Today was one of those days of seminars, lab meetings (L. was presenting his work today), and lots of odds and ends kind of work around the lab.  We also had a goodbye lunch for one of the students who is finishing up.  Then I had to prepare my cells for injection.  A. R. does these with me, so I had to wait for him after I prepared my cells.  Then we injected the cells into the mice.

Just to be clear, the mice are anesthetized for the procedure, so they aren’t in any pain.  Basically, we injected about 250,000 B16 melanoma cells under the skin on their right flank.  Those cells will grow into a tumor that we can monitor and study.

It’s our wedding anniversary!  So I’m taking a day off to wander Oxford.  Bodleian Library and Pitt Rivers Museum, here I come!  (Also, Nepalese food and ice cream…)

Well, today was another positive sort on my endothelial cells, the ones I started on Friday.

I also had to shave 30 mice and sort them into cages for Friday’s injections.  They were in a particular mood today, as well.  Not that I can blame them.  I’m pretty sure I wouldn’t be happy if a giant came at me with a shaver bigger than me.  Also annoying, the shavers were on the fritz.  One was missing.  After the first one stopped working, I hunted for it and found it.  Then that one stopped working.  I could only charge one at a time because one of the chargers was missing.  But eventually I conquered.

This day is one of those days that feels like a total waste because software is EVIL.

See, I’m trying find software which I can use to track the movement of my cells in the movies I took. I know ImageJ does this.  ImageJ is a free, open source software that many scientists use, because it’s free and open source.  They have this cell-tracking plugin that lets you click on a cell in each frame of your movie–that’s 144 frames for me–to track it as it moves.  It then stores coordinates for where it is in the frame in that movie and you can calculate speed and distance traveled, and what direction–all that good stuff.  That is, if it works.  I download the plugin, but find that I get an error message if I try to use it. It appears in the menu, but no luck.  I consult with one of my labmates. We try it on his computer. Nope.  Hrm.  I know other people use this.  I google.  No help.  On a whim, I rename the plugin.  Oh, now it works.  Of course.  Now, let’s see if I can make it track cells.

The thing is, in all its very nice documentation it fails to tell what file types it works with.  So I try the AVIs I spent all day exporting last week.  Nope.  So I go transfer the files in their original format–which is whatever the microscope generates.  Nope.  Then, hey, I get the bright idea that maybe I should update ImageJ.  (This is all on my personal laptop, of course, because my work laptop does not allow me to download programs, even plugins, and the IT department is in Glasgow. . .)  This, of course, crashes ImageJ, then stops it functioning at all.  OK.  I’ll just uninstall and re-install the newest version.  Nope.  Time to download 1 hour!  Seriously?

So, I go downstairs to the basement again (the microscope is in the basement, so I’m getting a workout here going up and down from the third floor).  I start playing around with the microscope software to figure out what file types it can make while I wait for ImageJ to download.  I’ve mentioned how delightful that software is.  It has two different options for a TIF based movie.  I choose one at random and of course can’t test it out on my computer, since ImageJ has self-destructed.  So I convert one to this TIF format then go upstairs to another person’s computer and try it out on there.  Voila!  It works!

Now back downstairs to sit and click the button every ten minutes to tell it to do the next of my 18 videos, then transfer them all to the portable hard drive.  Oh, ImageJ still downloading.  How about I cancel that and start over?  Oh, time to download, 30 seconds.

Time for lunch.

Now, ImageJ with plugin and correct file format?  Check.  Let’s generate some data and see how to analyze it.  Oh, looks like I need another piece o f software to crunch the numbers that ImageJ generates.  This is also free software from Ibidi.  They say that you can import data directly from ImageJ for analysis.  Great!  Let’s just click import here . . . and . . . nope.  Totally nonsensical error message, because what good is an error message if it actually tells you how to fix the problem?  Right, long story short, after some time of trial and error, I discover that by “directly import from ImageJ” they mean save the ImageJ data file, re-open it in Excel, delete three columns of data that just confuse their software, then save the ImageJ file as tab-delineated, then close that (clicking Yes and Ignore at all the weird error messages that pop up).  Then import it.  Directly.  After that, you just follow the step by step directions.  At least those were clear.

Also today, expand my cells more for injections into mice later this week.

Today, because all my endothelial cells I’d previously been growing died while I was in Rome, I had to repeat from the beginning the harvest of lungs to make endothelial cells.  I waited around until late so I could wash my cells that evening instead of having to come in Saturday morning to do it.

I also had to expand some cells that I took out of the liquid nitrogen yesterday.  Expanding cells is simple.  Basically, cells grow and multiply in the dishes until they’re out of room.  Then we lift them off the plate using something called trypsin.  We can then take the cells from one plate and spread them over more plates to make more cells.  Since these are cancer cells, you can keep them growing like this for a long time.

I spent my morning putting together the last bits of my presentation for lab meeting.  Lunch and then another seminar, by another person from the institute in Glasgow where I’m technically employed.  This is someone I collaborate with, so after his seminar he planned to come to our lab meeting.

The afternoon was mostly lab meeting, then a lot of discussion of where to go next with my project, some odds and ends around the lab. Not the most exciting day.  Oh, I also had a terrible headache that afternoon and left at about 4:30 with the intent to go home and lie down.  But I got half a block from work and discovered my bike tire was flat, so had to patch that.  At least it wasn’t raining.

