November 2014, Volume 3 Issue 3
Metabolic Profiling of Slowly Cycling Cells Reveals Differences in Glucose, Lactate, and Choline Concentrations in Melanoma Cells
Ishaan Parikh1*, Ms. Angelique Bosse2, and Sergey Magnitsky3
Student1, Teacher2: Montgomery Blair High School, 51 University Blvd. E., Silver Spring, MD 20901
Mentor: University of California, San Francisco, 185 Berry Street, San Francisco, CA 94107
*Corresponding author: firstname.lastname@example.org
A major problem with creating cancer therapies is the heterogeneity of the tumors. In melanoma, slowly cycling cells (SCC) comprise only 1-5% of the entire tumor cells and have demonstrated substantial drug resistance. These cells divide asymmetrically, produce a highly proliferating progeny, and exhibit a resistance to all conventional cancer drugs. Previous animal experiments showed that the selective elimination of the SCC subpopulation leads to tumor disappearance. Cells with similar properties have been identified in ovarian, brain, and breast cancers. Metabolic characterization of slowly cycling melanoma cells is the first step to understanding differences of SCC from bulk tumor cells and identifying roles of these cells in carcinogenesis. Further investigation of SCCs may allow oncologists to understand their characteristics and thus develop new therapeutic strategies. In the study, SCCs were separated from bulk tumor cells and the two subpopulations were compared in terms of glucose uptake and lactate and choline production. These measurements were taken because of each molecule’s relevance in either cancer detection or therapy. For the glucose uptake, a radio-tagged isotope of glucose was utilized and the amounts of radioactivity accumulated in different subpopulations were compared. To measure metabolite concentrations, cell extracts were prepared, analyzed with NMR spectroscopy, and compared the subpopulations of cells. A higher glucose uptake was observed in the slowly cycling subpopulation than in the bulk tumor. The bulk tumor’s uptake was extremely close to that of the unsorted tumor, which was consistent with the hypothesis, as the tumor is primarily the bulk tumor cells. Regarding lactate production, the SCCs produced more than twice the amount of lactate than the bulk tumor; while in choline concentration, the bulk tumor had the higher value. In melanoma, it is clear that the SCCs are integral to the survival of the tumor, and the subpopulation should be a treatment target. By learning more about SCCs, it brings us closer to understanding the heterogeneity of tumors and design better targeted treatments and for cancer and possibly a cure. It is important that the scientific community investigates the significance of the distinct subpopulations within cancers, and locally treats them. This research is thus the foundation for a new paradigm in designing cancer therapies.
Melanoma is an aggressive cancer that has a steadily rising incidence worldwide1. Melanoma can be stopped only if detected at an early stage; otherwise, the survival rate of affected patients falls just under nine months2. It can begin in any anatomical region containing melanocytes and is easily identified because of the abundance of melanin in affected areas. However, despite its relatively easy diagnosis and various treatment mechanisms, there exists no cure for melanoma3. The disease has been recognized for its unique intra-tumoral heterogeneity4,5, and recent studies indicate that there are several distinct subpopulations in melanoma tumors. One of these subpopulations consists of the slowly cycling melanoma cells (SCCs)6,7.
The Histone 3 K4 demethylase JARID1B was identified as a biomarker for the SCC subpopulation6. JARID1B is involved in chromatin regulation and has been associated with stem cell maintenance8-10. JARID1B is present in all tissues in the human body, but has a higher expression in those cells that are dividing more frequently or have regenerative properties11. In melanoma, however, JARID1B seems to be present at abnormally high concentrations in a few, singular cells. These few cells are what are known as slowly cycling cells6.
Based on these findings, one would expect a lower concentration of JARID1B in melanoma cells to be associated with a higher rate of division. 2-4(4-methylphenyl)-1,2-benzisothiazol-3(2H)-one (PBIT) was discovered as an inhibitor for the JARID1B demethylase. Researchers incubated the inhibitor in cells with a high concentration of JARID1B and, contrary to the expectation, observed total cell death11. This suggests that the inhibitor may have some potential for treating malignant melanoma.
