumor tissues. PS binding to the phosphatidylserine receptor (PSR) on immune cells initiates immunosuppressive pathways that can lead to immune evasion by cancer cells. Conversely, PSR activation of cancer cells plays an important role in their survival, proliferation and metastasis. Herein, we briefly summarize both recent advances in our understanding of the pathological roles of PS and its receptor in cancer biology, as well as relevant pharmacological approaches.

Error Tolerance of Machine Learning Algorithms across Contemporary Biological Targets

Machine learning continues to make strident advances in the prediction of desired properties concerning drug development. Problematically, the efficacy of machine learning in these arenas is reliant upon highly accurate and abundant data. These two limitations, high accuracy and abundance, are often taken together; however, insight into the dataset accuracy limitation of contemporary machine learning algorithms may yield insight into whether non-bench experimental sources of data may be used to generate useful machine learning models where there is a paucity of experimental data. We took highly accurate data across six kinase types, one GPCR, one polymerase, a human protease, and HIV protease, and intentionally introduced error at varying population proportions in the datasets for each target. With the generated error in the data, we explored how the retrospective accuracy of a Naïve Bayes Network, a Random Forest Model, and a Probabilistic Neural Network model decayed as a function of error. Additionally, we explored the ability of a training dataset with an error profile resembling that produced by the Free Energy Perturbation method (FEP+) to generate machine learning models with useful retrospective capabilities. The categorical error tolerance was quite high for a Naïve Bayes Network algorithm averaging 39% error in the training set required to lose predictivity on the test set. Additionally, a Random Forest tolerated a significant degree of categorical error introduced into the training set with an average error of 29% required to lose predictivity. However, we found the Probabilistic Neural Network algorithm did not tolerate as much categorical error requiring an average of 20% error to lose predictivity. Finally, we found that a Naïve Bayes Network and a Random Forest could both use datasets with an error profile resembling that of FEP+. This work demonstrates that computational methods of known error distribution like FEP+ may be useful in generating machine learning models not based on extensive and expensive in vitro-generated datasets.