Kristen M. Horstmann Week 10 Journal
"Transcriptional Regulatory Networks in Saccharomyces cerevisiae"
10 Unknown Vocab Words
- nucleate: the quality of having a nucleus
- myc epitote tag (or just epitote tag): "epitope is a portion of a molecule to which an antibody binds. In most cases, epitope tags are constructed of amino acids. Epitope tags are added to a molecule (usually proteins) which an investigator wants to visualize. Visualization can take place in a gel, a western blot or labeling via immunofluorescence"
- immunoblot analysis: led me to Western Blot Analysis, which I will assume is the same: "procedure in which proteins separated by electrophoresis in polyacrylamide gels are transferred (blotted) onto nitrocellulose or nylon membranes and identified by specific complexing with antibodies"
- peptone: "soluble and diffusible substance or substances into which albuminous portions of the food are transformed by the action of the gastric and pancreatic juices"
- dextrose: "sirupy, or white crystalline, variety of sugar, C6H12O6 (so called from turning the plane of polarization to the right), occurring in many ripe fruits. Dextrose and levulose are obtained by the inversion of cane sugar or sucrose, and hence called invert sugar. Dextrose is chiefly obtained by the action of heat and acids on starch, and hence called also starch sugar. It is also formed from starchy food by the action of the amylolytic ferments of saliva and pancreatic Juice"
- Chromatid immunoprecipitation: immunoprecipitation is "the small-scale affinity purification of antigens using a specific antibody and is one of the most widely used methods for antigen purification and detection" so I will assume that chromatid immunoprecipitation tested for chromatids or DNA in the genes probably through the immunoblot analysis
- thiamine: "Thiazolium, 3-((4-amino-2-methyl-5-pyrimidinyl)methyl)-5-(2-hydroxyethyl)-4-methyl- chloride A B vitamin that prevents beriberi; maintains appetite and growth.More commonly known as vitamin c"
- biosynthesis: "The production of a complex chemical compound from simpler precursors in a living organism, usually involving enzymes (to catalyze the reaction) and energy source (such as ATP)" so thiamine biosynthesis is basically the production of thiamine (vitamin c) through biological processes
- leucine: "most abundant amino acid found in proteins. Confers hydrophobicity and has a structural rather than a chemical role. a white crystalline amino acid occurring in proteins that is essential for nutrition; obtained by the hydrolysis of most dietary proteins"
- intergenic: "occurring between genes or occuring between multiple genes" I assumed this using simple clues but wrote down the word anyways because there's no harm in looking it up to make sure
- Scientists still trying to understand how cells control global gene expression
- Depends on recognition of specific promoter sequences by proteins
- Knowledge of the regulator sites could possibly give enough information to create transcriptional network models
- Lee et al first investigated how transcriptional regulators bind through genome analysis
- Strains of Saccharomyces cerevisiae were laced with myc epitote tag into the coding sequences (confirmed insertion through immunoblot test)
- From the immunoblot, 106 of the 124 proteins were detected
- Each strain was gorwn in three cultures: yeast extract, peptone, and dextrose. No further conditions or temperature were specificed
- Calculated p-value for each array spot. All results discussed had a p-value of .001
- False positive frequency is likely around 6-10%
- Around 1/3 of DNA interactions are not in p-value of .001
- Fig 1:
- A: Methodology
- Diagram of tagging yeast cells with c-myc tag
- Diagram of chromatin immunoprecipitation
- Promoters identified to microarray
- B: Effect of P Value threshold
- Stricter p-values reduce amount of data but also reduce likelihood of false-positives
- A: Methodology
- ~4000 regulator interactions observed
- 37% of promoter regions (2343 out of 6270) bound by one or more of the 106 regulators
- Many bound by multiple, which is typically only associated with regulation in higher-level eukaryotes
- Fig. 2
- A: Regulators bound per promoter region
- Distribution for actual location data shown along distributopm expected from same p-values
- B: Different Promoter Regions Bound by Each Regulator
- Number of promoter regions bound by each regulator ranged from 0-181
- A: Regulators bound per promoter region
- More than 1/3 of the promoter regions bound were by 2 or more regulators
- Regulator Abf1 bound the most regions (181)
- Regulators should be active with growth conditions different from yeast extract, peptone, and dextrose bound to the fewest promoter regions
- Identified 6 regulatory motifs:
- regulator which binds to promoter region
- Reduced response time to environmental stimuli
- Decreased biosynthetic cost of regulation w/ increased stability
- Multi-Component Loop
- Consists of regulatory circuit which requires multiple factors to close
- Perfect for feedback control
- Feedforward Loop
- Contains regulator which controls a second regulator with both regulators binding to common gene
- evolutionarily chosen during evolution for transcriptional yeast networks
- Can provide many features to a regulatory circuit
- Highly sensitive so can act as a switch
- Single-Input Motif
- Contain single regulator which binds to genes under specific condition
- Good for coordinating unit
- Multi-input Motif
- Set of regulators which binds to a set of genes
- Found 295 combinations of two or more regulators
- Each regulator bound can be responsible for regulating genes
- Regulator chain motifs
- Consist of chains of 3 or more regulators in which the previous regulator binds the next one
- Straightforward form in regulatory circuit
- Fig. 3- Pictorial diagram of these 6 motifs
Assembling Motifs into network structures
- Algorithm was created that examines over 500 expression experiments
