- 1 Xiaolan Chang Discussion Page
- 1.1 Nucleic acid detection with CRISPR-Cas13a/C2c2
- 1.2 5' UTR m6A Promotes Cap-Independent Translation
- 1.3 HSP90 Shapes the Consequences of Human Genetic Variation
- 1.4 Competing Memories of Mitogen and p53 Signalling Control Cell-Cycle Entry
- 1.5 TAD for chromosome architecture
- 1.6 piRNA expression in heterchromatin regions
- 1.7 Telomere-associated protein
- 1.8 Bad Luck and Cancer
- 1.9 Potholes and Identifiers
Xiaolan Chang Discussion Page
Nucleic acid detection with CRISPR-Cas13a/C2c2
1. Would this method really be useful for cancer diagnostic, one of the primary application of this technique to first world countries? It would not be useful! There are circulating tumor DNA (ctDNA) that are fragmented DNA derived from tumors and believed to be released into our circulation through tumor apoptosis, necrosis, or secretion. However, there are many limitations. One being that normal tumors also secretion normal ctDNA during cell tumor or stress situations, so one barrier is to have high specificity to differentiate between normal and cancer cells. A second limitation is that high levels of ctDNA associated to tumors are coming from malignant tumors. This defects the whole purpose of early detection in the benign tumor stage. A third consideration is that the ctDNA from tumors could be coming from necrosis cells, so random genome cleavage. Therefore, finding a specific sequence to probe for a particular cancer is difficult when there are random genome cleavages to ctDNA sizes of ~200bp. Fourthly, the stability of nucleic acid, especially with the small size of ctDNA, is very low and concentration will vary through time. Lastly, from single cell studies, we found large mutation heterogeneity for the same tumor type. Thus, a larger group of specific probing sequences need to be examined per cancer type. From my option, there are just too many limitations to using nucleic acid as a biomarker for cancer.
2. The purposed method aims to detect very low nucleic acids (aM) is various infectious diseases and cancers, however, most standardized test has a range and cut point of what is considered "low". Thus, is the aim of this group, to measure "very low" levels of disease biomarker really applicable?
Being the FDA, it would most likely require a range for what is considered a positive signal. And the lower end of the range will not be the "lowest" measurable concentration due to potential conflict with how one defines a background vs "lowest" signal. So to be on the safe side would be to have the lower range end be higher than what they may experimentally measure. This will give a more confident diagnosis, while aiming to find the lowest positive signal may lead to false positives. When it applies to third world countries and being quarantine during disease outbreaks, I would not want a false positive result to be the reason for infecting a health individual.
3.Could we increase the specificity for the test by including multiple sequences with minor mutations specific to the same disease?
One of the limitations to this technique is that there many be mutations within the conserved regions that identify it to be a particular pathogen or cancer. Thus, it would be interesting to see if we could transcribe those sequences in cis to increase the specificity and probability of diagnosing a particular disease. So say that our synthesized strand contains three similar sequences with minor mutations between them, but they are all conserved for zika virus. The patient that we test has one particular mutation, so rather than a one to one match, there are three possible sequences to find a match. This will count down on the number of screening tests we need to conduct for the patients.
5' UTR m6A Promotes Cap-Independent Translation
1. Do you believe there is a different pathway for mRNA nuclear export for capped and uncapped mRNA?
Probably not because the only difference between capped and uncapped is a methylated G and a three phosphates linkage between the methylated G and the rest of the mRNA. The charge would still be negative from the backbone and RNA will still have self folding stability. The quality of uncapped mRNA being exported out of the nucleus, a function of capping, may be lower than capped mRNA. However, the polyA tail has the same function and may compensate the uncapped 5' end. It would be interesting to do experiments to study this.
2. One function for the 5' cap is to keep it stable and from degradation of the various cytosol components, do you believe that translation locations (within the cytosol) for capped and uncapped mRNA would be different?
There will still be RNA folding, like hairpin, if it's capped or uncapped mRNA, so most likely the uncapped mRNA will be shielded in similar manners as capped mRNA.
3. One of the studied evolutionary advantage for uncapped RNA is efficient translation during stress times. However, how is the kinase of the 5' UTF m6A methyltransferase, METTL3, changed during stressful situations (higher temperature)? Could METTL3 be stress (UV, temperature, salt, etc) resistance or are the m6A alternations already found in proteins destined for upregulation during stress?
