Periodic Reporting for period 1 - OVADEX (Defining the role of the tumor microenvironment in treatment resistance of OVArian cancer through patient-Derived EXplants (OVADEX))
Período documentado: 2022-09-01 hasta 2024-08-31
Treatment resistance in ovarian cancer (OC) is common. Recently, new treatment approaches with the potential to improve OC management have been identified, but their therapeutic efficacy is limited due to treatment resistance. In this regard, the tumor microenvironment (TME) is emerging as a key contributor to treatment resistance, being the metabolic crosstalk between TME and OC one of the recently proposed mechanisms of contribution. Evidence shows that platinum-resistant OC cells have defects in serine biosynthesis, leading to an enhanced serine uptake (Nat Commun 2022;13(1):4578). Importantly, exogenous serine is also required for some TME components, suggesting a possible competition that could affect their anti-cancer function. Unfortunately, the lack of knowledge on the exact functional role of TME in treatment resistance in OC is mostly because current preclinical models of OC do not fully recapitulate the TME.
In this context, OVADEX established four scientific objectives:
• To generate patient-derived explants (PDEs) from 50 treatment naïve OC patients and evaluate their response to different treatments, including standard chemotherapy and emerging therapies.
• To define metabolic changes, especially in serine biosynthesis, in the tumor before and after treatment.
• To define cellular and phenotypic changes in TME components before and after treatment.
• To associate changes in TME and OC cells to serine metabolism and treatment resistance.
In this context, OVADEX established four scientific objectives:
• To generate patient-derived explants (PDEs) from 50 treatment naïve OC patients and evaluate their response to different treatments, including standard chemotherapy and emerging therapies.
• To define metabolic changes, especially in serine biosynthesis, in the tumor before and after treatment.
• To define cellular and phenotypic changes in TME components before and after treatment.
• To associate changes in TME and OC cells to serine metabolism and treatment resistance.
Tumor resections were collected from 36 OC patients. Solid tumor lesions from different areas of the resected tissue were selected by an experienced pathologist and brought to the lab where I processed them by manually cutting into small tumor fragments of 1–2 mm3 size. Fragments were subsequently frozen according to Nat Med. 27(7) 2021:1250-1261, which allows for the cryopreservation of viable tissue. In each vial, we added 10 to 20 fragments to maintain tumor heterogeneity. We also created a RedCap project to store clinical and preclinical patient data, including demographic data, diagnosis, histology and translational evaluation of the tumor, treatment, follow-up, and survival. Clinical and preclinical data is continuously registered and updated in the database.
Next, I learned to establish and evaluate treatment responses in PDEs during a secondment. During this formative period, I was trained in (i) the culture technique, including thawing of the tissue, preparation of the culture matrix, establishment of cultures, ex vivo treatment, harvesting of fragments for analysis, (ii) read-outs, including sample processing for fluorescence-activated cell sorting (FACS), cytokine/chemokine analysis and scRNA-seq, and (iii) experimental design, especially about number of fragments required and tumor heterogeneity. Back in my lab, adaptation of the protocol to gynecological samples and non-immunologic therapies was required. The first optimization consisted of the identification of the culture conditions that lead to higher viability over time but preserving the original tumor microenvironment as much as possible. This included the usage of human plasma-like medium (HPLM), serum, and hormone supplementation when required. The second stage included the determination of appropriate drug concentrations and readout methods. In this regard, we set up two methods to evaluate tumor response: AlamarBlue and ToxiLight assays. Using AlamarBlue we measured the metabolic capacity in terms of reduction capacity over time, whereas the usage of ToxiLight allowed the determination of cell death. Treated and control (untreated) PDEs were collected and i) carboxymethylcellulose embedded for spatial metabolomics (WP3), ii) formalin-fixed and paraffin-embedded for IHC (WP3 and WP4) or iii) digested for FACS analysis.
