From OpenWetWare


Home        People        Materials        Schedule        Help        Discussion        Science in the News        Productivity Tools/Apps       

Tim Bates Discussion Questions

(wow so good!)

  • Maureen 19:30, 11 October 2017 (PDT) Way to make us look :)

Week 1: Replicative Errors in Cancer

Please see presentation 1 for this week.

Week 2: Persistent Data Identifiers

1) What is the best way to ensure links will continue to work if the hosting service closes?

  A: Always use services with clear data archiving policies discussing the fate of existing data in the case of closure.

2) Who should be responsible for paying the cost of data cataloging and retention?

  A:  This is a very complex issue that has no real answer, but instead depends on the situation and the parties involved.  In the context of publicly funded research, it is important that the data be made available to all taxpayers.  However this entails significant extra costs due to management and distribution that would also be the responsibility of the taxpayer to fund.  The case of private data is more simple, requiring only that data be preserved and made available to those who payed for the research that generated the data.  However, many privately funded studies intermingle with public studies, sharing data and resources.

3) Is the use of alphanumeric characters and symbols to reduce the length of unique identifiers warranted compared to simply using longer strings of numerals only?

  A:  The increase in data storage capacity increases dramatically with the number of increased characters available, however the additional complexity in terms of assigning and processing these more complex identifiers requires some cost benefit analysis.  For example, if a service uses 6 character identifiers of only numbers, then the possible number of identifiers is 10^6 = 1,000,000.  If that same service now allows lower case letters than the possible identifiers becomes 36^6 = 2,176,782,336; however, adding letters will require additional processing to watch for real words, functional code, and combinations which appear to be real words.  If the more complex identifiers require only twice as much code to handle, then processing of the same number of requests could take twice as long.

Week 3: Telomere Length Limiters

1) What is tamoxifen and why did they use it in their assays?

  A:  Tamoxifen is a telomerase activator and acts to lengthen telomeres which have been excessively shortened.  They seem to have been using it in order to test how exactly tamoxifen causes this change.

2) Is their use of a polyclonal antibody here warranted?

  A:  I think their assay did a good job of demonstrating specificity, but I was not able to find any other tests where they actually used this antibody.

3) Do you agree with the conclusion of this article: TZAP is preferentially recruited to and causes trimming of telomeres?

  A:  While it does look like TZAP and TRF1/2 are negatively correlated I am not completely convinced by their evidence that TZAP somehow prevents TRF1/2 from binding.  They looked at binding in western style gel shift assays, but it is unclear how relevant this is outside of the physiological context.  I would have liked to see more colocalization in cell assays.

Week 4: Transposons

1) What is the purpose of the deadlock-nanobody fusion?

  A:  The nanobody is specific for GFP and becomes active inside the cell.  GFP is fused to TRF2.  The purpose of this fusion is that it allows deadlock to bind directly to TRF2 and rescues the lethal phenotype caused by the lack of moonshiner.

2) How can Pol II transcribe efficiently in heterochromatin when it has been well established that general rates of transcription are lower?

  A:  This is the issue that is addressed in this paper.  It seems that the complexes discussed move the chromatin apart and provide additional access for the transcription machinery.  Since rhino binds to H3K9me3, the complex is anchored using this heterochromatin mark.

3) Transposons are highly regulated using the heterochromatin and piRNAs, what would be the consequences of loosing some or all of this regulation?

  A:  I would guess that the transposons would start jumping around the genome more frequently and would disrupt many more genes as well as duplicating, which would cause more of the genome to be made up of transposons.  Hybrid dysgenesis occurs when two species of drosophila are crossed.  The egg provides its own transposons in addition to the piRNA system for their silencing.  The sperm then introduces a new set of transposons from a different species, but without the regulating piRNA.  The result of this is many active transposons that jump around frequently until the piRNA system from the sperm genome becomes active.

Week 5: Topological Chromatin Domains

1) What is Hi-C and how does it inform us about chromatin topology?

  A:  Hi-C is not just an orange flavored beverage, it is also a method for looking specifically at large scale topological structures in chromatin.  It stands for Chromosome conformation assay, and involves several steps: Crosslinking of chromatin, Restriction digestion, Ligation of the digested ends, Reversal of the crosslinks, Measurement of the resulting products, usually by sequencing, microarray, or qPCR.  The results of this experiment give information about the association in space of segments of DNA which may be separated by long sequences.  In theory, this method could provide a complete 3D reconstruction of the nucleus and give valuable information about which DNA sequences are in close proximity.