METTL3 functions in normal cellular conditions and stressful times, such as increase heat, may disrupt it's natural function and stability. It may be interesting to study the kinetics of this enzyme during various stressful situation because it seems that uncapped mRNA is compensating for unfavorable biological functions, and METLL3 helps to methylate the uncapping. At the same time, the methylation marks may already be found in stress prone genes for regulation during stress. Additionally, there may be another methyltransferase that dominate over METTL3 during stress.
4. The paper only goes into one protein, HSP70, that is upregulated during stress and depends on uncapped mRNA. If we expand this experiment into a genome wide scanning for proteins translated primarily with m6A uncapped mRNA, do you believe that the proteins would service similar functions?
Uncapped mRNA seems to be an evolutionary compensation for survival during stress. Thus, if we did a genome wide scanning, like CHIP, we may find upregulation of various protein folding proteins, heat shock proteins, and epigenetic marks like methylation or acetylation. These proteins could serve the same purpose to help with survival. It would be interesting to do single cell analysis on this aspect so that we can confidently identify these subgroups in genes with similar functions.
5. Do you believe that we could induced m5A uncapped mRNA translation to help with aging or disease or when introduced in stressful conditions?
Most events that happen post stress and trauma are short lived until the cellular function can edit or regain homeostasis. Therefore, I think it would be inefficient to use this method for age and disease preventing. Most importantly, we do not know the long term consequences of having uncapped mRNA.
6. Reflecting back to the paper with HSP90 and HSP70, with phenotypic correction of FA pathway when associating with HSP90 > HSP70 and the fact that m5A uncapped mRNA increases translation of HSP70, do you believe that m5A regulated uncapped mRNA could promote or enhance disease pathway?
7. We learned in class that the 5'cap and 3'polyA tail shared similar purposes: stabilize mRNA, help with transporting it out of the nucleus, and help start translation. Additionally, histone mRNA lack polyA tail but sequence signal within the ORF helps to signal the mRNA out of the nucleus. With these two aspects, do you believe that there are additional factors in an uncapped mRNA that are helping to compensate for the lack of capping?
I believe that the poly A tail is playing a large role in helping maintain the regulation and function of uncapped mRNA because the 5'cap and 3' poly A tail share similar functions. But on the other end, maybe the cellular stage at the time of uncapped mRNA synthesis allows for the mRNA to be uncapped while maintaining normal function. It would be interesting to experiment the polyA tail length to see if there is a correlation between length and normal activity of uncapped mRNA. Additionally, the polyA tail helps to direct the 5' end towards the 3' end, itself, to allow for fast turnover of translation between the ribosomes could quickly reassembly on the AUG if they are close in space. Having the 5' and 3' end in space (a loop RNA) will allow for efficient translation. Hence, maybe the polyA tail is playing a larger role in initiation translation than the 5' cap.
HSP90 Shapes the Consequences of Human Genetic Variation
1. The introduction gave a good idea of the importance of HSP90, particularly during the developmental stage, so why is the expression of HSP90 only modestly induced compared to other chaperones?
Answer: The paper talks about the difference between HSP70 and HSP90: HSP70 binds to hydrophobic regions of unfolded/nonfunctional proteins to help with refolding while HSP90 binds to partially structured proteins to help buffer it’s defect. Hence, our body decided that HSP70 and many other chaperones that are critical for primary correction of misfolded proteins require more regulation and induction than chaperones that helps to “buffer”. Especially when partially structured proteins are still functional and may not be seen by the body as important areas to regulate when there are misfolded proteins that the body needs to attend.
2. The paper suggests that HSP90 could help buffer mutant defects, leading to weaker or normal phenotypic defects from mutants. Do you believe we could use gene therapy or chemicals to induced HSP90 expression? Would having a high concentration of HSP90 in the body be a good therapeutic approach?
Answer: Yes and no. The paper definitely showed normal FA pathway from mutants when buffered with HSP90 more than HSP70. However, the authors also proposed that HSP90 promotes oncogenic hyperactivity. It seems that their “buffering” abilities not also assist with disease phenotypic correction but also help to support irregular proteins involved with cancer promotion. Therefore, there is a fine spot between having too much HSP90 that could help us verse help oncogene proteins.
3. In Fig2F, it is interesting to see that the same amino acid change at 936 (E) could lead to different association to HSP70 or HSP90. E936G associated with HSP70 while E936K associated with HSP90. Do you believe that a focus on single amino acid change that switches a protein’s association to varying chaperone would be a good treatment approach?
