Malignant Neoplasms

Example 7

The MTT Cell Proliferation assay determines cell survival following apple stem cell extract treatment. The purpose was to evaluate the potential anti-tumor activity of apple stem cell extracts as well as to evaluate the dose-dependent cell cytotoxicity.

Principle: Treated cells are exposed to 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT). MTT enters living cells and passes into the mitochondria where it is reduced by mitochondrial succinate dehydrogenase to an insoluble, colored (dark purple) formazan product. The cells are then solubilized with DMSO and the released, solubilized formazan is measured spectrophotometrically. The MTT assay measures cell viability based on the generation of reducing equivalents. Reduction of MTT only occurs in metabolically active cells, so the level of activity is a measure of the viability of the cells. The percentage cell viability is calculated against untreated cells.

Method: A549 and NCI-H520 lung cancer cell lines and L132 lung epithelial cell line were used to determine the plant stem cell treatment tumor-specific cytotoxicity. The cell lines were maintained in Minimal Essential Media supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (100 μg/ml) in a 5% CO₂ at 37 Celsius. Cells were seeded at 5×10³ cells/well in 96-well plates and incubated for 48 hours. Triplicates of eight concentrations of the apple stem cell extract were added to the media and cells were incubated for 24 hours. This was followed by removal of media and subsequent washing with the phosphate saline solution. Cell proliferation was measured using the MTT Cell Proliferation Kit I (Boehringer Mannheim, Indianapolis, IN) New medium containing 50 μl of MTT solution (5 mg/ml) was added to each well and cultures were incubated a further 4 hours. Following this incubation, DMSO was added and the cell viability was determined by the absorbance at 570 nm by a microplate reader.

In order to determine the effectiveness of apple stem cell extracts as an anti-tumor biological agent, an MTT assay was carried out and IC50 values were calculated. IC50 is the half maximal inhibitory function concentration of a drug or compound required to inhibit a biological process. The measured process is cell death.

Results: ASC-Treated Human Lung Adenocarcinoma Cell Line A549.

Results: ASC-Treated Human Squamous Carcinoma Cell Line NCI-H520.

Results: ASC-treated Lung Epithelial Cell Line L132.

Summary Results: Cytotoxicity of Apple Stem Cell Extracts.

Apple stem cell extracts killed lung cancer cells lines A549 and NCI-H520 at relatively low doses: IC50s were 12.58 and 10.21 μg/ml respectively as compared to 127.46 μg/ml for the lung epithelial cell line L132. Near complete anti-tumor activity was seen at a dose of 250 μg/ml in both the lung cancer cell lines. This same dose spared more than one half of the L132 cells. See Tables 7-10. The data revealed that apple stem cell extract is cytotoxic to lung cancer cells while sparing lung epithelial cells. FIG. 6 shows a graphical representation of cytotoxicity activity of apple stem cell extracts on lung tumor cell lines A549, NCIH520 and on L132 lung epithelial cell line (marked “Normal”). The γ-axis is the mean % of cells killed by the indicated treatment compared to unexposed cells. The difference in cytotoxicity levels was statistically significant at p≤05.

Example 9

The experiment of Example 7 was repeated substituting other plant materials for ASC. Plant stem cell materials included Dandelion Root Extract (DRE), Aloe Vera Juice (AVJ), Apple Fiber Powder (AFP), Ginkgo Leaf Extract (GLE), Lingonberry Stem Cells (LSC), Orchid Stem Cells (OSC) as described in Examples 1 and 2. The concentrations of plant materials used were nominally 250, 100, 50, 25, 6.25, 3.125, 1.562, and 0.781 μg/mL. These materials were tested only for cells the human lung epithelial cell line L132 (as a proxy for normal epithelial cells) and for cells of the human lung adenocarcinoma cell line A549 (as a proxy for lung cancer cells).

A549 cells lung cancer cell line cytotoxicity results for each of the treatment materials.

DRE-Treated Lung Cancer Cell Line A549 Cells.

AVJ-Treated Lung Cancer Cell line A549 Cells.

AFP-Treated Lung Cancer Cell line A549 Cells.

GLE-treated Lung Cancer Cell line A549 Cells.

LSC-treated lung cancer cell lines A549 cells.

OSC-treated Lung Cancer Cell line A549 Cells.

L132 cells (“normal” lung epithelial cell line) cytotoxicity results for each of the treatment materials.

DRE-Treated Lung Epithelial Cell Line L132 cells.

AVJ-Treated Lung Epithelial Cell Line L132 cells.

AFP-Treated Lung Epithelial Cell Line L132 cells.

