Skip to main content

Young Scientist Journal

Young Scientist Journal

Regulation of the PPARꝩ Pathway using Antagonists and In-verse Agonists across 2D and 3D Models

ABSTRACT

Lung cancer is the leading cause of cancer-related deaths and is known to effectively colonize other parts of the body, often evading the immune system as it metastasizes or spreads. Preliminary data from our lab indicated that the PPARγ pathway may enhance lung cancer cells’ ability to adapt and survive in new microenvironments resulting in metastasis. To test this, three lung cancer cell lines were treated with PPARγ antagonists (GW9662, T0070907) and the inverse agonist BAY-0069. In 2D monolayers, all three compounds significantly reduce cell proliferation in a dose-dependent manner in two of three lines. However, growth inhibition was markedly weaker when cells were cultured as neurospheres or 3D tumor organoids, suggesting that PPARγ’s role in proliferation is highly dependent on model complexity and tumor microenvironment. These findings highlight the importance of using complex relevant models for drug testing. a complex model such as a cerebral organoid could be utilized in future studies to test whether proliferation and colonization of the brain could be reduced with PPARγ pathway inhibition. If successful, targeting PPARγ signaling may offer a novel therapeutic strategy to improve outcomes in patients with lung cancer brain metastasis.

INTRODUCTION.

Lung cancer is the leading cause of cancer-related death worldwide and the 3rd most common diagnosed cancer [1]. There are two types, Small Cell Lung Cancer (SCLC) and Non-Small Cell Lung Cancer (NSCLC). 80-85% of lung cancer cases are NSCLC, growing at a slower rate than SCLC. Growth starts in the lung(s) and expands to various parts of the body as the initial tumor cells colonize other organs. Lung cancer has the capability to metastasize using a process called EMT or epithelial-mesenchymal transition [2]. This is a process where epithelial tumor cells lose their cell polarity and cell-cell adhesion and become mesenchymal cells with acquired migratory and invasive properties. Cancer cell plasticity enables metastatic cells to colonize secondary tumor sites such as the brain, opposite lung, and other organs [3].

Both non-small cell lung cancer and small cell lung cancer are predominately diagnosed in smokers, with > 70% of NSCLC diagnoses being considered smokers. Tobacco smoke has over 7,000 chemicals and 70 known carcinogens, with carcinogens being classified as anything that causes cancer [4]. Shortness of breath, chest pain, and lung infections are all symptoms of lung cancer and a result of heavy smoking. Furthermore, smoking has effects at the cellular level, causing direct DNA damage which contributes heavily to genetic mutations, ultimately initiating tumor formation and progression [4]. Previous studies show that smoking and lung cancer development are closely related, secondhand smoke also causing an increase in diagnosis frequency in comparison to non-smokers [5]. Carcinogens linked to tobacco increase tumor resistance to treatment and increase the cancers aggression, making lung cancer more common. The amount of cancer mutations overrides the tumor suppressors (p53, K-ras, etc.) and the cumulative mutations make it challenging to find an effective target for treatment.

Current treatments for lung cancer target rapidly dividing cells and often use prophylactic irradiation to prevent metastasis in organs such as the brain. Although the acute side effects of prophylactic cranial irradiation can be manageable (e.g., hair loss, fatigue, nausea, etc.), long-term side effects can be debilitating to the patient and their family members [6].  For example, radiation can degrade the myelin sheath, known as demyelination, which leads to personality changes, motor problems, and overall neurocognitive decline [7].

Chemotherapy is another therapeutic option that works to target rapidly dividing cancer cells; however, this method of treatment is often non-specific and frequently causes an overall negative effect on the body through off-target effects and cellular toxicity to normal, noncancerous cells [8]. For instance, many cancer patients lose their hair or begin to have digestive issues, due to the rapid cell division in the hair follicles and gastrointestinal tract and, thus, increased susceptibility to chemotherapy. As an alternative to chemotherapy, several targeted therapy drugs have been developed and approved for lung cancer patients.  Notable developments in targeted cancer therapy include categories such as monoclonal antibodies, small molecule inhibitors, and immunotoxins [9].

Despite advancements in targeted therapy, approximately 90% of promising pre-clinical drug candidates fail in clinical trials [10]. Pre-clinical models that do not accurately represent the patient disease, is the reason. It is important to model the complexity of tissue morphology and host-tumor interaction.  As such, we utilize an array of culture models such as neurospheres (see S1), tumor organoids and cerebral organoids to better represent tumor complexity and the microenvironment of lung cancer. These 2D and 3D models allow for more accurate drug testing results and provide enhanced insight into what patient treatment response could look like.

