Sofia's Blog: Brain Tumor and Alzheimer's Research at MSKCC
Sofia's Blog: Brain Tumor and Alzheimer's Research at MSKCC
Hey guys, my name is Sofia Dimitriadoy, I'm a rising senior, and this summer I have the great opportunity to intern at Memorial Sloan Kettering Cancer Center in New York City (yes, the commute is terrible) in their Molecular Epidemiology Lab. Epidemiology is simply the science that studies incidence and prevalence of disease in large populations. Molecular epidemiology specifically focuses on the contribution of potential genetic and environmental risk factors at the molecular level to the source, distribution, and prevention of disease. So evidently, this is an instrumental field in the research of cancer!
Pictured to the left isn't actually the
building I work in (which is the Schwartz Research Building), but rather the newest MSKCC building, the Zuckerman Research Center (but I really only included a picture of this building in particular because it's pretty).
The project I've decided to undertake (one of many) is investigating various aging-related genes that may contribute to cognitive dysfunction in brain tumor and Alzheimer's patients, especially in those treated with radiotherapy or chemotherapy. What has been found is that brain tumor patients treated with radio/chemotherapy often develop cognitive dysfunction, but little known about the mechanism or risk factors for this. It was recently found that the є-4 allele of the apolipoprotein (APOE) gene moderates cognitive outcome in patients with brain tumors. So the new investigation I am partaking in is focused on examining other aging-related genes that may contribute to cognitive dysfunction, and to see if the APOE gene is associated with increased amyloid cortical retention in brain tumor patients treated with radio/chemotherapy. This latter information is crucial as having the ability to know which patients would be more susceptible to the negative impacts of treatment would necessitate other targeted interventions for them.
That's just a general overview of what my project entails--in future posts I'll go over the components of the research and what I do (and learn in the process) to contribute. Until then!
- Sofia
Week 1: 07/13-16/15
To start off I'll talk about the general environment of the lab I'm in--everyone is a lot friendlier and closer than I thought they'd be. On the first day we all (the lab technicians and the two other interns here) ate lunch together, and on the second day we even went out (NYC pizza is A+). Although everyone in the lab tries to help each other, most of the time they all have their own projects and tasks to work on. Every Monday there is a meeting to go over the progress of each project, and the items that must be completed for upcoming deadlines.
So now for the work I was actually involved in. Before anything else, I performed a few pipetting exercises. Pipettes are used all the time in the lab, since this lab works with DNA and RNA constantly. I learned that multi-channel pipettes can be a real hassle sometimes, but once you get the hang of them, they're a great time saver (although sometimes you must use single channel pipettes--fear of cross contamination with DNA samples or reagents is taken very seriously).
After this, I actually didn't do much lab work--instead, I did a lot more research on the computer. I was responsible for helping out in the completion of a large file with information about the single nucleotide polymorphisms (SNPs) and genes we wanted to investigate for the project. I had to find things such as a gene's official name, its ID, aliases, the genomic context of the SNPs (whether they're intronic, exonic, or intergenic upstream/downstream), chromosome coordinates for each SNP, which allele was the major and minor one of the SNP, the minor allele frequency (MAF) for differing populations, and finding potential proxies for the SNPs (i.e. other SNPs that are in high linkage with the original SNP, or in linkage disequilibrium). I used websites commonly used by researches in this field such as NCBI, dbSNP, Ensembl, RegulomeDB, and SNAP. The information in itself isn't all too hard, but it was a lot of busy work, and learning which databases were best for certain pieces of information, as well as learning how to navigate them, was pretty crucial. Below is a picture of a SNP shown in its genomic context. In this case, the SNP is intergenic and upstream (before the promoter region) of the gene.
I then received publications on SNP characteristics in general, multiplex PCR protocol, and user guides on how to navigate the MassARRAY System and the Assay Design Suite program (used for genotyping) for upcoming aspects of this project.
I also helped run a gel utilizing a method called gel electrophoresis, which essentially separates DNA (or RNA and proteins) and its fragments by its size and charge. This then allows you to analyze many different aspects of the DNA, depending on what you're looking for (such as quality of the DNA). I also had to start preparing a presentation for a lab meeting of next week on PCR additives (substances that help facilitate the PCR reaction).