During the project, we managed to create tissue microarrays from the cultured fragments to evaluate the evolution of the fragment composition and cellularity, tumor architecture, tissue morphology, and the expression of several markers. Preliminary data shows that we will be able to use these microarrays to measure PHGDH levels using IHC, but the optimization of the staining is not complete. To further investigate the metabolic changes associated with resistance and serine biosynthesis, we performed spatial metabolomics using mass spectrometry imaging (MSI). In contrast to bulk metabolites analysis by mass spectrometry, this platform allows the mapping of the spatial distribution of metabolites, drugs, and proteins from a tissue section at the single-cell level. To do so, we combined two state-of-the-art techniques: Matrix Assisted Laser Desorption Ionization (MALDI)-MSI and Desorption Electrospray Ionization (DESI)-MSI. Treated and untreated PDEs were embedded in optimal cutting temperature (OCT) compound. Cryosections were obtained following standard procedures using the Epredia™ Microm HM525 NX cryostat. For MALDI-MSI, we applied a matrix directly to cryosections, forming co-crystals with metabolites. Upon radiation with a laser beam, the matrix was ionized and charges were transferred to the metabolites, resulting in their desorption and ionization. Using this approach, we detected crucial metabolomics pathways, including serine, arginine, glutamine, nucleotides, free fatty acids, and many others. Complementary, we used DESI-MSI. DESI-MSI uses electrospray ionization, whereby a fine spray of charged solvent droplets extracts metabolites from tissue. Importantly, DESI-MSI imaging provided information on the spatial distribution of small metabolites (glucose, lactate, amino acids, etc.) without any additional sample preparation.
We characterize tumor samples collected at baseline, and after 3 days of culture to determine cell composition and viability. HE slides were generated from the tumor tissue reserved to be formalin-fixed and paraffin-embedded (FFPE), uncultured tumor fragments, and all treated and untreated PDEs. In addition, ki67 (cell proliferation) and Caspase-3 (apoptosis) were evaluated using standard IHC methods. To further evaluate cell composition, T cell state and activation, and the presence of myeloid, fibroblast, and endothelial cells, cultured and uncultured PDEs were digested into a single-cell suspension using a digestion mix (RPMI medium supplemented with 1% P/S, 1 mg/mL DNAse I, and 100 mg/mL collagenase type IV) for 1 h at 37 °C under slow rotation. Cells were analyzed using antibody panels that I set up in the hosting lab. Data was collected using BD FACSDiva 8.0.1 software and further analyzed with FlowJo v10.8.2 (FlowJo LLC). Unfortunately, the application of multiplex IHC using the MILAN platform and single-cell RNA sequencing has been delayed until sample collection is completed.
Finally, we also looked for associations between changes in the metabolism of cancer cells (WP3), TME changes (WP4), and patients’ responses, expecting to find differences in TME components and OC metabolism in the PDEs obtained from the resistant and refractory patients in comparison to the PDEs established from responders.
Next, I learned to establish and evaluate treatment responses in PDEs during a secondment. During this formative period, I was trained in (i) the culture technique, including thawing of the tissue, preparation of the culture matrix, establishment of cultures, ex vivo treatment, harvesting of fragments for analysis, (ii) read-outs, including sample processing for fluorescence-activated cell sorting (FACS), cytokine/chemokine analysis and scRNA-seq, and (iii) experimental design, especially about number of fragments required and tumor heterogeneity. Back in my lab, adaptation of the protocol to gynecological samples and non-immunologic therapies was required. The first optimization consisted of the identification of the culture conditions that lead to higher viability over time but preserving the original tumor microenvironment as much as possible. This included the usage of human plasma-like medium (HPLM), serum, and hormone supplementation when required. The second stage included the determination of appropriate drug concentrations and readout methods. In this regard, we set up two methods to evaluate tumor response: AlamarBlue and ToxiLight assays. Using AlamarBlue we measured the metabolic capacity in terms of reduction capacity over time, whereas the usage of ToxiLight allowed the determination of cell death. Treated and control (untreated) PDEs were collected and i) carboxymethylcellulose embedded for spatial metabolomics (WP3), ii) formalin-fixed and paraffin-embedded for IHC (WP3 and WP4) or iii) digested for FACS analysis.