A potential target for specific drug treatment for lung cancer patients is the peroxisome proliferator-activated receptor gamma or PPARγ [11]. This pathway is primarily expressed in adipose tissue, or fat found virtually everywhere in the body [12]. Research shows that the PPARγ pathway can function as both a tumor suppressor and tumor-promoting factor, depending on the type of cancer [13]. Specific ligands or drugs can bind to PPARγ and regulate the pathway to either promote or prevent cancer cells from gaining migratory and invasive characteristics required for metastasis. Utilizing the antagonists GW9662 and T0070907, as well as a PPARγ inverse agonist BAY-0069, we aimed to investigate how suppression of the PPARγ signaling pathway affects the growth of two NSCLC cell lines and one SCLC cell line grown in varying 2D and 3D conformations.

MATERIALS AND METHODS.

Cell Line Selection.

In our experiments, three human cell lines were provided by the National Cancer Institute (NCI) which are all approved for research purposes. The H1048 small cell lung cancer (SCLC) cells are from a female, non-smoking patient from 1985. The H1299 cell line is a non-small cell lung cancer (NSCLC) cell line from a 43-year-old, white, non-smoking, male patient who also had prior radiation therapy. A549 is a non-small cell lung cancer line (NSCLC) from a 58-year-old white male. All three of these cell lines are epithelial cell types and have adherent growth properties, meaning the cells can bind to one another or stick to the surfaces of their respective cultureware (more info in Test Subjects section). The type of study we conducted was observational to analyze cellular phenotypes in response to inhibition of the PPARγ pathway using three different lung cancer lines.

Inhibitor Selection.

We also used three different inhibitors (GW9662, BAY-0069, and T0070907).  Among these were two PPARγ antagonists (GW9662 and T0070907) and one inverse agonist (BAY-0069), which binds to PPARγ. We identified these ligands or drugs by reviewing previous literature on cancer and the PPARγ pathway.

Cell Viability.

The cell viability was measured using an AlamarBlue assay. Briefly, this assay exposes cells to a resazurin-based reagent that, upon entering live cells, is reduced to a highly fluorescent compound known as resorufin. The raw fluorescent units were measured by the microplate reader and then converted into normalized numbers corresponding to viable cells during analysis.  All 2D and 3D models were incubated with the alamarBlue reagent for three hours. Some examples of factors that could interfere with our analysis could be compensatory mechanisms of other signaling pathways or simply contamination or a lack of nutrients due to insufficient maintenance of cultures.

Test Subjects.

Typical preparation before experimentation involved thawing and culturing each cell line in medium made with 10% Fetal Bovine Serum, 5% Penicillin/Streptomycin, and a bottle of RPMI with glutamine.  All three cell lines were cultured and allowed to expand over the first week. The sampling method used is biological sampling or repeated sampling to reduce variability. This was done by using a multitude of samples for each condition (6 wells per condition, per cell line). The drug concentrations were our experimental units and the target in each model was the PPARγ pathway. This was a comparative study, comparing the effects of three drugs (BAY-0069, GW9662, and T0070907) and one control. Samples were divided up using group allocation into a control group, a BAY-0069 treated group, a GW9662 treated group, and a T0070907 treated group for each cell line, neurosphere, and tumor organoid.

Data Collection.

Types of data collected included 1) microscope images, which were collected daily to monitor the morphology of the three models: cell lines, neurospheres,(see S1) and tumor organoids, and 2) plate readings, using a Synergy H1 plate reader, to measure cell viability three days post-drug treatment.

Our measured variable was cell viability, measured using the alamarBlue assay through fluorescence readings after the drug treatment. This allowed us to observe the effects of each inhibitor on the three cell lines, whether inhibition of the PPARγ pathway increased or decreased cell viability. To process the 96-well plates, we used a Synergy H1 microplate reader (BioTek); the specific assay we used was the alamarBlue fluorescence cell viability assay. The overall fluorescence in each well is directly proportional to the viability of the cells per well.  Relative Fluorescence Units (RFU) were measured after 3 hours of incubation with alamarBlue using the BioTek Gen5 software.  For descriptive statistics, the fluorescence intensity was measured and displayed in graphical representation (See figures 1-2).