Slowly Cycling Cells
The slowly cycling cell subpopulation is a small portion of the tumor consisting of about 1 to 5% of the total tumor’s mass (Figure 1a). These cells can divide asymmetrically, producing one bulk tumor cell (JARID1B negative) and one slowly cycling cell (JARID1B positive) during each replication (Figure 1b). The bulk tumor cells then proceed to replicate quickly and produce other JARID1B negative cells, while the slowly cycling cells repeat their division every >4 weeks12.
Figure 1: (a) Cell Sorting Population Distribution (Image obtained from (6)). The graph shows intensity of Green Fluorescent Protein expression vs. number of cells (The cell line used was a the WM3734-JARID1Bprom-EGFP cell line. This cell line co-expresses JARID1B and GFP). This is a bimodal curve; the first represents the bulk tumor cells, while the second represents the slowly cycling cells that have the distinct extremely high expression6. (b) Microscopic Images of Slowly Cycling Cell Division (Image obtained from: (6)). This group of cells was originally only JARID1B+, or slowly cycling, cells (green). However, this image was taken 15 days after a single slowly cycling cell was isolating, indicating that the slowly cycling cell produces both slowly cycling and bulk tumor (grey) cells6.
The slowly cycling cells are also able to suddenly “switch” into a rapid-proliferative state as a result of asymmetric cells division. An isolated subpopulation of slowly cycling cells divides just as fast as a subpopulation of bulk tumor cells for about eleven days; then, the slowly cycling cells begin to double at a far greater rate than before (Figure 2). This indicates a strong role in tumor recurrence following therapy. The bulk tumor cells that are produced from the now fast-dividing slowly cycling subpopulation not only naturally double extremely quickly, but also are more resistant to subsequent therapies13.
Figure 2: Fold Growth for Isolated GFP+ and GFP- Subpopulations (image obtained from (6)). The dark line represents the isolated SCC subpopulation, while the gray represents the bulk tumor. The dark line grows at a must faster rate after eleven days, which is the point when the SCCs switched into a rapidly proliferative phase 6.
Chemotherapy and radiation, among other common cancer treatments, work by mutating the cancer cell’s DNA during replication when the DNA is exposed. If a mutation is lethal (successful, in this case), the cell induces apoptosis, or cell death, instead. Previously, researchers were under the impression that since every cell in the tumor was dividing, all cells would be killed by the treatment. However, the slowly cycling subpopulation is rarely dividing, so the cancer drug can barely even interact with the cells9. Therefore, the only subpopulation that is really hindered by therapies such as chemotherapy is the bulk tumor14. The extremely long doubling time of slowly cycling cells gives this population a form of intrinsic resistance to most modern cancer therapies4. Previous experiments found that isolated JARID1B-positive subpopulations are more tumorigenic than JARID1B-negative subpopulations (Figure 3). This suggests a promising potential for the SCCs as future therapeutic targets.
Figure 3: Tumor Emergence from Isolated Subpopulations (image obtained from (6)). The dish on the left originally contained only SCCs, while the dish on the right contained only bulk tumor cells. Each of the red marks is considered a tumor, and this shows the extreme tumorigenic property that the SCC subpopulation has 6.
One of the most important properties of the SCC subpopulation is the essential role it plays in tumor maintenance. In a previous study, researchers removed the entire SCC subpopulation from a tumor, and then let it grow. Initially, an increase in rate of growth of the tumor was detected compared to the rate of growth of a normal tumor, but the experimental tumor ultimately ceased to thrive6. Essentially, this proved that in the absence of the JARID1B-positive cells, a melanoma tumor is unable to survive.
In order to conduct the aforementioned experiment, researchers at the Wistar Institute engineered the WM3734 human melanoma cell line in order to counteract the dynamic JARID1B phenotype that is normally present15. It is extremely important to note this dynamic phenotype when comparing the SCC subpopulation to a traditional “Cancer Stem Cell” (CSC). CSCs are fairly new in the scientific community and they suggest an entire new mode of tumor metastasis. These cells asymmetrically divide and can differentiate into various types of cells that compose the rest of the tumor. These CSCs specialize in self-renewal and despite being such a small subpopulation, they are the source for the majority of the tumor16. The hypothetical presence of these cells in cancer would create an entire new target for therapies7. The SCC fulfills all of the criteria to be a CSC, except for hierarchy organization and the dynamic changes of phenotype of slowly cycling cells6.