One last major item of the week was a service meeting I attended with many other epidemiologists about the projected benefits and harms of mammography. It was interesting to see a room filled with experts on varying types of cancers coming together to analyze such an issue.
That about wraps up the first week. Until next time!
- Sofia
Week 2 (07/20-23/15)
So for this week I began getting more involved in the intricacies of the project. Obviously I won’t mention every detail so as not to make this dry, but I’ll try to add as much relevant detail as possible!
First item on the list for this week was to watch a webinar on genotyping with the iPLEX® platform, which is used with the MassARRAY® System. This webinar explained the basic chemistry behind the MALDI-TOF mass spectrometer, important and relevant case studies, and a basic overview on how to conduct assay designs using the Assay Design Suite.
So before I move on, I’ll give a brief overview of what all of this means. The Sequenom MassARRAY® iPLEX platform is basically a system with varying steps that enables you to genotype samples--or for the sake of my project--find out which allele is present in the SNPs being studied. Below is the general workflow.
So first you perform a PCR (polymerase chain reaction) with forward and reverse primers, which will amplify the region of DNA you want to analyze. Once this step is done, another PCR reaction takes place with a shrimp alkaline phosphatase (SAP) enzyme. This is basically a clean-up step, as the SAP enzyme degrades unincorporated dNTPs (a reagent used in the first PCR) and extra nucleotides to prepare for the SNP analysis. After that, one last PCR is performed with the extension primers. These primers are the ones that bind directly before or after the SNP, and will facilitate the replication of one nucleotide (i.e. your SNP). Once this is done, resin is added, which removes salts such as Na+, K+, and Mg2+ ions. All of the steps mentioned so far require lots of vortexing, spinning down (in centrifuges), and in some cases, mixing on a rotator (for the resin step) before proceeding onto subsequent steps. Below is how an extension primer would work:
Once all of this is done, you’ll be left with a 384 plate filled with your amplified DNA samples. However, the next step utilizes a machine that reads samples from a chip. To get your samples onto a chip, a Nanodispenser must be used. The machine will take up liquid from the plate, and then spot them onto a chip. The machine can be very finicky sometimes because of how sensitive it is to the moisture levels in the air, and must be adjusted (as well as cleaned!) daily.
The last step utilizes the MALDI-TOF mass spectrometer, which is what enables us to detect which nucleotide (i.e. allele) is present in the DNA. I’ll go more in depth on how this machine works (because it’s rather interesting!) when I actually get to that part of the research, in order to keep things flowing more chronologically.
But before any of this can happen, a design must be created using the Assay Design Suite program--which is what I did this week! The basic idea behind the whole process is that you want to replicate numerous specific sections of the DNA you want to analyze (the sections with the SNPs). To do this you must order specific primers for each SNP (because you need a primer to bind to unwound DNA for it to replicate) that will bind directly before the SNP, which will then extend one nucleotide (i.e. the nucleotide you are looking for, which is the allele of your SNP). The mass spectrometer then detects what this one allele is. Each SNP needs its own primer, and all DNA samples will be mixed with a cocktail mixture of all the extension primers for that one well (reaction). The issue is that sometimes certain primers may form primer dimers with other primers, or may bind to different parts of the DNA sequence other than before your SNP. Additionally, if there are other SNPs nearby the SNP of interest, it is impossible to predict which nucleotide will be there, thus preventing you from making a primer--an issue like this would usually result in a rejection of that particular SNP. Thus, the Assay Design Suite is able to analyze the SNPs you input and create primers for it, while at the same time providing feedback as to whether the primer is possible. If the primer interferes with another one, the design will be split into two or more “wells,” i.e. reactions (necessitating two or more mixtures of primers to be added to the same set of DNA samples). The amount of wells/reactions you have should be minimized, since reagents for separate reactions are extremely expensive (tiny 5mL tubes filled with reagent can cost up to $2,000!). So, my goal was to create a design with as little rejects and wells as possible. After working on this for a couple of days, I finally came up with the most optimal design (which happened to be the 11th design). Because we were testing so many different SNPs, there had to be two wells (the maximum per well is forty and we had over that number anyway, so it was not a loss). Unfortunately, some SNPs needed to be put off (including a poly-T and a VNTR, which are impossible to test using this method of genotyping).