Regarding sample size, each drug concentration had an n of 5-6.  We used 3 cell lines in total, with 3 different architectural models (2D cell lines, neurospheres, and 3D tumor organoids).  For each drug, the means for the 1 µM, 10 µM, and 20 µM treatment groups were normalized to the 0 µM group and, thus, graphed as % of control.  Error bars were based on the standard error of the mean (SEM) to reflect the precision of the sample mean as an estimate of the population mean.   Furthermore, we used independent two-tailed T-tests to compare individual treatment groups to the control for each cell line and model. A p-value threshold of .05 was used to find statistical significance. To address confounding variables, three different cell lines were used with different origins, as well as three different model types, including 2D cell lines, neurospheres, and 3D tumor organoids.  Key bias control methods that we utilized included vehicle controls, assay standardization, clear metrics for experimental evaluation, and using at least 5-6 replicates per condition to minimize variability and ensure that the resulting data was truly representative of each cell line’s drug sensitivity.  All experiments were performed under the same conditions and environments. Cell counts and dilutions were all calculated using the LUNA-II Cell Counter by Logos Biosystems. No missing data was reported; wells with lack of a model or contamination were excluded from the analysis. Fluorescence was measured using the Synergy H1 microplate reader.  Data analysis and t-tests were performed using Microsoft Excel, and figures were also designed on Excel.

RESULTS.

Three lung cancer cell lines (H1048, A549, and H1299) were each treated with three different PPARꝩ ligand’1µM, 10 µM, and 20 µM.

Cell Lines.

For the H1048 cell line, PPARꝩ ligand GW9662 had a 53.43% cell viability for the 1 µM concentration, 29.43% for the 10 µM, and 4.14% µM for the 20 µM concentration (p = 0.0001 for 10 µM and 20 µM). This indicates that the higher concentration of drug or ligand decreases cell viability. 1µM of GW9662 was enough to see a decrease of ~55% compared to controls, indicating the success in inhibiting PPARꝩ. The BAY-0069 Ligand had a 73.35% cell viability for 1 µM, 45.52% for 10 µM, and 40.69% for 20 µM (p = 0.0001 for 10 µM and 20 µM). This indicates when the H1048 cell line is treated with the BAY-0069, as concentration increases cell proliferation is reduced, indicating success in inhibiting the PPARꝩ pathway. However, in comparison to the GW9662 drug, BAY-0069 has a lesser effect on reducing cell viability. For example, at 1 µM of GW9662, cell viability was at 53.43% compared to 1 µM of BAY-0069 with a 73.35% cell viability at the same concentration. T0070907 had a 69.37% cell viability for 1 µM, 29.93% for 10 µM, and 5.18% for 20 µM (p = 0.0001 for 10 µM and 20 µM). Consistent with previous results, T0070907 also indicates inhibition of the PPARꝩ pathway as the cell viability rapidly decreases upon the increase of drug concentration. Furthermore, our results have strong statistical significance, backing our hypothesis that inhibition of the PPARꝩ pathway results in reduced tumor cell viability. (See Fig. 1A)

Figure 1. Lung cancer cell lines and the effects of PPARꝩ inhibitors. (p=0.05*, p=0.01**, p=0.001***, p=0.0001****)

The A549 cell line’s cell viability for the GW9662 PPARꝩ ligand was 100.20% for 1 µM, 98.63% for 10 µM, and 103.63% for 20 µM (p=0.05 for 20 µM). These results for the A549 cell line contradict previous results, with added drug GW9662 causing an increase in cell viability for 1 µM and 20 µM, which may be a result of the activation of the PPARꝩ pathway resulting in cell growth and proliferation. Due to the reduced significance, we are 95% confident that inhibition of the PPARꝩ pathway occurred. Inhibition of PPARꝩ can be noted at the 10 µM concentration where there is a 1.37% decrease in cell viability from our 100% control threshold. For the BAY-0069 ligand cell viability was 101.94% for 1 µM, 105.66% for 10 µM, and 126.55% for 20 µM (p=0.0001). The pattern of cell viability increases in A549 as a result of increased drug concentration remains consistent with the GW9662 and BAY-0069 results for the A549 cell line. This enforces the hypothesis that when PPARꝩ is activated it promotes cancer cell proliferation. In comparison to control, overall cell viability increases by 34.1% across 1-20 µM BAY-0069 concentrations. Finally, the T0070907 ligand had a cell viability of 91.77% for 1 µM, 89.41% for 10 µM, and 100.59% for 20 µM concentration. Similar to the increase in viability at 1 and 20 µM for GW9662, T0070907 had an 11.18% increase in cell viability at the 20 µM concentration in comparison to the 10 µM concentration’s cell viability which could be the result of nonspecific ligand binding to other molecular receptors, encouraging cancer cell growth (See future works for more information). (See Fig. 1B)

The H1299 cell line’s cell viability for the GW9662 PPARꝩ ligand was 98.75% for 1 µM (p=0.01), 96.08% for 10 µM (p=0.01), and 66.99% for 20 µM. For the 1 µM and 10 µM experiment groups For the BAY-0069 ligand cell viability was 98.24% for 1 µM, 89.4% for 10 µM, and 70.21% for 20 µM. Finally, the T0070907 ligand had a cell viability of 100.39% for 1 µM, 97.41% for 10 µM, and 39.48% for 20 µM concentration. (See Fig. 1C)

Neurospheres.

