Vivian Sun :: Week 4 :: As Busy As a Bee

My fourth week at my lab got off to a busy start on Monday! I jumped right in with my mini-project given to me just last Friday, which took me the better part of a long (yet anticipatory) day.

I'll take this moment to backtrack a bit and provide some information on my experiment. I am comparing the LD column and the autoMACS bead selection to determine which will result in maximized cell enrichment, as well as which will allow for (noncontaminated) cell culture; I am also going to determine what volume of selection beads to use to filter out as many unnecessary cells as possible. If I were to simplify the description completely, the LD column is basically a magnetic field. The user adds magnetic "beads" to the sample and these beads adhere to the cells that have a certain marker (e.g. CD45+ cells); the magnetism of the LD column basically holds the magnetic beads (and thus the attached wanted cells) while letting all other unmarked, non-magnetic cells filter through. The autoMACS is a faster, automated way to perform this selection.

The autoMACS (left) and the LD column (right). The orange holder is the "magnetic field" of the LD column! 

The first day was super long but very interesting! I processed two mice spleens and made a histopaque (liquids with different densities) - it's very useful for centrifuging for the isolation of various components of the processed spleens, like red blood cells or granulocytes. We took the mononuclear layer from the histopaque results (this encompasses all cells with one nucleus, as opposed to having no nuclei, three nuclei, etc.). There was a lot of washing and centrifuging involved to make sure that I would only have the cells that I wanted and nothing else. I learned how to perform cell counts manually, as well as automatically. Manual cell counts entail counting the number of cells in an area of about 16 squares under the microscope, then multiplying that by 10^4, the dilution factor, and the original volume of the solution in order to obtain the total number of cells. It's a rough estimate; the automatic cell count is faster and easier but more expensive due to the equipment involved. 

I ran the cells through an LD column to isolate the Ly6G- cells (all cells that aren't neutrophils), and I did the same thing with the autoMACS as well. Fast forward through a lot of washing, centrifuging, etc., I then isolated the CD11b+ subset of cells with an LS column and compared it to isolation with an autoMACS. (LS columns are used for positive selection or depletion, aka getting rid of excess cells, whereas LD columns are mainly for depletion, but both are manual selection processes, unlike the autoMACS.) Once I had my final cells (the CD11b+ cells), I made a cell culture and put the wells into the incubator. I then stained the cells, as well as created the isotype and compensation controls, in preparation for flow cytometry the next day. Note that I selected for the cells in the hood aseptically to avoid any contamination, and Monica cleaned the autoMACS machine with ethanol in between selections to avoid any contamination. 

Running flow cytometry was a foreign process to me - I hadn't even heard of the term before coming to this lab, but it's actually a widely used (albeit expensive) method in the field. Katie ran the flow cytometry machine with me while Snow supervised because Katie had some experience with the machine but needed some practice, whereas I was just completely new to it all. Snow provided pointers here and there and gave advice on what to do or what not to do. I learned how the machine works, how to run flow (in terms of the program on the computer), how to set the machine up, and how to clean the machine at the end of the session. 

After obtaining the raw numerical data from flow cytometry, Snow, Katie, and I went back and used FlowJo to analyze said data. We wanted to identify the live monocyte and macrophage populations. 

This is the layout of flow - it's like a table of contents. You can see the overall population of cells, as well as the "daughter gates" (subsets of the population) that we then gated out from the original raw data.

We usually gate this way: raw data > super gate (all the cells we want, minus the debris or noncellular material > singlets (single cells, excluding cells that might stick together to form doublets or triplets, etc.) > live (focusing on live cells as opposed to dead ones). You can then choose whatever markers you want to isolate, but you normally go from a "large umbrella" (a general population) and narrow it down to a "smaller umbrella" (a more specific subset of that general population). 

Here's an example of the super gate (the black "box) I did - each dot is one cell or thing, and the graph is like a density plot with the most cells in red in the center. You can gate out the bottom left corner or the left part because those are usually debris. 