2) This paper discusses many physical deformities, what are the definitions of the relevant conditions?

  A:  Brachydactyly - shortened bones in the fingers and toes.
Syndactyly - fusion of two or more digits
Hypoplasia - incomplete development of a tissue Dysgenesis - abnormal development of a tissue
Synostosis - fusion of two adjacent bones through the growth of additional bony substance
Polydactyly - the presence of additional digits on one or more of the hands and feet
F-syndrome - a form of acropectorovertebral dysgenesis which results in hypoplasia and dysgenesis of the fingers and frequently syndactyly of the first and second fingers and synostosis of the fourth and fifth fingers. F designates name of the family in which this specific subset of the condition was identified.

3) Do you think that these results would be better explained by a theory that does not involve 3D structures?

  A:  This is more of a discussion topic, but one could imagine a situation with a highly complex orchestra of protein interactions.  I feel like this paper did a good job of showing that the topological domain theory is the most reasonable solution that fits the data.

Week 5: Topological Chromatin Domains

1) Is Rad21 required only for making the loop structures or also for their maintenance?

  A:  The experiments in this paper make clear that Rad21 is required for the chromatin loop structures to remain intact.  When Rad21 is knocked down by ubiquitinylation, the complexes which were formed previously fall apart.  No new complexes appear to be made during this time, indicating that Rad21 is also necessary for setting up these loops.  Experiments with NIPBL, which is a cohesin loader protein may show different results.

2) Auxin is a plant hormone; What is it doing here in mammalian cells?

  A:  In this system, auxin is being added exogenously as an inducer of the AID (auxin-inducible degron).  Essentially, endogenous Rad21 is replaced with a version containing an AID domain and a ubiquitin ligase is inserted at a different locus under the control of an auxin-inducible promoter.  When auxin is added to the media containing the cells it finds the promoter and activates the production of the ubiquitin ligase which will signal for degradation of any protein with the AID domain.  Auxin is simply used because it is a powerful inducer which is not present in mammalian cells.

3) Why does the genome structure not fall apart completely when the cohesin proteins are knocked down?

  A:  The reason that genome structure is somewhat maintained during the periods of cohesin knock down is that there are several other parallel methods of genome organization.  TADs are set up by alternative methods that do not rely on cohesin loops, and thus are relatively unaffected by this change, though many TADs are thought to contain loops and may show some level of disturbance.  Further, there are other loop producing proteins such as the superenhancers which are dependent on a different set of proteins, the condensin complex. Both of these factors mean that even if all CTCF based loop structures are destroyed then larger genome structure and even some other loops are maintained.

Week 6: Hsp70/90

1) Why might the E936G mutant be so different from the E936K mutant FANCA?

  A: It is interesting that different mutations in the same location would give different phenotypes.  Normally, one would expect a mutation that leads to a significant defect would affect protein stability or the energy of folding. These two possibilities might be the case for FANCA, but there is one additional option because of the fact that one mutant encodes a glycine, while the other codes for a lysine.  These amino acids have very different properties, and the original amino acid glutamate.  This amino acid location must be key to the overall function of FANCA and changing out the normal acidic amino acid for the flexible glycine or the basic lysine may have different effects due to the fact that this key interaction could happen in many ways, but is generally strongly biased by the chemistry of the normal glutamate.

2) What was the reason for performing so many viability assays?

  A: Viability assays are very easy to perform, which allowed the authors to screen many different proteins and provided a consistent readout.  While I understand the reasons for this choice, I would have liked to see a more functional assay.  Cell death can be caused by many different factors, and not all types of cell death are the same.  Apoptosis and necrosis will, however, show up similarly in the PI assay.  As they are looking at chaperones, it would have been preferable to look for the general state of protein folding in the cell, potentially using something like 8-Anilinonaphthalene-1-sulfonic acid, which changes its fluorescence upon binding of hydrophobic regions of unfolded proteins.

3) What was the purpose of testing multiple chemotherapy agents at multiple temperatures?

  A: Each of these chemotherapy drugs promote DNA damage of some form such as crosslinking or alkylation.  The authors were looking at this because the fanconi anemia pathways are known to involve DNA repair.  Increasing the temperature decreases protein stability, and as they were looking at mutants of the fanconi anemia, one would expect some temperature sensitivity. It was interesting to compare the different mutants under these conditions to both wild type and null.  The different chemotherapy drugs used may help indicate which DNA repair pathways are most affected by each mutation.

Week 7: Cell Cycle

1) Why did they use single cell imaging?