The H1048 neurosphere’s cell viability for the GW9662 PPARꝩ ligand was 92.79% for 1 µM, 68.79% for 10 µM, and 44.69% for 20 µM. For BAY0069 ligand cell viability was 101.88% for 1 µM, 97.11% for 10 µM, and 85.57% for 20 µM. Finally, the T0070907 ligand had a cell viability of 96.42% for 1 µM, 58.85% for 10 µM, and 52.40% for 20 µM concentration. The H1299 neurosphere’s cell viability for the GW9662 PPARꝩ ligand was 83.53% for 1 µM, 93.51% for 10 µM, and 84.49% for 20 µM. For BAY0069 ligand cell viability was 70.48% for 1 µM, 65.12% for 10 µM, and 84.33% for 20 µM. Finally, the T0070907 ligand had a cell viability of 103.05% for 1 µM, 106.14% for 10 µM, and 129.31% for 20 µM concentration. The A549 neurosphere’s cell viability for the GW9662 PPARꝩ ligand was 46.09% for 1 µM, 58.33% for 10 µM, and 48.41% for 20 µM. For BAY0069 ligand cell viability was 97.17% for 1 µM, 108.70% for 10 µM, and 111.20% for 20 µM. Finally, the T0070907 ligand had a cell viability of 101.46% for 1 µM, 86.02% for 10 µM, and 127.45% for 20 µM concentration.

Tumor Organoids.

The H1048 tumor organoid’s cell viability for the GW9662 PPARꝩ ligand was 108.6% for 1 µM, 97.22% for 10 µM, and 126.10% for 20 µM. For BAY-0069 ligand cell viability was 96.45% for 1 µM, 90.33% for 10 µM, and 111.51% for 20 µM. Finally, the T0070907 ligand had a cell viability of 91.93% for 1 µM, 81.99% for 10 µM, and 82.49% for the 20 µM concentration (See Fig. 2A). The H1299 tumor organoid’s cell viability for the GW9662 PPARꝩ ligand was 96.36% for 1 µM, 76.43% for 10 µM, and 50.33% for 20 µM. For BAY-0069 ligand cell viability was 117.46% for 1 µM, 86.19% for 10 µM, and 49.77% for 20 µM. Finally, the T0070907 ligand had a cell viability of 106.29% for 1 µM, 90.00% for 10 µM, and 48.09% for the 20 µM concentration (See Fig. 2B). The A549 tumor organoid’s cell viability for the GW9662 PPARꝩ ligand was 151.59% for 1 µM, 153.17% for 10 µM, and 118.97% for 20 µM. For BAY-0069 ligand cell viability was 89.18% for 1 µM, 52.65% for 10 µM, and 85.11% for 20 µM. Finally, the T0070907 ligand had a cell viability of 120.69% for 1 µM, 72.89% for 10 µM, and 92.34% for the 20 µM concentration (See Fig. 2C).

Figure 2. Lung cancer tumor organoids and the effects of PPARꝩ inhibitors. (p=0.05*, p=0.01**, p=0.001***, p=0.0001****

DISCUSSION.

In H1048 and H1299, our results demonstrated a significant reduction in tumor cell growth when cells were treated with increasing concentrations of PPARγ inhibitory ligands. These drugs failed to inhibit proliferation in A549 cells, however, with GW9662 and BAY-0069 increasing proliferation at a dose of 20 mm. Such data mimics the variable treatment responses that are often observed amongst lung cancer patients. As the models across experiments increased in complexity, the effectiveness of the drugs decreased. These results provided a partial confirmation of my hypothesis which stated that inhibition of the PPARγ pathway in 2D and 3D lung cancer models will result in a decrease in tumor cell viability. A notable dose-dependent decrease in tumor cell viability was observed in two out of three tested cell lines; however, 3D lung cancer models were significantly less sensitive to PPARγ inhibitory compounds. These findings highlight the importance of evaluating tumor cell drug sensitivity in a variety of microenvironments, not solely in 2D cell lines.

FUTURE WORKS.