During the project, we managed to create tissue microarrays from the cultured fragments to evaluate the evolution of the fragment composition and cellularity, tumor architecture, tissue morphology, and the expression of several markers. Preliminary data shows that we will be able to use these microarrays to measure PHGDH levels using IHC, but the optimization of the staining is not complete. To further investigate the metabolic changes associated with resistance and serine biosynthesis, we performed spatial metabolomics using mass spectrometry imaging (MSI). In contrast to bulk metabolites analysis by mass spectrometry, this platform allows the mapping of the spatial distribution of metabolites, drugs, and proteins from a tissue section at the single-cell level. To do so, we combined two state-of-the-art techniques: Matrix Assisted Laser Desorption Ionization (MALDI)-MSI and Desorption Electrospray Ionization (DESI)-MSI. Treated and untreated PDEs were embedded in optimal cutting temperature (OCT) compound. Cryosections were obtained following standard procedures using the Epredia™ Microm HM525 NX cryostat. For MALDI-MSI, we applied a matrix directly to cryosections, forming co-crystals with metabolites. Upon radiation with a laser beam, the matrix was ionized and charges were transferred to the metabolites, resulting in their desorption and ionization. Using this approach, we detected crucial metabolomics pathways, including serine, arginine, glutamine, nucleotides, free fatty acids, and many others. Complementary, we used DESI-MSI. DESI-MSI uses electrospray ionization, whereby a fine spray of charged solvent droplets extracts metabolites from tissue. Importantly, DESI-MSI imaging provided information on the spatial distribution of small metabolites (glucose, lactate, amino acids, etc.) without any additional sample preparation.
We characterize tumor samples collected at baseline, and after 3 days of culture to determine cell composition and viability. HE slides were generated from the tumor tissue reserved to be formalin-fixed and paraffin-embedded (FFPE), uncultured tumor fragments, and all treated and untreated PDEs. In addition, ki67 (cell proliferation) and Caspase-3 (apoptosis) were evaluated using standard IHC methods. To further evaluate cell composition, T cell state and activation, and the presence of myeloid, fibroblast, and endothelial cells, cultured and uncultured PDEs were digested into a single-cell suspension using a digestion mix (RPMI medium supplemented with 1% P/S, 1 mg/mL DNAse I, and 100 mg/mL collagenase type IV) for 1 h at 37 °C under slow rotation. Cells were analyzed using antibody panels that I set up in the hosting lab. Data was collected using BD FACSDiva 8.0.1 software and further analyzed with FlowJo v10.8.2 (FlowJo LLC). Unfortunately, the application of multiplex IHC using the MILAN platform and single-cell RNA sequencing has been delayed until sample collection is completed.
Finally, we also looked for associations between changes in the metabolism of cancer cells (WP3), TME changes (WP4), and patients’ responses, expecting to find differences in TME components and OC metabolism in the PDEs obtained from the resistant and refractory patients in comparison to the PDEs established from responders.
In OVADEX, we effectively developed and optimized a PDE culture system that preserves the original TME of the tumor, allowing us to investigate its role in OC chemotherapy response. In this new PDE platform, we dissected fresh tumor tissues from surgical resections into small fragments (~1 mm3), intending to allow sufficient nutrient and reagent access while preserving the cellular context and tumor architecture. Furthermore, the use of small-sized tumor fragments allowed the mixing of PDEs from different tumor areas, thereby reflecting tumor heterogeneity. The optimization of the culture system included the identification of the culture conditions that lead to higher viability over time, including media, serum, and hormone supplementation. In addition, to prevent immune cell efflux, PDEs were embedded into an artificial extracellular matrix. By comparing PDE cultures with and without the matrix, we confirmed that embedding in the matrix did not limit immune cell functionality or induce unspecific immune activation.