Cellular respiration may be aerobic or anaerobic. Both pathways initially utilize glucose and go through glycolysis, to produce 2 molecules of ATP. The resulting two pyruvate molecules either go through the TCA cycle or lactic acid production in aerobic or anaerobic respiration, respectively. If the TCA cycle and further oxidative phosphorylation are utilized, up to 38 molecules of ATP per glucose molecule can be produced. However, even in the presence of oxygen, it is believed that the cancer cells use glycolysis excessively3,17-19. In cancer, large amounts of lactate can be detected in affected tissue and this has been suggested as a biomarker for cancer20. In the 1920s, Otto Warburg suggested that certain components in TCA or the electron transport chain in cancer cells’ mitochondria could be deficient, hence resulting in the elevated lactate level, but no components have since confirmed the source of the excess lactate21. For our study, we hypothesize that slowly cycling cells are responsible for the high lactate concentration. It is likely that since the SCC subpopulation is so small, no researcher has ever examined these cells individually.
Positron Emission Tomography Gamma Counters
In positron emission tomography (PET) scans for humans, a radioactive isotope is injected into the body. The isotope emits a positron, and this positron interacts with an electron from tissue. This reaction leads to an emission of two high-energy gamma ray packets, or gamma quants. Gamma quants can be detected and an image can be reconstructed. For example, in standard nuclear medicine, a high concentration of a radioactively (F-18) labeled isotope of glucose, fludeoxyglucose (FDG), could suggest a tumor in that area. In order to measure glucose uptake in cells, the same concept could be utilized. FDG is an isotope-labeled analog of glucose that will still be taken up by cells, but cannot go through glycolysis. In simple terms, it gets “stuck”, and then the scanner can read the positron emissions. By incubating FDG with a cell suspension and measuring the amount of gamma quants, one could measure how much radioactivity is present in the sample. If the cell count is known, this could be converted into how much glucose the cell is utilizing.
Nuclear Magnetic Resonance Spectroscopy
Nuclear Magnetic Resonance (NMR) spectroscopy is one of the best methods of studying metabolite concentrations both in vivo and in vitro. When samples are placed inside of the NMR spectrometer, most of the protons’ nuclei align with the magnetic field created by the machine. The magnet releases an electromagnetic pulse to align the nuclei with the opposite pole, and then measures the fluctuation in the magnetic field as the nuclei revert back to the relaxed (alpha) state. The signal obtained during the shift is emitted at a certain frequency that is dependent on electron shielding. As a result, information about the nucleus’ chemical environment can be derived from its resonance frequency. Through experimentation and calibration, the scientific community has figured out at which points certain metabolites are, and with this knowledge we can determine the concentrations of individual metabolites.
The primary goal of this project was to characterize the slowly cycling cells in a melanoma cell line using the metabolic profiling techniques and methods above.
Materials and Methods
Tu2% Media: 1) 400 mL of dH2O with 50 mg of MCDB153. 2) 100 mL of Leibovitz’s L-15. 3) 10 mL of Fetal Bovine Serum. 4) 0.5 mL of Bovine Insulin. 5) 0.42 mL of Calcium Chloride. The aforementioned ingredients were mixed and filter sterilized using a 0.22 mm filter.
LV15 + FBS 10% Media: 1) 500 mL of Leibovitz’s L-15. 2) 55.6 mL of Fetal Bovine Serum. The aforementioned ingredients were mixed and filter sterilized using a 0.22 mm filter.
The cryovial containing the WM3734 human melanoma cell line was removed from liquid nitrogen storage Dewar, and transferred to an ice bath. The cell line we used coexpressed the JARID1B biomarker with the green fluorescent protein (GFP). After five minutes, the cryovial was placed in a cold-water bath, and subsequently in a warm water bath. Using the LV15 + FBS 10% media, the thawing cells were pipetted out of the cryovial into a 15 mL centrifuge tube. After centrifugation, the WM3734 cells were aspirated, re-suspended in 5mL of the same media, and centrifuged one more time. The cells were aspirated and mixed with 3 mL of Tu2% media. The solution was transferred to a T25 flask for growth and placed in a 37° C incubator.