Though our studies used three compounds that are known inhibitors of PPARγ activity, all three compounds are also capable of inhibiting other molecular targets that are independent of the PPARγ pathway.  To fully elucidate the role of PPARγ in our model systems, we will need to knock down or knock out PPARγ using either shRNA or a CRISPR-Cas9 system specific to PPARγ.  Also, to address our second hypothesis regarding PPARγ inhibition and reduced tumor invasion in the brain microenvironment, our future studies will examine the role of PPARγ in the cerebral organoid model that Dr. Linkous’ lab has previously described [14].

CONCLUSION.

From our findings, we can conclude that treatment with chemical inhibitors of PPARγ significantly reduced proliferation in two of three tested lung cancer cell lines.  However, depending on the cell line, drug, and conformational complexity of the samples, variable responses to PPARγ inhibition were observed.  For example, in the H1048 cancer models, the GW9662 inhibitor decreased cell viability for the 2D cell line and neurospheres (See S2) but increased cell viability for the most complex model (i.e., tumor organoids), although the increase was not statistically significant.  In addition, A549 cells exhibited both increased and decreased cell viability, depending on the 3D conformation of the cells and the PPARγ inhibitor that was tested.  This pattern of varying sensitivity based on cell model, drug, and drug concentration, illustrates the ongoing challenges that exist in the quest to develop novel, universally efficacious drugs for lung cancer.

ACKNOWLEDGEMENTS.

Thanks to the Research Experience for High School Students program and the Collaborative for STEM Education Outreach, Dr. Nicolas Means, Dr. Angela Eeds, and Dr. Nathaniel Freymeyer, for their mentorship and the opportunity to collaborate with the Linkous Lab.  Special thanks to Dr. Amanda Linkous, Jing Hao, and Mary Kate Macedonia, for mentoring me and allowing me to work alongside them.

SUPPORTING INFORMATION.

Supporting information includes imaging depicting neurosphere morphology and the culturing process as well as graphed results from testing on the neurosphere model.

REFERENCES.

  1. J. Fletcher, Metastatic lung cancer life expectancy: Survival rates and more (2021).
  2. D. Xiao, J. He, Epithelial mesenchymal transition and lung cancer. Journal of Thoracic Disease 2 (2010).
  3. J. D. Minna, J. A. Roth, A. F. Gazdar, Focus on lung cancer. Cancer Cell 1, 49–52 (2002).
  4. M. Raoof, Carcinogens (2023).
  5. CDC, Lung cancer among people who never smoked, Lung Cancer (2025).
  6. S. Penitzka, S. Steinvorth, S. Sehlleier, M. Fuß, M. Wannenmacher, F. Wenz, Erfassung kognitiver Funktionen nach prophylaktischer und therapeutischer Ganzhirnbestrahlung mittels neuropsychologischer Testverfahren. Strahlentherapie und Onkologie 178, 252–258 (2002).
  7. L. Helson, Radiation-induced demyelination and remyelination in the central nervous system: a literature review. Anticancer Res 38, 4999–5002 (2018).
  8. U. Anand, A. Dey, A. K. S. Chandel, R. Sanyal, A. Mishra, D. K. Pandey, Cancer chemotherapy and beyond: Current status, drug candidates, associated risks and progress in targeted therapeutics. Genes & Diseases 10, 1367–1401 (2023).
  9. T. A. Baudino, Targeted cancer therapy: the next generation of cancer treatment. Current Drug Discovery Technologies 12, 3–20 (2015).
  10. G. A. Van Norman, Limitations of animal studies for predicting toxicity in clinical trials. JACC Basic Transl Sci 4, 845–854 (2019).
  11. S. W. Han, Anticancer actions of PPARγ ligands: Current state and future perspectives in human lung cancer. WJBC 1, 31 (2010).
  12. M. Ahmadian, J. M. Suh, N. Hah, C. Liddle, A. R. Atkins, M. Downes, R. M. Evans, PPARγ signaling and metabolism: the good, the bad and the future. Nat Med 19, 10.1038/nm.3159 (2013).
  13. A. T. Reddy, S. P. Lakshmi, R. C. Reddy, Pparγ as a novel therapeutic target in lung cancer. PPAR Res 2016, 8972570 (2016).
  14. A. Linkous, D. Balamatsias, M. Snuderl, L. Edwards, K. Miyaguchi, T. Milner, Modeling patient-derived glioblastoma with cerebral organoids. Cell Reports 26, 3203-3211.e5 (2019).

Posted by on Friday, May 15, 2026 in May 2026.

Tags: , , ,