  A: Single cell techniques provide a great look at how an individual cells are responding to treatment.  This helps prevent the smoothing of signals caused by averaging many cells and the bias introduced by many cells not being exactly aligned in the cell cycle.  The drawback is that the statistical power is significantly reduced and vastly increased noise.  It would have been nice to perform more replicates to gain statistical power.

2) Why are they looking at γH2AX, and how is this related to NCS?

  A: γH2AX is a marker for DNA damage.  More specifically, γH2AX is a phosphorylated histone variant that seems to localize to double stranded breaks in live cells.  This is generally recognized with a specific antibody.  NCS is Neocarzinostatin.  It induces double stranded breaks without causing interstrand crosslinks.

3) In supplemental figure 2c, what is meant by maximum amplitude vs average EGF activity?

  A: It appears that max amplitude in this figure refers to plot b and is looking at the amplitude of the tallest peak only from each trace.  The average is probably the just looking at the total average brightness from these traces.  The reason these are so different and have differing trends is because the peak intensity seems to peak at intermediate EGF concentrations, while at high EGF concentrations the signal increases overall, but the peaks become shorter and less defined.

Week 8: Cap Independent Translation

1) What is the toe printing assay?

  A:  My understanding is that the toe printing assay involves in-vitro transcription of an mRNA followed by binding of the mRNA to a solution of proteins.  In this case the proteins used were ribosomal subunits and initiation factors.  Then, a primer is used which binds to the very beginning of the mRNA, near the cap (or lack of cap).  Reverse transcriptase is used to generate DNA, but stops once it reaches an obstruction in the form of a protein bound to the mRNA.  This gives a very good look at where and if proteins are binding to the mRNA.  This technique is unidirectional, however, which means that you may only look at the 5' end of where the protein interacts, and if multiple proteins are interacting, then you also only see the point closest to the 5' end which is bound.

2) Why are the scales different in figure 2?

  A:  This is more of a comment than a question.  I am not sure why the authors decided that the subfigures in figure 2 should be on different scales in the sense that A,C,D,E are all on absolute scales based on the luminescence from the luciferase assay, but B is on a relative scale based on a control which is not included in the figure.  I think the control is analogous to the furthest right bar on A, but one cannot know for sure.  I believe it would be better to use an absolute scale for all of these plots.

3) Hsp70 is introduced in the last figure, and the authors claim relevance to this cap independent translation mechanism. I would like to discuss with the class regarding if this makes sense, and if they did a good job showing that Hsp70 is important.

  A:  Before the discussion, here are my thoughts.  The Hsp70 link is weak in my opinion.  I am not entirely sure how it ties into the cap independent translation work that they were doing earlier in the paper other than that Hsp70 is induced during times of stress, which is a proposed reason for the CIT pathway.  If they are saying that the Hsp70 gene is involved in CIT then I would like to see more showing that it responds similarly to the constructs used in previous studies.  If they are saying that the Hsp70 protein is involved in the CIT pathway, then they should have included purified Hsp70 in their previous experiments to show its effects.  However, I may be confused as to the reason why this set of experiments was included in the first place. 

Week 9: CRISPR Nucleotide Detection

1) What is collateral cleavage?

  A: I feel that the paper did a poor job of explaining exactly what was meant by this term.  It sounded to me like they meant that once the sgRNA's cognate sequence is found, the Cas protein becomes activated, and is not immediately de-activated after cleaving the cognate sequence.  The continued activity allows cleavage of the collateral quenched oligos.  If this is a general property of all Cas proteins, then it could end up being a reason for some of the non-specific cleavage noted in in vivo systems.

2) Given the specificity profile of Cas proteins, what kind of accuracy can you expect from this system?

  A: We learned from Jennifer Doudna's talk that different positions in the guide RNA have different propensities for mismatching.  Generally, the first nucleotides are more specific and the later nucleotides are less specific.  Each Cas protein has a different range, and you would need to quantify this for the SHERLOCK Cas proteins in order to determine the false positive rate, and which sequences are going to be more likely to be problematic.  It would be best to design guide RNA's to have your SNP of interest in the highest fidelity location.

3) What laboratory tasks will be improved by this method?

  A: This will allow detection of SNPs in any oligo and could potentially assist with detection of very low quantities of proper pcr product in an experiment that may otherwise be labeled as a failure.  It may also help detect specific DNA directly from cell lysates, 2 aM is incredibly sensitive.

Many have asked, so here are some pictures of my new kitten!

  • Maureen 12:50, 7 November 2017 (PST)this kitten has two modes. Super surprised and asleep.