After the T25 flask became confluent, all of the Tu2% media was removed. Saline was added in order to wash off excess serum and quickly removed 30 seconds after. To detach the cells from the surface, 0.5 mL of trypsin (0.05%) was added, and the flask was incubated for 3 minutes. Following incubation, the trypsin was neutralized by 2.5 mL of LV15 + FBS 10%, and then the entire suspension was pipetted into a 15 mL centrifuge tube. After centrifugation, the cells were aspirated and then suspended in 5 mL of LV15 + FBS 10%. The cells were aspirated after another centrifugation, and then resuspended in 13 mL of Tu2%. The entire solution was then transferred to a T75 flask.
After the cells reached confluence in the T75 flask, they were removed and washed in the same manner as before. However, 1.5 mL of trypsin was used and 6.5 mL of the final Tu2% media was placed into one T150 flask and the other half was placed into another T150 flask. Another 13.5 mL of Tu2% was added to each flask so that the surfaces were fully covered with the media. This procedure was repeated until we had 6 confluent T150 flasks with the WM3734 cell line. In addition, one T25 flask of WM3734 cells was created in the same manner as previously discussed. Figure 4 shows the cells in vitro during attachment.
Figure 4. Attached WM3734 GFP+ Cells. This is a microscopic (light microscope) image of the melanoma cells after they attach to the flask. They maintain a linear shape during this time (they have a circular morphology when detached). The few circular figures are likely dead cells.
Cell Sorting Preparations and Execution
Four of the confluent T150 WM3734 GFP+ flasks and one T25 WM3734 GFP- flasks were removed from the incubator. Using the same method of removing the cells as mentioned in step 2, each of the cell lines were trypsinized and washed twice. Following the second centrifugation, the GFP+ cells were suspended in 1 mL of saline. The GFP- cells were suspended in 1.5 mL of saline, and 0.5 of the cell solution was placed in a T25 flask already containing 2.5 mL of Tu2% (in order to have a constant supply of the GFP- cells). The remaining 1 mL solution of GFP- cells, 1mL solution of GFP+ cells, and a 10 mL aliquot of Tu2% were placed on ice and taken to the cell sorting facility.
The cell-sorting machine separates the cells based on their fluorescence when different lasers pass through single-cell droplets. Depending on if there was a high expression or not, an electrical plate will move the droplet so that it will fall into a separate tube with cells of its kind (Figure 5a). After obtaining cells with high GFP expression (High GFP), medium GFP expression (Mid GFP), and low expression of GFP (Low GFP) (Figure 5b), they were centrifuged, and re-suspended in 13 mL of Tu2% and placed in a T75 flask for 3 days for recuperation.
Figure 5(a). Cell Sorting Mechanism. All of the GFP+ cells start out in the same solution, but then proceeded down a cone-like shaft so that single cells were isolated. A series of lasers passed through the cell and into two detectors; one for size and one for fluorescence. If the size detector only detects one cell, the fluorescence detector signals the two electromagnetic plates to change their polarities such that the cell can fall into the similar GFP expression cell containers.
Figure 5(b). GFP Expression Curve. (Image obtained from UCSF Cell Sorting Facility). The graph shows the distribution of GFP intensity among the cells in the sample given to the machine. The section labeled “High” would contain the highest 10% of GFP expression and so the highest concentration of slowly cycling cells. The section labeled “Low” is the bottom 10% of GFP expression, and hence the bulk tumor.
Fludeoxyglucose (FDG) Uptake: Preparations and Cell Counting
The following flasks were used for the FDG uptake experiment: T75 of Low GFP, Mid GFP, and High GFP and a T150 of GFP+ Unsorted cells. All of the cells were trypsinized and removed from their flasks. Following the first centrifugation, all four cell groups were aspirated and re-suspended in 5 mL of saline. Twenty µL from each solution was removed and mixed with 20 µL of Trypan Blue (a vital stain for counting live cells). The number of cells was counted using a hemacytometer. The 3 mL solutions were all centrifuged, aspirated, and then suspended in 1 mL of saline in an eppendorf tube.
FDG Uptake Experiment
About 1.5 mC FDG was obtained and stored in a lead container for safety. Depending on the sample volume, FDG with 10 µC of radioactivity was added to the cell suspension and incubated at room temperature for 30 minutes. Following incubation, the supernatants were extracted post-centrifugation, and the cells were re-suspended in 1 mL of saline. Finally, all eight samples (four cell solutions and four supernatants) were placed into gamma counter for measurement of radioactivity.