One final thing I had to do this week was attend an hour and a half long Biology and Fire Safety Training class. Some of the things I learned about were the correct procedures to enact when threatened by a fire for the building, the rules and regulations associated with maintaining a laboratory, and the importance of consistently updating the correct labeling for all chemicals within the lab and chemical supply closets. Things like these are handled very seriously, and if a lab has too many violations (or maybe even just one major one), it can be shut down. There’s a lot to constantly remember, so I got to appreciate more of what it means to run and maintain your own lab.
That’s a lot of information for one post, so I’ll continue next time with what I do next week!
Until then!
- Sofia
Week 3 (07/27-31/15)
This week I continued working on the many different aspects of my project. However, of course, my full time here was utilized and I received materials and helped out with a variety of different things. I ran a PCR for an unrelated project, learned how to clean the Nanodispenser, learned how to spot a chip using the Nanodispenser (this was done for yet another project), ran this chip in the mass spectrometer, and received some literature, including an article on Lee Jones (whose work is related to exercise and its effects on cancer), a paper on molecular pathological epidemiology (a newly developing discipline that combines the fields of epidemiology with pathology to truly understand the source and spread of diseases), Primer of Epidemiology by Gary Friedman, and The Molecular Mechanics of Cancer Predisposition by Neil Sullivan. I also created a PCR ladder reference for the lab’s sake.
As for my project, I received a document detailing how to dilute primers (because when primers are shipped, they don’t come as solutions, but rather in a solid form the size of a speck of dust). I had to dilute the forward and reverse primers to 100µm, and the extension primers to 500µm. After this, I created primer mixes (two of each for well 1 and 2). We then ran a primer optimization, which is when you test the primer mix in the mass spectrometer to ensure all primers output around the same signal. It is then calculated for those that are too low how much more of the primer should be added to the mix. To the left is an example of what optimized primer signals should look like.
Perhaps most importantly, however, was that I had to retrieve the DNA we would be using and make sure everything was in order. This DNA had already been extracted from a different study, so I did not have to worry about that. What I discovered, though, was that many of the 150 samples I would be testing needed to be normalized (set to a standard concentration) again. Although there are specimen logs that tracked the concentrations and volumes of all samples, just to be sure, I re-measured all 150 samples using the Nanodrop 8000. This spectrophotometer enables the user to measure the concentration of macromolecules (including DNA, RNA, and proteins). It looks like the image to the right, and to the bottom is what it would look like to place samples to be read.
Some samples did indeed need to be diluted, so after doing that and reaching an optimal concentration range for all the DNA samples, I was ready to create a motherplate. Before, all the DNA samples are stored in individual tubes, which makes transferring DNA for separate trials a hassle. So after transferring all the samples to one single motherplate (with an ample amount of DNA), using these samples becomes much easier in downstream applications.
Once this was done, a PCR was ran with the forward/reverse primers, then with the SAP enzyme, and finally with the extension primers. Of course, it wasn’t that quick--PCR reactions take a few hours each. Once all the PCR reactions were finished, the samples needed to be mixed with resin.
Then it was finally time to place the samples on a chip and have it read by the mass spectrometer--and here is where I made my first encounter with failure in the lab. The results were pretty terrible in fact. My mentor commented she had never seen results so poor (which was encouraging to say the least). There were hardly any peaks and everything was mostly background noise (meaning the mass spectrometer was not able to detect any of the nucleotides we were looking for). We looked at the peaks very closely for troubleshooting, and we saw mixed indications of what could have gone wrong. Some SNPs had huge peaks for the extension primers, and no peaks for the nucleotides (indicating the extension primer never bound, or there was no/not enough DNA to begin with). However, other SNPs showed no peaks at all! After some troubleshooting, we think we pinpointed the problem (although one can never really know for sure). With that in mind, we prepared for a second trial for next week.