- Defines a regulator-bound set of genes, G, which is used as “core” profile
- Bound by set of regulators, S (all P values .001)
- Genome is scanned for genes common to set G. Significant matches are examined for S
- P value is then relaxed to “recapture” data that wasn’t used because of P value
- Algorithm repeated for all gene combinations have been measured
- Used motifs to create yeast cell network structure automatically
- Ultimate goal: to create mathematical approach to create replica of cell cycle based only on the location/data of the regulators with no prior knowledge
- Fig. 4- Model for the yeast cell cycle transcriptional regulatory network
- Transcriptional regulatory network created from binding and expression data
- Boxes correspond to when peak expression occurred
- Blue Box: set of genes w/ common regulators
- Ovals: regulators connected to their genes w/ solid line
- Arc: defines time of activity
- Dashed line: gene in the box encodes outer ring regulator
- Created model based on peak expression of common expression multi-input motifs
- Three notable aspects:
- Model correctly assigned all the regulators to previously proven stages of the cell cycle
- Two relatively unknown regulators could be assigned based strictly on binding data
- Required no prior knowledge and was completely automatic
- Hopefully can use as a general outline for creating more complex network models
- Fig. 5- Regulatory binding network
- All 106 regulators displayed in a circle
- Sorted into functional categories (color coded)
- Lines follow regulators binding to each other/itself
Coordination of cellular processes
- Cycle regulators bind to other cycle regulators
- Multiple regulators within each category were able to bind to gene regulators that are responsible for control of other processes
- Stress response gene Yap6 bound to the Rox1 repressor and vice versa, creating positive and negative feedback loops
- Their findings were consistent with previous research’s findings
- Control of most processes characterized by complex networks
Significance of regulatory network information
- Identified network motifs reveal regulatory strategies that were selected through evolution
- Motifs can be used as building block to model larger network structures
- Transcriptional regulators that control other transcriptional regulators is highly connected
- Better understanding of transcriptional networks will require further knowledge of growth condition binding sites
Additional Questions and Information
- How did they treat the cells?
- They did not discuss detailed treatment of the cells as the cell growth was not the main aspect of the experiment. Mostly the cells were tagged then grown in 3 different mediums.
- What strains of yeast did they use? Halploid or diploid?
- The only information given about the cells was that they were S. cerevisiae
- What media did they grow them in? Under what conditions and temp.?
- Grown in yeast extract, peptone, and dextrose. No further information about growing conditions was given.
- What controls were used?
- No controls were discussed. I believe this is because they were more studying the overall common transcriptional cycle. They weren’t wondering about specific parts. They simply wanted a natural run with natural data in order to model it.
- How many replicates?
- They had 106 tagged strains and 3 different mediums, so it was repeated around 318 times.
- What mathematical/statistical method did they use to analyze the data?
- They used false positive analyzing and finding a small p value (<.001) with statistics, then also created an algorithm which scanned a genome multiple times to find the data to create a model.
- What transcription factors did they talk about?
- They hardly discussed individual transcription factors, but they briefly mentioned Thi2, Abf1, PUT3, UGA3, Gcn4, Yap6, and Rox1. They also had a list in a figure which categorized the regulators with what roles they play.
Not sure if electronic notebook is needed this week so I am writing one anyways.
- Wrote summary of article online only to have internet stop working as I pressed submit. Had to rewrite. Huge lesson learned.
- A little on the long side because I felt like this article was really more of two experiments: tagging and growing the yeast strains and creating an algorithm and computational model to analyze the results.
- Felt like article was very sparse on repeatable information and hard to see how the growth was conducted/how it applies to the computational model they keep discussing. Furthermore, the model they created is impossible to replicate. Really the only information on how they created it was that it was "derived from a combination of binding and expression data" which really doesn't mean anything specific.
- Thought this article was fairly cool and interesting at first, but as deeper reflection was required, I became very frustrated with how non descriptive and disjunct it seems. For example, there were very few answers for the overall questions Dr. Dahlquist wanted to make sure we addressed.
- Met up with Tessa and Lucia today to start creating our project.
- My available times only briefly overlapped theirs, but we were able to split up the subsections amongst us in order to start developing a presentation.
- I wrote and analyzed information on creating the algorithm and computational model. It was about 6 slides, but tomorrow we will need to find places where I can talk in the beginning or the end for a discussion.
- We had far more than 20 slides and will need to pare down the information a bit
- Planning on meeting with Dr. Dahlquist on Monday to ask her questions we had on the article in order to make our presentation more seamless.
- completed and refined powerpoint slides
- Will be presenting on p-value and statistics and also on the algorithm/computational model of the cell cycle
- Confirmed by Tessa and Lucia today that I did not need an electronic notebook but I didn't want to delete it, so I guess disregard my paranoia when grading/