Cell Extract Preparations and Layer Separation
Metabolites from different sub-cell populations were prepared as described previously 22. The cells were centrifuged following gamma counting, and then suspended in 4 mL of ice-cold methanol and transferred to a glass tube. After extensive vortexing, 4 mL of ice-cold chloroform was added and the solution was vortexed again. For the final step, 4 mL of ice-cold water was added and the entire solution vortexed once more. The samples were placed at 4° C overnight so they could be separated into different layers the following day. The solution was separated into a lipid, protein, and aqueous-metabolites fractions and placed in an -80° C freezer.
Lyophilization and Nuclear Magnetic Resonance Insert Preparations
The separated layers were removed from the -80° C freezer and placed directly into a glass jar attached to the lyophilizer, which sublimed the ice so only the metabolites remain in powder form.
Twenty µM trimethylsilyl-2,2,3,3-tetradeuteropropionic (TSP) D2O solution was used as a standard during NMR spectroscopy, which was prepared by dissolving 68.9 mg of TSP in 100 mL of dH20. Then 100 µL of this solution was mixed with 20 mL of D2O. The metabolites in powder form were re-suspended in 400 µL of D2O in each of the centrifuge tubes and transferred to 5mm diameter NMR tube inserts to measure metabolite concentrations.
Preforming NMR and Analysis of Spectra
NMR spectroscopy has been used to measure metabolite concentrations in tumor and brain tissues in 23 and 24. The Bruker 600 Hz Nuclear Magnetic Resonance Spectrometer was used in order to obtain metabolite concentration. The machine was set to proton NMR frequencies, and immediately calibrated to suppress excess water signal. Signal was collected over a two hour time period to maximize the signal-to-nose ratio.
The software “MustReNova” was used to analyze NMR spectroscopy data. MustReNova allows for simple normalization and baseline correction of the spectra. In combination with a library provided by “Chenomx” containing the ppm values of metabolites, specific peaks were detected on the spectra that correlate to the metabolites and the lactate and choline concentrations were measured.
Each experiment involved four cell suspensions: high GFP (slowly cycling), medium GFP, low GFP (bulk tumor), and unsorted GFP (entire tumor). The software “RiaCalc WIZ” returned how much radioactivity was in each suspension, and then the FDG uptake was calculated. We found that the high GFP, or slowly cycling, subpopulation had a mean uptake of 3.291 FDG/min which was higher than the low GFP, or bulk tumor, mean uptake of 1.824 FDG/min. The medium GFP subpopulation’s uptake was much closer to the high, but the unsorted’s was almost the same as that of the low (Figure 6).
To obtain metabolite concentrations, D2O samples were run in the Bruker 600 MHz Nuclear Magnetic Resonance spectrometer (Figure 7). Lactate peaks were observed at 1.3 and 4.1 ppm. At 1.3 ppm, there was a doublet, and at 4.1 ppm a quadruplet. After adding the areas under all of these peaks, and dividing by the area under our standard’s, TSP’s, peak (at ppm=0), we obtained the ratio of lactate to TSP in our D2O solution. By dividing by the total cell count that was in this suspension originally, we determined the concentration of lactate per cell. The choline peak is typically a small singlet located at 3.2 ppm.
Figure 7. Screenshot of MestReNova Software with Adjusted NMR Spectrum. This is an image of the types of spectrum that we analyzed in order to get the area under the desired peaks. The orange boxes show the general location of the lactate peaks, while the black box shows where the choline peak was located. The green box contains the distinct TSP signal that was integral to calculating the final concentrations of the metabolites. The large blue box contains the water signal post-suppression. Prior to running the metabolites accumulations, we suppressed water signal for as much as the machine would allow us to do, and yet the signal was typically very large. Metabolites from about 4.5 ppm to 6.5 ppm were generally undetectable because of the immense water signal.
Due to time constraints, the sample size of our data was limited. However, we were able to obtain four data points for the lactate concentration per cell, and we were able to detect a statistically significant difference even with the miniscule number of trials (Figure 8). We were unable to find a standard value for the volume of a typical melanoma cell, and so moles per cell were determined instead. Based on our results, the SCCs produce significantly more lactate than the bulk tumor. The SCCs produce about 2.98 x 10-9mMoles of lactate, while the bulk tumor produces 1.32 x 10-9 mMoles. There was a general upward trend, but the Medium GFP amount was significantly closer to that of the bulk tumor than the others.