GLE-Treated Lung Epithelial Cell Line L132 cells.

LSC-Treated Lung Epithelial Cell Line L132 cells.

OSC-Treated Lung Epithelial Cell Line L132 cells.

Calculated values.

Example 9

An analysis of gene ontology (GO) categories associated with ADAR1 dependent cells revealed that NCI-H1650 and HCC366 (“HCC-366”), two ADAR1 dependent cell lines, both have elevated basal expression of interferon inducible genes (FIG. 35). The expression levels of interferon-inducible genes were also elevated in NCI-H196 cells (FIG. 36).

In light of the correlation between ADAR1 dependency and the expression of interferon-inducible genes, additional cancer cell lines from the Molecular Signatures Database (MSigDB) (Liberzon et al. (2015) Cell Systems 1:417-425) was examined. Cancer Cell Line Encyclopedia (CCLE) clustering was performed based on the Type I/Interferon-a gene set, which contained 97 genes including PKR. The resulting cluster included HCC366, NCI-H1650 and 9 additional lung cell lines. Among these cell lines, HCC1438 and NCI-H596 were sensitive to knockout of ADAR1 by lentiviral CRISPR-Cas9 (FIG. 37).

All the above-identified ADAR1 dependent cancer cell lines showed elevated interferon signaling markers, e.g., phosphorylation of STAT1 and expression of interferon-stimulated gene (ISGs) (FIG. 38). Elevated interferon signaling in the ADAR1 dependent cancer cell lines did not necessarily lead to PD-L1 overexpression (FIG. 38). Cell lines in the high interferon signaling cluster (LN215_CENTRAL_NERVOUS_SYSTEM, NCIH596_LUNG, HCC1438_LUNG, T3M10_LUNG, NCIH1869_LUNG, SW900_LUNG, HCC366_LUNG, SKLU1_LUNG, NCIH1650_LUNG, HCC4006_LUNG, and NCIH1648_LUNG) displayed high IFN-β, but not IFN-α (FIG. 39). As such, cancer cell lines sensitive to ADAR1 or ISG15 knockdown displayed elevated interferon secretion and downstream signaling. To further investigate the relationship between ADAR1 and IFN-β secretion, it was found that ADAR1 knockout led to amplified IFN-β secretion in cell lines primed with high basal interferon activation (FIG. 40). It was also found that ADAR1 dependent cell lines do not show enhanced sensitivity to IFN-α or IFN-β alone (FIG. 41).

Example 1

The MCA-miner method disclosed herein in FIGS. 2A-2C, when used together with BRL, offers the power of rule list interpretability while maintaining the predictive capabilities of already established machine learning methods.

The performance and computational efficiency of the new MCA-miner is benchmarked against the “Titanic” dataset, as well as the following five (5) datasets available in the UCI Machine Learning Repository: “Adult,” “Autism Screening Adult,” “Breast Cancer Wisconsin (Diagnostic),” “Heart Disease,” and “HIV-1 protease cleavage,” which are designated as Adult, ASD, Cancer, Heart, and HIV, respectively. These datasets represent a wide variety of real-world experiments and observations, thus enabling the improvements described herein to be compared against the original BRL implementation using the FP-Growth miner.

All six benchmark datasets correspond to binary classification tasks. The experiments were conducted using the same set up in each of the benchmarks. First, the dataset is transformed into a format that is compatible with the disclosed BRL implementation. Second, all continuous attributes are quantized into either two (2) or three (3) categories, while keeping the original categories of all other variables. It is worth noting that depending on the dataset and how its data was originally collected, the existing taxonomy and expert domain knowledge are prioritized in some instances to generate the continuous variable quantization. A balanced quantization is generated when no other information was available. Third, a model is trained and tested using 5-fold cross-validations, reporting the average accuracy and Area Under the ROC Curve (AUC) as model performance measurements.

Table 1 presents the empirical result of comparing both implementations. The notation in the table follows the definitions above. To strive for a fair comparison between both implementations, the parameters rmax=2 and smin=0:3 are fixed for both methods, and in particular for MCA-miner μmin=0:5 and M=70 are also set. The multi-core implementations for both the new MCA-miner and BRL were executed on six parallel processes, and stopped when the Gelman & Rubin parameter satisfied {circumflex over (R)}≤1.05. All the experiments were run using a single AWS EC2 c5.18×large instance with 72 cores.

It is clear from the experiments in Table 1 that the new MCA-miner matches the performance of FP-Growth in each case, while significantly reducing the computation time required to mine rules and train a BRL model.