One last thing I did this week was attend a talk by Sheena Iyengar which MSK hosted. She is known for her book The Art of Choosing, and her talk, also titled this, focused largely on how we make decisions and why we choose the way we do. It was very interesting and I’m thankful for MSK having given me the opportunity to see her in person. If you want to view more of her work, you can check out a TED talk she gave here.
That’s all for this week!
- Sofia
Week 4 (08/03-05/15)
So in this final week, I was able to redo the entire process, and with really great results this time! So of course, I had to get DNA from the motherplate again, do all the necessary PCR reactions, mix it with resin, and spot the plate onto a chip. Since never pictured before, below are what the Nanodispenser and chips look like.
As promised before, I’ll go into a bit more depth about how the mass spectrometer works. Pictured to the left is what
it looks like on the outside. Nothing fancy, right? In fact, what this machine utilizes is something called matrix-
assisted laser desorption/ionization - time of flight mass spectrometry, or MALDI-TOF. The chip is coated with a matrix that allows for crystallization of the PCR product (now your analyte) on its surface. A laser is fired at the analyte to ionize the molecules. These ions move through a vacuum to an ion detector based on their mass (smaller molecules travel faster than larger ones). Time of flight measures the difference in time the molecules hit the detector, which enables the software to analyze the mass of the fragments. Pretty neat! Pictured below is a diagram that shows what I just explained.
After retrieving the results, it’s time to analyze the clustering. The traffic light is a quick assessment tool for the successful call rate per well, i.e. your samples. Dark green is best, and red is very poor. For my project, most wells were dark green except for two light green and one red!
The cluster plot is, what I think, the most interesting and visually rewarding, because it shows you which alleles were found in the samples for each SNP! It plots the low mass allele vs. the high mass allele. It also calculated Hardy-Weinberg values for each population per assay, and it allows you to click on any individual point to review the spectra and manually determine the quality of the assay. Additionally, it provides a whole population assessment of assay behavior and quality.
The well data provides detailed information on the processed data and allows you to manually interact with the data. You can look at one well at a time for genotype calls and the confidence for each call. You can also see information for the entire plate, and can sort this by different headers. Information such as area, resolution, and primer peak scores are stored here.
The spectrum shows the analyte signals, genotypes, and mass range. It provides annotations for all peaks, and a rough judgement of intensities, resolution, and signal-to-noise ratio.
Also useful is a histogram which has four categories: no calls, low mass homozygous, heterozygous, or high mass homozygous. It allows for the quick analysis of calls per assay over all samples within the plate.
So once the data is retrieved, it needs to be reformatted for easier analysis. This is done with Statistical Analysis Software (SAS). It’s organized in a way to more easily see the alleles found in the DNA samples per SNP. The next part is sadly something I wasn’t part of--all this information is sent to biostaticians, who will analyze it and perform the necessary calculations for determining the significance of the findings. In the future I’ll be able to see these results once processed through though!
One final activity I participated in was a DNA extraction from saliva samples for an project related to pancreatic cancer. This was a pretty fun experience, as the process is very visual and therefore in a sense more rewarding (you’re actually able to see the DNA and pellets formed during some of the steps, which is unlike some of the more abstract steps involved with genotyping). It’s a long process though (something that would usually take three days, considering overnight incubation steps), so those in the lab try to maximize their time and do as many samples at once as possible. Below are a few pictures of what some of the steps would look like:
And finally, for some takeaways.
I learned a lot from this experience--more of what molecular epidemiology is really about, what it's like to handle human specimens, but more importantly, what the life of a lab technician is like. Work can get pretty dull, but it's important to always be as precise as possible--cutting corners and not checking when you could have are things you should never let happen. It's also interesting to note that many of the things I worked on here I actually learned about in Science Olympiad in an unlikely event--Forensics! I learned about PCR reactions and gel electrophoresis in this event, and was able to apply this knowledge during my internship. I also learned what biological research (especially genetic research) entails, and it's definitely affected what I want to pursue in college and as a career later in life One final thing I learned about was how to commute to and navigate the city!
With that, this'll be my last time signing off!
- Sofia Dimitriadoy
