- 1 Alec's Discussion Questions
- 1.1 Week 2 Discussion Questions on Bad Luck and Cancer
- 1.2 Discussion Questions - TZAP
- 1.3 Discussion Questions - Rhino and piRNAs
- 1.4 Discussion Questions - TADs
- 1.5 Discussion Questions - Cohesin and Loop Formation
- 1.6 Discussion Questions - Hsp90 and Variation
- 1.7 Discussion Questions - P53 and Cell Cycle Memory
- 1.8 Discussion Questions - Cap Independent Translation
- 1.9 Discussion Questions - CRISPR-Rx
Alec's Discussion Questions
Week 2 Discussion Questions on Bad Luck and Cancer
1. Is there any experimental evidence supporting this model stating R contributes to the majority of the mutations for oncogenesis when compared to H and E? What would an experiment like this look like?
There isn't much experimental evidence supporting the conclusions in the papers, the conclusions were drawn from data analysis and making conservative assumptions about variables that weren't known, the main one being total stem cell division. Experiments that could test the conjectures made in the papers could vary, but would need to test the correlation between stem cell division number and oncogenesis/tumorigenesis (such as growing stem cells under the same conditions, except one group would be exposed to mild UV light and the other group would not; if R is the major factor contributing to tumorigenesis then the UV exposure would not affect tumorigenesis development over time).
2. These papers seem to be examining population data for all cancer types. Were there any similar studies correlating total number of stem cell divisions (which correlates with age) with cancer incidence for one cancer type? For example, is there a correlation between age and incidence of pancreatic cancer, or cancers that are prevalent in children such as ALL?
Again there doesn't seem to be much more studies about this, it is a relatively new idea that R is a major contributing factor to cancer formation in tissues.
3. During the presentation, a graph was shown with three different categories of correlation - groups of tissue where a major contributing factor is environment (making cancer incidence higher that expected for the total number of stem cell divisions), tissues where stem cell division and cancer incidence is well correlated, and a group where the cancer incidence is lower than expected for the number of stem cell divisions. How do the authors explain the last group?
The authors determined that there must be some factor that helps prevent oncogenesis in these tissues, such as better DNA repair mechanisms or less error rate during replication.
After thinking about the answer for a bit (And discussing it with other classmates), a follow-up to this answer would be why isn't there emphasis on these tissue types? There seems to be some factor that is helping prevent cancer formation in these tissues, and if we can elucidate these mechanisms this could be a potential therapy path by introducing some of the preventative mechanisms into other tissues.
- Maureen 12:20, 11 October 2017 (PDT) cool idea. I wonder how you might go about this?
- Maureen 12:20, 11 October 2017 (PDT)nice job, reasonable questions/answers and comments.
Discussion Questions - TZAP
1. Why can't the ChIP experiment determine telomere length dependent Shelterin density? The paper states:
"As a result, cells with long telomeres have a lower density of the shelterin complex compared to cells with shorter telomeres, a difference that is not detected by ChIP (11)."
ChIP will only enrich for bound DNA up to the point that all of the proteins are bound - therefore ChIP couldn't detect unbound extra DNA and so a different technique needed to be used.
2. How did the authors figure out that three last zinc finger motifs are all needed for binding, instead of just one (or different combinations of two such as 9 and 10 or 9 and 11)? The supplementary material as well as the article indicates that this was not tested.
Both the presenters as well as other students answered that they assume these were tested but the results were not published.
I don't find this an acceptable response. I feel that if the authors tested many different combinations of the zinc finger motifs they would have published it in the supplemental information at the very least. Just saying "well I think that it was done" is not evidence that it was done - it is pure speculation.
3. How does the new model of telomere regulation apply to somatic cells that don't have active telomerase?
The idea now is that telomerase is now transiently active in all cells, and telomere degradation is thought to occur due to a buildup of mutations in the shelterin and TZAP proteins instead of chromosomal shortening due to replicaiton.
Discussion Questions - Rhino and piRNAs
1. There is a conserved region between the moonshiner protein and TFIIA-L of unknown function. Has there been any experiments probing the function of this region in TFIIA-L (which seems to have more information on its structure/function)?
There doesn't seem to be any.
2. The volcano plot in Figure 1c shows some proteins that were enriched during the co-IP/MS experiment that were not identified in the paper. Is there any information on these proteins? Why were the top four the only ones that were identified?
3. Why don’t all rhino localized piRNA regions/clusters recruit moonshiner? Do these locations have the deadlock protein localize to them as well, or just rhino?
4. If there are no potential genes in the human genome with sequence homology to some of these proteins, how would one go about finding an orthologous gene?
There wasn't enough time in the Q/A session to ask all of these questions.