Yes, even the authors proposed “single amino acid changes in FANCA can increase the dependence of the entire FA pathway on HSP90”. I think it’s very interesting to witness how a single amino change and their binding to chaperones align with the types of proteins those chaperone binds to. As I mentioned earlier, HSP70 binds to hydrophobic, nonfunctional proteins. E936G is now a new hydrophobic residue and the dramatic spatial difference from an E to G must have disrupted the protein’s regular function. On the other hand, E936K may have changed from an acidic to basic residue, it seems that the mutant is still viable. Therefore, it would be interesting to study single amino acids and it’s effects in binding proteins, and use huge data banks of information to manipulate our proteins to treat disease phenotypes.
Competing Memories of Mitogen and p53 Signalling Control Cell-Cycle Entry
1. The findings in the paper suggest that the decision for proliferation or quiescence of a daughter cell depends on the ratio of cyclin D1 and p21. Could we use this information to screen for infant health (by finding an associating between health and earlier cellular proliferation verse quiescence)
Answer: I think it would be very interesting to conduct a longitudinal study to see if there is an association between proliferation or quiescence of cells to the body's overall wellness. Since this paper suggest two factors, p53 and cyclin D mRNA, to be indicators for cell fate, we could measure that and see how a person's overall health or a particular disease phenotype could change with varying p53 or cyclin D mRNA levels. Ideally, proper or enhanced cellular division would signal for good health, in the action of replenishing old and damaged cells, especially with the findings that DNA damage does not carry into the daughter cells. It would also be interesting to see if cell fate for proliferation or quiescence would change with aging. Ideally, there should be more proliferation during development and more quiescence when we are older.
2. With findings that show factors that could enhance cellular proliferation, do you believe that drug companies will start to produce supplements containing factors like growth factors, cyclin D1 or other proteins in the cell cycle regulation pathway, to promote as an anti-aging supplement that enhance cells division? As researchers publishing data that highlights the important roles of isolated proteins, how much responsibility falls onto the researchers in making sure that companies don't alter to the truth for financial gain?
Answer: We talked about this issue in earlier classes where drug companies would charge hundreds for anti-aging pills or creams to uninformed consumers. I once encountered a skin product company that said that their researchers have identified the single protein responsible for aging and have engineered the protein into their skin care creams. Ridiculous! Realistically, these supplemented factors would not be able to maintain its stable or active form through the skin or intestinal barriers, hence, consumers are paying hundreds for useless supplements. As a researcher, it is hard to control how people interprets my findings, especially if media finds out about it and blows it out of portion (with headlines such as "coffee can prevent cancer" and "acrylamide in coffee can cause cancer"). This is why I believe that scientific outreach is important to educate and engage individuals outside of the science field.
3. Fig 2E shows that DNA damage (measure in H2AX puncta for phosphorylated histones) is higher before mitosis and lower after mitosis. The authors wanted to further clarify cells immediately after mitosis, they used hypo- and hyper- phosphorylated Rb proteins. Normally, phosphorylated Rb is associated with active E2F transcription, however is Rb protein a good protein to use in order to answer this question for measuring DNA damage after mitosis?
Answer: No, because you will not be able to differentiate G0, G1, and G2 from each other. Rb protein would be a good indicator for the S phase where transcription is turned on and off. If this question is trying to show that DNA damage is lost after mitosis, it would be more reasonable and even interesting to measure DNA damage (H2AX puncta level) throughout the many steps of cell division. Thus, rather than just saying that DNA damage is not carried over to the daughter cells, we can see exactly when this process occurs in. Additionally, we could study how long each process takes in order to correct a DNA damage or if there is a correlation between cell cycle pausing and amount of DNA damage in the parental cells.
TAD for chromosome architecture
1. Could a newly formed TAD boundary cause negative effects in a cell even though the function of TAD seems to provide genome stability?
Ideally, TAD is a good aspect of chromosome architecture: it provides stability to chromosome structure and it helps with the interaction of enhancers and promoters that are not close in sequence but required in space interaction. However, new TAD formation could lead to unpredictable consequences, good or bad or benign, depending on what kind of genes and/or proteins contacts it facilitates with it's new production. To witness it's affect, each TAD would most probably need to be examined independent.
2. In the review attached, it said that TAD is not disrupted by transcriptional activities in the cell? How is the TAD boundary not changed during transcription when in general chromosome is structurally changed for transcription (when different transcription activating proteins associate to the genome)?