Figure 8: Lactate Production in Different Subpopulations. The high GFP cells, or SCCs, are producing a significant amount of lactate when compared to the bulk tumor subpopulation (p=0.0496). The Mid GFP was significantly closer to the bulk tumor, which suggests that the distinctly high JARID1B expression could result in the excess lactate that is being produced in the SCCs.
In recent studies, a higher choline uptake has been detected in various cancers25,26. We wanted to determine if there was a difference in choline concentration with the different subpopulations in melanoma. After measuring choline singlet at 3.2 ppm, we found that the SCCs had an average choline amount of 4.19 x 10-10 mMoles, which was much lower than that of the bulk tumor at 6.68 x 10-10 mMoles (Figure 9).
Figure 9: Choline Production in Different Subpopulations. The bulk tumor had a much higher concentration of choline than the SCCs (p=0.008). Even though the Mid GFP and the SCCs had a lower concentration, this data suggests that 95% of a tumor could still have a higher concentration of choline, which corroborates previous studies that tumors have a high overall choline level. The small SCC subpopulation would not have a large effect on the reading from a PET scan detecting choline levels.
To create an effective drug against heterogeneous tumors, it is important to understand differences between subpopulations of tumor cells. In our experiments, we tried to identify metabolic differences between different subpopulations of melanoma cells. We analyzed FDG uptake and lactate and choline production in SCCs and compared amounts to those of the bulk tumor cells. Our results showed that the SCCs produced significantly more lactate compared to the fast proliferation subpopulation (bulk tumor cells). Similar results were detected in the FDG experiment: SCCs uptake a higher amount of glucose compared to bulk tumor cells.
An unusually high glucose uptake is often a biomarker for cancer27. Our data corroborates this, as our lowest uptake (bulk tumor) with the cancer cells was almost twice the uptake of normal cells. In addition, the SCCs more than tripled the normal cell’s uptake, which shows an even greater irregularity in the energy pathways of the SCCs.
We expected a lower uptake in the SCCs when compared to that of the bulk tumor because the SCCs are not actually dividing. There was no apparent reason for the SCCs to be utilizing so much glucose when cellular division was not taking place. It raises the question as to what is the need for the excess glucose, and there is a large input into glycolysis, but we do not know what happens from there.
In the 1970s, Otto Warburg won the Nobel Prize for discovering that there was a large lactate concentration in affected regions in cancer. He hypothesized that this was because of an error in the TCA cycle and oxidative phosphorylation in the energy pathways, or in the mitochondria of cancer cells. However, intensive investigations of the Kreb’s cycle’s components and mitochondrial enzymes of cancer cells revealed very minor alterations28. We hypothesize that these changes were not detected because they were present only in a small subpopulation of the entire tumor, namely SCCs. Our results show that the SCCs are clearly producing far more lactate than the bulk tumor, and so it could be that in fact the SCCs that are producing the majority of the tumor’s lactate. It is possible that no one has examined individual subpopulations like SCCs in the past.
To draw further conclusions, one would need to measure the ATP concentrations in each of the subpopulations. If a lower ATP concentration is detected in the SCCs than in the bulk tumor, we could confirm that these cells are primarily utilizing the energy production pathway of lactic acid production. However, if a higher ATP concentration were detected, further analysis would be required of other metabolites within the SCCs.
In the future, we plan to study the catalytic activity of the Kreb’s Cycle and mitochondrial activity of SCCs in order to test our hypothesis. This could be the first step to developing localized treatments for melanoma. The specific target of the SCC subpopulation in combination with traditional cancer therapies could result in an overall more effective treatment.
Choline has recently been identified as a biomarker for cancers in the breast and prostate. The presence of unusually high concentrations of choline in cancerous tissue can be detected using a PET scanner. Our findings show that the bulk tumor does have a high concentration of choline, while the SCCs do not. This is consistent with what is already known, as the weighted average of all the cells’ concentrations would still be much closer to the bulk tumor’s concentration, which is relatively high 25,26.