I didn't show a few gates (when I jumped from the super gate to this gate), but you can see in this picture that I've gated out the myeloid, neutrophil, and lymphocyte populations. I focus on the myeloid population since the monocyte and macrophage populations are a subset thereof.

After focusing on the myeloid population, I get this density/pseudocolor plot here. I added the black boxes to focus on the Ly6Chi/Ly6Clo/Ly6C- monocytes, respectively. Be warned, however: I did this on my own because FlowJo didn't save when I was doing it with Katie, so I'm not sure if I did the gating correctly. However, the concept is correct, that much I am sure of! :) 

We checked on our cell cultures every day. The first day after culturing, the wells looked fine, but the second day was very different. The autoMACS-selected cell wells were contaminated (just as Monica had said!), but the LS/LD column cell wells were perfectly fine. As such, we determined that the autoMACS is not a sterile way to select for cells! It's fine to use if you're culturing cells overnight (since there was no contamination after one day), but culturing for two days is a whole other story. 

                              

     

The uncontaminated wells from the LS/LD columns (left) versus the contaminated wells from the autoMACS column (right). You can see the cloudy color of the wells in the right, indicating that there's bacterial contamination! 

After getting the data and analyzing it, I rewrote my protocol to make it more accurate! I also prepped for genotyping the mice by obtaining the materials for real-time PCR, as well as going to the animal room and watching Snow take small tail samples ("tail snips") from the mice we wanted to find the genotype of. We wanted to out a new way to genotype the mice to see if it would be more efficient than our current method, or if it would even work! 

We took the tail samples and prepped them, isolating and amplifying the DNA, but we used a different enzyme (SYBER Green instead of Dream Taq), as per the new protocol we were trying out. The qPCR results were very strange. We had 18 samples, but only one of the samples' melt curve was even close to ideal - the other 17 were way off, and similarly so. Snow hypothesized that it might have something to do with SYBER Green - perhaps the way that we isolated and amplified the DNA may not have been compatible with the enzyme itself, but we weren't entirely sure. 

At our weekly lab meeting, I presented a scientific paper I read (this week's was one from Nature Medicine called "A heart-brain-kidney network controls adaptation to cardiac stress through tissue macrophage activation"). It was also Snow's turn to present on her research! She provided information on her current research findings, as well as solicited advice on the experiments she's designing right now. 

Wrapping up the week, I learned how to make emulsions. These are colloids, a mix of oil and water (as contradictory as that sounds) - the oil and water form extremely tiny droplets in a homogeneous mixture. In the context of my lab, they are usually made up of PBS solution, complete Freud's adjuvant (an immune system stimulator found in vaccines), and pertussis toxin (a heat-killed version of the bacteria that causes whooping cough) to induce myocarditis in mice. Snow was making them to immunize her mice for identifying the ideal time frame of when transferred cells would best traffic to the heart. I also learned how to do IP and IV injections (injections into the peritoneum/body cavity and tail vein injections, respectively), although I practiced with PBS (an isotonic, non-harmful saline solution) instead of Snow's emulsion solution! 

On Saturday, I went to Fell's Point and walked to the Inner Harbor. I saw Oriole Park at Camden Yards (from the outside) - there just happened to be a game (the Orioles vs the Chicago Cubs), so I saw lots of baseball fans outside, in and around the Harbor area as well. 

The Oriole Park at Camden Yards. 

On Sunday, I went to Annapolis for my mother to meet up with some of her highschool classmates! I had some of Maryland's famous blue crabs - it's the peak of crab season, so the crabs were delicious. I had to learn how to eat a crab first (my family does not eat a lot of crabs, usually), but it was definitely a very fun experience! Annapolis was a beautiful little city and the views were lovely.

We ordered a dozen crabs - it was a lot of (very filling) crab meat...

A view of Annapolis' harbor.