They are unclear about the relationship between TAD and transcription. Does TAD affect transcription or would transcription affect TAD? Or maybe their interactions are in a cycle that it is hard to differentiate which matters more: TAD boundary for genome stability or transcription
3. In the section " Conservation of TAD structure across species, tissues, and rearrangements", it said that "the observed chromatin interacts appear to be independent of gene expression level" because EPH4A and PAX3 are upregulated, WNT6 is downregulated, and IHH is not detected. Why doesn't protein level, hence transcription, affect the genomic architecture of TAD because transcription generally should affect local chromosome structure? In general, does chromosome structure affect protein expression or does protein expression affect chromosome structure?
TAD boundary most likely provide a stable region for gene regulation of an uniquely identified gene region. The four genes listed are within one TAD boundary, next to one and another, and ideally would be transcribed in identical rate. However, different regulatory proteins may be in play to affect the expression of these gene, hence, you get differences in gene expression of genes even though they are adjacently in the same TAD region.
4. Could we just restore the needed factors for proper TAD boundary (DNA binding proteins, methylations, etc) to restore TAD function and normal phenotype? Would it be more efficient than studying the new TAD boundary that arises from chromosome changes?
Ideally, if we introduce the same players of a TAD back into a disrupted TAD, the structural function and gene interactions should also be restore (form fits function). However, that is not necessarily true all the time. It is hard to predict the changes of a disrupted or restore TAD boundary because chromosomes and our DNA are so dynamic. Additionally, I feel like it is worth studying both the function/consequence of new TAD boundaries and if we could restore disrupted TAD boundaries so that we could use this information to 1. predict future TAD boundary function, 2. use this information for potential therapeutic treatment is we find that associating certain enhancer to promoter show our desired phenotype, 3. use this information for therapeutic treatment to reverse diseases, 4. recognize patterns/outcomes so that we could manipulate TAD boundaries, and so many more possibilities.
piRNA expression in heterchromatin regions
1. In Figure 4D, the knockout of the left and right promoters still show expression in the center region of cluster38C1. The paper suggest an alternative mechanism for transcription initiation from within the cluster region. How would a research go about investigating this promoter and Moonshiner independent transcribed region?
The level of piRNA in the central region of cluster38C1 in both promoter knockout is lower than with both or just one (left or right) promoter, hence, the outer promoters still play a role in control central cluster transcription, but not the primary role. We could systematically knockout the central regions to see which segment of it is important for this alternative, outer promoter, and Moonshiner independent transcription initiation.
2. Moonshiner is a protein that binds to other proteins to assemble a transcription-initiation complex in many sites on both strands. Why do you believe that many transcription initiation site is needed when already a long strand of mRNA strand is being produced and then later cut into tiny 10-20bp piRNA?
Many transcription sites and bidirectional transcription may help to produce a high diversity and high quality of piRNA that can find and silent transposons quicker than if it is only transcribed by one strand and initiated in one site. PiRNA is very unstable because it is RNA and very small, so I would expect degradation to occur frequently. Additionally, complementary between the piRNA and transposon is an important step to silent transposons. Therefore, our body could have evolved in ways that allowed for bidirectional transcription and many site transcription initiation to better equip piRNA in silencing transposons.
3. The authors of the review said that "transposons are present in many copies in the genome, both inside and outside piRNA clusters". However, if a transposon is outside a piRNA cluster, then how could piRNA be made to bind to transposon and silent that region (by recruiting proteins needed to mediate histone methylation mark heterochromatin formation)?
This is honestly a difficult to answer question that I do not know the question for and forgot to ask the class. The function of piRNA is to bind to transposon and make that region of the chromosome heterochromatin (silent gene region). Additionally, piRNA sequences match with DNA sequence of the host and the transposons. If transposons are also found outside of piRNA production clusters then I do not see how existing piRNA can show complementary sequence with these transposons. Additionally, I do not see how these piRNA could physically travel to these transposon locations given that piRNA are unstable and extremely small RNAs. One explanation that I can think of is that new piRNA cluster may be found around these new transposons or these transposons may be relatively safe and not disrupting functional opening reading frames for the body to repair.
1. The experiments in Fig 2 use TRF2 WT and MT to study TZAP binding to telomeric repeats, why do you think is their decisions to use TRF2 instead of TRF1? Both the TRF1 and TRF2 are important for telomeric repeats, so is there a reason for choosing TRF2 or is it just for convenience?