Our results by themselves are just the beginning when it comes to understanding the SCC subpopulation in melanoma. These are just a few of the properties that the medical community looks at when analyzing any cell. However, even with just a few results, one can see the disparity between SCCs and the bulk tumor. It is crucial that investigation into this subpopulation continues, as the potential of these cells as possible treatment targets later is immense.
Conclusion and Future Work
The SCC subpopulation in malignant melanoma is clearly an important subpopulation. Previous research has already asserted their necessity in tumor maintenance and their “stem cell”-like properties such as asymmetric division do suggest a larger role in tumor recurrence 6. Investigating the glucose uptake and metabolite concentrations was the first step in understanding the energy production mechanism in these cells.
Our experiment confirms that the energy pathways in the SCCs are different from those in the bulk tumor cells. In combination with previous knowledge, it suggests that these cells should be therapy targets in the future. More research needs to be done to sufficiently metabolically profile these cells. Lactate and choline were only the first step; measurements of concentrations of other metabolites will bring scientists closer to understanding why these cells act the way they do and why and how they play such an important role in melanoma development.
A melanoma tumor needs the SCC subpopulation for survival. By creating an environment that is toxic to the SCCs, scientists could ensure no SCC presence at all. If the JARID1B+ cells are inhibited with specific molecules such as 2-4(4-methylphenyl)-1,2-benzisothiazol-3(2H)-one (PBIT) 11 to deplete the SCCs and regular chemotherapy is used for the treatment of the bulk tumor, more effective methods of eliminating the cancer can be developed. Each subpopulation would be locally attacked and produce a better overall reduction of the tumor. Further research needs to be conducted to observe the behavior of SCCs in the JARID1B inhibitors environment, as this could allow scientists to develop treatments that target a smaller population, but have better overall effects.
In general, subpopulations with “cancer stem cell” properties have been detected in various cancers. It is of utmost importance that the medical community analyzes these different subpopulations of cells like the SCCs that are so integral to tumor maintenance and function are detected. These cells are next targets for cancer therapies.
These cells will likely bring about a paradigm shift to the entire way cancers and cancer treatments are thought of. The presence of different subpopulations of cells like SCCs in tumors not only heightens the complexity of the tumor, but also expands the possible approaches of treatments. If the one or two integral subpopulations that are key in tumor maintenance are identified and understood per cancer, it is not before long that new, successful, and efficient therapies are produced.
1. Markovic SN, Erickson LA, Rao RD, et al. Malignant melanoma in the 21st century, part 1: epidemiology, risk factors, screening, prevention, and diagnosis. Mayo Clinic proceedings. Mar 2007;82(3):364-380.
2. Gogas HJ, Kirkwood, J.M., and Sondak, V.K. Chemotherapy for metastatic melanoma: time for a change? Cancer. 2007(109):445-464.
3. Tsao H, Chin L, Garraway L, Fisher D. Melanoma: from mutations to medicine. Vol 26. Genes & Development: Cold Spring Harbor Laboratory Press; 2012:1131-1155.
4. Roesch A, Vultur A, Bogeski I, et al. Overcoming intrinsic multidrug resistance in melanoma by blocking the mitochondrial respiratory chain of slow-cycling JARID1B(high) cells. Cancer Cell. Jun 10 2013;23(6):811-825.
5. Zabierowski SE, Herlyn M. Learning the ABCs of melanoma-initiating cells. Cancer Cell. Mar 2008;13(3):185-187.
6. Roesch A, Fukunaga-Kalabis M, Schmidt EC, et al. A temporarily distinct subpopulation of slow-cycling melanoma cells is required for continuous tumor growth. Cell. May 14 2010;141(4):583-594.
7. Fukunaga-Kalabis M, Roesch A, Herlyn M. From cancer stem cells to tumor maintenance in melanoma. J Invest Dermatol. Aug 2011;131(8):1600-1604.
8. Takeuchi T, Watanabe Y, Takano-Shimizu T, Kondo S. Roles of jumonji and jumonji family genes in chromatin regulation and development. Developmental dynamics : an official publication of the American Association of Anatomists. Sep 2006;235(9):2449-2459.
9.Wilsker D, Probst L, Wain HM, Maltais L, Tucker PW, Moran E. Nomenclature of the ARID family of DNA-binding proteins. Genomics. 2005;86(2):242-251.