Discussion Questions - TADs
1. How are the TAD boundaries determined? CTCF seems to bind at the proposed boundaries but it seems to also bind at other places as well that don’t correspond to the boundary regions. How do we know that the TADs delineated in the paper are not subdivided into smaller regions?
The Hi-C protocol includes an analysis method that determines boundaries via directional interactions between DNA sequences (i.e. there are interactions on one side but none on the other).
2. How are TADs thought to stop interactions between components of different TADs? Is it largely structural sequestration?
Structural sequestration seems to be one of the primary methods of preventing interactions of DNA sequence elements (genes, regulatory elements, etc) between TADs, however it may not be the only mechanism of preventing this sort of inter-TAD interactions.
3. Are TAD rearrangements thought to contribute to the imprinting on chromosome 15 that could lead to Prater-Willi/Angelman syndrome? There seem to be some similarities, such as CTCF binding to control differential gene expression.
"Diploid Hi-C Maps Reveal Homolog-Specific Feature, Including Imprinting-Specific Loops and Massive Domains and Loops on the Inactive X Chromosome"
It seems that this may be the case after all - though it may not be specific for the imprinting for Prater-Willi/Angelman syndrome.
Discussion Questions - Cohesin and Loop Formation
1. Is there another cross-linker other than formaldehyde that can be used?
Yes, but formaldehyde is one of the best crosslinkers.
2. It was shown that not many genes were changed in expression after loss of the loops. What, therefore, are the functions of the loops? If it is for 'fine tuning' of expression, would we expect to see a greater change in gene expression with the loss of loops if the studies were done in a single-cell format?
Probably not - since the general structure of the chromatin is preserved it is thought that the loops aren't directly changing expression.
3. Why, in the introduction (Whit et al, 2015), was there not a loss of cohesin binding in one of the DNA areas when CTCF motifs was deleted?
This is promoter region and CTCF independent - an example of a different function of cohesin other that CTCF loop formation.
4. Is cohesin recruited to specific histone modifications or is it a more general method - are the 'fast forming' loops due to targeted recruitment or more from chromatin accessibility?
Both - cohesin is recruited to H3K27Ac sites to begin extrusion, and it is also thought that since to the more open nature of the chromatin at regions with active marks cohesin will bind and begin to extrude the DNA at these regions more rapidly when compared to the more closed and inaccessible regions of DNA.
Discussion Questions - Hsp90 and Variation
1. Would the variation in Hsp90 sequence itself play a role in masking variation?
2. If Hsp90 is one of the most abundant proteins in the cell how would overexpression therapy apply?
It would probably be more of a therapy for patients with defective expression of Hsp90, thus causing the phentoype.
Discussion Questions - P53 and Cell Cycle Memory
1. Is the p53-mediated quiescence reversible in daughter cells? We learned that in cells with DNA damage, activated p53 will arrest the cell cycle, and can resume if the damage is repaired - would this be the case in daughter cells?
Quiescence is definitely reversible, so this would be the case for daughter cells that inherited activated p53 during mitosis.
2. This was talked about in the discussion on Tuesday, and is more of a comment instead of a question. I, along with others, don't believe that the inheritance of cell cycle determinants boils down to p53 or CyclinD1, as proposed in the paper. For instance, CyclinD is not necessary for cell cycle progression (there are redundancies), and p53 is not the only protein that arrests the cell cycle. I believe that this inheritance is probably more complex, but I do agree with the paper's conclusion that a quiescence state can be inherited in mitosis. It is also possible that, at least in some instances, p53 and CyclinD levels can be indicators of whether or not the quiescence state is inherited.
3. What are the implications of this new finding. Would it be possible that this heritable memory plays a role in certain phenomena such as age-related clonal hematopoiesis?
Discussion Questions - Cap Independent Translation
1. Is there any evidence suggesting 5' capping and 5' UTR N6-methylation of adenosines are mutually exclusive?
It seems there is evidence that they are not mutually exclusive.
2. Do the uncapped but 5'UTR methylated mRNAs exhibit a more rapid degredation (much like the 3' UTR mRNAs that have methylated adenosines do)? If yes, would this suggest a more rapid, but shorter-acting, response to stress such as heat shock?
3. Do the methylated but uncapped mRNAs show a difference in nuclear export when compared to a capped and unmethylated counterpart?
Discussion Questions - CRISPR-Rx
1. This technology seems to add a fast, accurate, simple, and cheap diagnostic tool for clinicians. What are some of the limitations of SHERLOCK?
2. It seems that the cost of the crRNA/Cas system used is minimal, but what about the RPA reagents? Does the cost of every reagent factor into the $0.61/test cost stated by the authors?
3. Would there be any way to make the test able to assay multiple DNA/RNAs in parallel (i.e. modifying the Cas enzymes to only act upon certain RNAs in an unspecific manner so multiple fluorophores could be used so that each spot on the paper could test multiple SNPs/virus serotypes/anything else)?