TRF1 and TRF2 seem to both show similar and important activity in the shelterin complex. It seems that the authors just decided to use TRF2. A future approach, which may have been in the supplemental information is to do the same experiments with TRF1 to show that you can get the same activity with TRF1 knockout and TRF1 + Znf1-11 as TRF2 knockout and TRF2 + Znf1-11.
2. For Fig 2D, what is the purpose of including Znf1-11 in the experiment? It shows that just Znf1-11 only behaves like TRF2knockout, hence, chromosome packaging is unlike WT. So this may suggest that the zinc fingers-only cannot give proper chromosome packaging but the association of the zinc fingers with other proteins (here, both TRF2 and TZAP).
I feel that the author included the solo 11 zinc fingers to show that the activity of these zinc fingers alone do not give the activity that we are witnessing. It shows that the zinc finger needs addition assistant (i.e. bound to another protein) to give the proper chromosome packaging. A good experiment to do is to attach the zinc finger to a nonessential protein to a telomeric repeat of potential identical size, charge, structure of TZAP or TRF2 to witness if these zinc fingers are the primary component to chromosome packing.
3. The finding to this paper shows that we can potentially control the concentration of TZAP and shelterin complex to regulate telomere length. Could we apply this information for translational treatments to help control telomere length, hence control the aging of our cell.
This finding introduces a protein involved in regulating telomere length. This finding can definitely be used to add onto our growing knowledge of telomere length regulation but it in itself cannot be used for translational treatment. It's very hard to control the concentration of TZAP and shelterin complex. We could use gene therapy to increase gene expression of one. Say we make more TZAP enzymes of our body are trimming too much of our telomere, but then how can we control the aftermath of having a high TZAP concentration because length telomere is also disease prone. Thus, it is hard right now to make any disease approach using this new information.
Bad Luck and Cancer
The authors say that R is the major source of cancer. They define R as the number of random mistakes that occurs during DNA replication and that R increases as we age. How does this model fit into cancer for infants and children?
Answer: the paper doesn't talk about age, just that the R value increases as a stem cell continues dividing (older people would've divided more often than children and would have a higher R value). But by that standard, the R values for most tissues are relatively low for children, so what explains their cancers? Their E is also very low as well. The only explanation left is H, but what if H risk factors are low and not found in the family? There are too many factors involved in the causation of cancer and this paper simplified a lot of these factors when making their model. Overall, the model introduced cannot explain cancer in children.
The authors say that “exhaustively documented fact that about three mutations occur every time a normal cell divides and that normal stem cells often divide throughout life”, so does the collection of mutations per cell lead to more dramatized mutations per cell division (ie. an exponential increase in mutations per cell division rather than a linear increase of the consistent three mutation/ cell division)?
A: Like I mentioned before, the model was made from a lot of generalizing and theoretical thinking. The consideration of mutations in a cell that caused more mutations than the suggested 3 mutations per division was not considered in the development of this model.
Do you think the results from this paper will change the directions of future cancer research focuses?
I hope so. I learned from the presentation that the funding to research for early detection is very low. That was very surprising and unexpected. I believe that causation, diagnosis, and treatment should be funded at equal rates and findings for each type of cancer research could spark collobation and inspiration of new ideas.
Potholes and Identifiers
1. Do you believe that we should have groups look through old identifiers and delete old and unused ones? The Author proposed to never reassign or delete identifiers, but what if an identifier was not used at all and just created with linkage to no useful information? Should we delete those and free up identifier space and confusion?
A: I don't believe there are actual groups who manage identifiers to this extent, so searching the web to find unused identifiers may be a long process. Rather than focusing on the old, we could put more effort into education to make sure that everyone has a uniform knowledge of how to use identifiers.
2. Similar to NIH's requirement for ethics course works, do you believe that the NIH should require scientists to take classes related to the creation, usage, and reference of identifiers?
A: That would be very helpful. I feel that a one-day or half-day workshop for this topic is enough to help strengthen the knowledge of identifiers.
3. This paper was published in the Biology session of the PLOS, however, the Authors mentioned in the introduction that identifier management problems are not just in the life science fields but range from astronomy to law. If this paper was published in PLOS-Biology and gave a call to action for a more uniform and clear way of managing identifier, how would this information get to other disciplines?That'
A: That's a difficult task to do because professionals get their post-graduation education through their specialized groups. I feel that the easier way to approach a uniform education is to start early on, say undergraduate level when everyone is still taking similar classes and socialize in the same groups.