10. Lee TI, Jenner RG, Boyer LA, et al. Control of developmental regulators by Polycomb in human embryonic stem cells.[see comment]. Cell. 2006;125(2):301-313.
11. Sayegh J, Cao J, Zou MR, et al. Identification of small molecule inhibitors of Jumonji AT-rich interactive domain 1B (JARID1B) histone demethylase by a sensitive high throughput screen. J Biol Chem. Mar 2013;288(13):9408-9417.
12. Sigalotti L, Covre A, Nicolay HJ, Coral S, Maio M. Cancer testis antigens and melanoma stem cells: new promises for therapeutic intervention. Cancer Immunol Immunother. Mar 2010;59(3):487-488.
13.Zabierowski SE, Herlyn M. Melanoma stem cells: the dark seed of melanoma. J Clin Oncol. Jun 2008;26(17):2890-2894.
14. Blagosklonny MV. Why therapeutic response may not prolong the life of a cancer patient: selection for oncogenic resistance. Cell Cycle. Dec 2005;4(12):1693-1698.
15. Magnitsky S, Roesch A, Herlyn M, Glickson JD. In vivo and ex vivo MR imaging of slowly cycling melanoma cells. Magn Reson Med. Nov 2011;66(5):1362-1373.
16. Al-Hajj M, Becker MW, Wicha M, Weissman I, Clarke MF. Therapeutic implications of cancer stem cells. Curr Opin Genet Dev. Feb 2004;14(1):43-47.
17. Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science. May 2009;324(5930):1029-1033.
18. Thompson CB. Attacking cancer at its root. Cell. Sep 2009;138(6):1051-1054.
19. Thompson CB. Metabolic enzymes as oncogenes or tumor suppressors. The New England journal of medicine. Feb 19 2009;360(8):813-815.
20. Shestov AA, Mancuso A, Leeper DB, Glickson JD. Metabolic network analysis of DB1 melanoma cells: how much energy is derived from aerobic glycolysis? Advances in experimental medicine and biology. 2013;765:265-271.
21. Warburg O, Wind F, Negelein E. The Metabolism of Tumors in the Body. The Journal of general physiology. Mar 7 1927;8(6):519-530.
22. Glunde K, Raman V, Mori N, Bhujwalla ZM. RNA interference-mediated choline kinase suppression in breast cancer cells induces differentiation and reduces proliferation. Cancer Res. Dec 1 2005;65(23):11034-11043.
23. Magnitsky S, Belka GK, Sterner C, Pickup S, Chodosh LA, Glickson JD. Lactate detection in inducible and orthotopic Her2/neu mammary gland tumours in mouse models. NMR Biomed. Jul 5 2012.
24. Magnitsky S, Vite CH, Delikatny EJ, et al. Magnetic resonance spectroscopy of the occipital cortex and the cerebellar vermis distinguishes individual cats affected with alpha-mannosidosis from normal cats. NMR Biomed. Jan 2010;23(1):74-79.
25. Hara T, Kosaka N, Kishi H. PET imaging of prostate cancer using carbon-11-choline. J Nucl Med. Jun 1998;39(6):990-995.
26. Arlauckas SP, Popov AV, Delikatny EJ. Direct Inhibition of Choline Kinase by a Near-Infrared Fluorescent Carbocyanine. Mol Cancer Ther. Jul 15 2014.
27. Larson SM, Schwartz LH. 18F-FDG PET as a candidate for "qualified biomarker": functional assessment of treatment response in oncology. J Nucl Med. Jun 2006;47(6):901-903.
28. Gaude E, Frezza C. Defects in mitochondrial metabolism and cancer. Cancer & metabolism. 2014;2:10.
I would like to acknowledge Dr. Geetha Mohan for helping me in the lab throughout my stay at UCSF. In addition, I thank Melanie Regan for teaching me how to use the Gamma Counter. I acknowledge Mark Kelly, Dr. Jose Izquierdo, and Mr. Alan Wang for assisting me with the NMR Spectrometer. I would like to thank the Wistar Institute for supplying the WM3734J human melanoma cell line, and the UCSF Cell Culture Facility for supplying media. Finally, I would like to thank Dr. Sharmila Majumdar for providing key edits to my paper and supporting me throughout my stay at UCSF.