To assess the stability of the tissue during culture and the tumor response to treatment, we have analyzed immune, nonimmune, and cancer cell compartments with different techniques at different time points. Using fluorescence-activated cell sorting (FACS), we have determined the immune composition, T cell state and activation, and the presence of myeloid, fibroblast, and endothelial cells. This is important to know if a lesion from a patient has been infiltrated by immune cells and to evaluate the response of the immune compartment to specific treatments. Importantly, we have observed stable levels of the cellular compartments up to 3 days of culture, indicating that this system allows for the analysis of early effects of ex vivo treatment. To evaluate tumor response to treatment, we used the AlamarBlue and ToxiLight assays. Using these approaches, we have measured cell damage and the evolution of the reduction capacity of different fragments derived from the same tumor over time. We noticed that intratumor heterogeneity is reflected in the fragments, with noticeable differences between fragments in the fall of their reduction capacity over time. In addition, the decrease in the reduction capacity is not only fragment-dependent but tumor-dependent. Different tumors show different evolutions in their reduction capacity over time reflecting intertumor heterogeneity. Using these approaches, we have assessed the tumor response to carboplatin considering different concentrations and exposures. Importantly, fragments derived from patients with refractory disease that progressed during treatment displayed no differences in the evolution of the reduction capacity or cell damage between the treated and the untreated fragments. On the contrary, fragments derived from potential responders to carboplatin exhibit a clear increase in cell damage and a decrease in the reduction capacity after treatment compared to untreated fragments. In addition to the previous readouts, we have also managed to construct tissue microarrays from the cultured fragments used to evaluate the evolution of the fragment composition and cellularity, tumor architecture, tissue morphology, and metabolism after carboplatin exposure over time. Moreover, we have optimized the algorithm to measure the expression of different markers in the immune and tumor compartments, being able to track their evolution in fragments processed at different time points.
The development of this model is an important achievement since ex vivo models need to be predictive of clinical response to allow researchers to directly test the functional importance of specific TME components to treatment response. Several ex vivo cancer models have been previously devised to investigate TME function in response to therapy. Unfortunately, whether these models predict response to chemotherapy (whether the response ex vivo correlates to the response in the patient) was not established. Our results highlight the utility of PDEs i) to study the role of TMEs in the context of treatment response in OC and ii) to be used as preclinical models to predict response in OC tumors.
To assess the stability of the tissue during culture and the tumor response to treatment, we have analyzed immune, nonimmune, and cancer cell compartments with different techniques at different time points. Using fluorescence-activated cell sorting (FACS), we have determined the immune composition, T cell state and activation, and the presence of myeloid, fibroblast, and endothelial cells. This is important to know if a lesion from a patient has been infiltrated by immune cells and to evaluate the response of the immune compartment to specific treatments. Importantly, we have observed stable levels of the cellular compartments up to 3 days of culture, indicating that this system allows for the analysis of early effects of ex vivo treatment. To evaluate tumor response to treatment, we used the AlamarBlue and ToxiLight assays. Using these approaches, we have measured cell damage and the evolution of the reduction capacity of different fragments derived from the same tumor over time. We noticed that intratumor heterogeneity is reflected in the fragments, with noticeable differences between fragments in the fall of their reduction capacity over time. In addition, the decrease in the reduction capacity is not only fragment-dependent but tumor-dependent. Different tumors show different evolutions in their reduction capacity over time reflecting intertumor heterogeneity. Using these approaches, we have assessed the tumor response to carboplatin considering different concentrations and exposures. Importantly, fragments derived from patients with refractory disease that progressed during treatment displayed no differences in the evolution of the reduction capacity or cell damage between the treated and the untreated fragments. On the contrary, fragments derived from potential responders to carboplatin exhibit a clear increase in cell damage and a decrease in the reduction capacity after treatment compared to untreated fragments. In addition to the previous readouts, we have also managed to construct tissue microarrays from the cultured fragments used to evaluate the evolution of the fragment composition and cellularity, tumor architecture, tissue morphology, and metabolism after carboplatin exposure over time. Moreover, we have optimized the algorithm to measure the expression of different markers in the immune and tumor compartments, being able to track their evolution in fragments processed at different time points.
The development of this model is an important achievement since ex vivo models need to be predictive of clinical response to allow researchers to directly test the functional importance of specific TME components to treatment response. Several ex vivo cancer models have been previously devised to investigate TME function in response to therapy. Unfortunately, whether these models predict response to chemotherapy (whether the response ex vivo correlates to the response in the patient) was not established. Our results highlight the utility of PDEs i) to study the role of TMEs in the context of treatment response in OC and ii) to be used as preclinical models to predict response in OC tumors.