This was one of those days.  I had to deal with all those tumors I harvested, changing them out of formaldehyde, rinsing them, then putting them in ethanol.  They’ll need to go over to the histology department so they can be prepped for analysis.  (Histology is staining tissues so you can see different things in the tissue.  This is often used to diagnose cancer in humans.)

I spent much of my day, also, converting the time lapse videos of my cells crawling around (from before I went to Rome).  The software was designed by a moderately clever eggplant, so you can’t actually just click on all the files and say “convert and compress all” or something like that.  No, I have to click on every video (18 in all) and tell it where to save it and to save it as an AVI and to compress it.  Of course, each video takes about ten minutes.  So I multi-tasked and sat downstairs at the microscope, converting files one computer while I did a bunch of data analysis on the experiments that I just finished with the mice.  This involves a lot of data entry and number crunching.  It was all, depressingly, as I expected from earlier measurements.  I had three experiments going:

1) I had previously observed that tumors grew faster in mice carrying extra copies of the gene I study.  I hypothesized that this was because of increased formation of blood vessels. This was supported when I stained the tumor for blood vessels and found that they did indeed have more than my control mice.  But maybe the increased growth was caused by something else?  To test that, I repeated the experiment, but once the tumors took, I started giving them doses of a drug that is meant to stop vessel formation.  Unfortunately, the tumors that got drug and that didn’t get drug all grew at the same rate in all my mice, so I couldn’t get any data from the experiment to answer my question.  I spent a little time going through the scientific literature to find out if anyone else has done this experiment, but didn’t have much luck.

2) So when you try to study a gene that you think might affect cancer, you usually take two different tactics.  First, you look at what having too much of the gene does.  If that makes the tumor grow faster, then in theory, having less of the gene than normal should make the tumor grow slower.  Unfortunately, it is very hard to make a mouse that has less of the gene I study.  It’s kind of complicated to explain without telling you too much of what I study, but we have to take a more indirect route and try and reduce the gene that makes the gene I study.  However, if you get rid of it entirely, the mouse dies before it’s born.  This is not very helpful.  However, because we get two copies of every gene, one from Mom and one from Dad, we can make a mouse that has only one copy of the gene.  Unfortunately, once this mouse was made, we found there didn’t seem to be a decrease in the amount of protein the gene made, even though the mouse has only one copy instead of two copies of the gene.  We hoped that perhaps there might still be some effect on tumor growth–that perhaps when the tumor grows and demands higher production that one gene wouldn’t be able to keep up–but when I put the tumor cells in my mouse, there was no difference between the mouse with half the genes or all of them.  Have to think about what I want to do next.

3) The tumors I was talking about in experiments 1 and 2 were a type of lung cancer.  To really convince other scientists that this is a real phenomena, it is good to check that a different cancer type shows the same effect, so I repeated my experiments using a melanoma (skin cancer) cell type.  This tumor is disgusting.  When you dissect it out, the tumor is never the size it looked like from the outside because it’s surrounded by a layer of this yellow jelly.  The tumor itself is this mushy black ball.  Even worse, it grows so fast we had to really reduce the number of cells we injected, but that meant some of the mice didn’t even develop tumors.  Of those that did, one had a tumor explode while being measured, so the mouse was culled (out of mercy to the mouse).  In two other mice, the tumor vanished.  I don’t know if it was reabsorbed or popped somehow between measurements, but there was nothing there when I went to harvest.  That left very few mice with measurable tumors, but I had to cull them because some had grown too big to keep the mice alive and have it still be humane.  So when I finally got to analyze what little data I got from the experiment, it seems that the experiment almost worked.  The tumors were, on average, three times bigger in the mice with the extra copies of the gene.  However, because there was a lot of variability, the statistics came out not quite significant.  The cut off for publication is generally a P-value of less than 0.05, which basically means that there is a 5% possibility that the result of the experiment was due to chance.  My P-value was 0.06, so a 6% possibility that the result was due to chance.  As the statistics were so close, even with very few mice, and the difference was so large, I’ve decided I’ll repeat the experiment but with more mice, since I lost so many in the first run.

In addition to analyzing this data, I also had a lot of other data to analyze and put together in charts, graphs, and pictures, because tomorrow I have to give lab meeting to keep the rest of the lab up to date one what I’ve been doing.

An example of blood vessels (green) with and without the red dextran.

After work, I went home, walked the dog, ate, then sat down to finish some analysis to present tomorrow.  This was of some images I scanned in on the (despised) Ariol of tissue sections (slices of tumor) that I had stained with a pair of antibodies that would make blood vessels glow green.  Previous to that, I had injected tumors in mice with something called dextran which was attached to something that glows red.  I let that pump through the mice for exactly 5 minutes before culling them.  If the blood vessels in the tumor were functional, then the dextran should have flowed into them and they would glow red. That means anything that glowed both red and green (and was shaped like a blood vessel) should be a functional vessel.

An example of a blood vessel with red dextran. The blue spots are the nucleus of cells that have been stained with a blue dye.

This is a scary face.

This morning the lab was doing journal club.  Calling this a “club” makes it sound . . . well, more recreational than it really is.  Journal club is this:  Every other week, someone takes a turn at choosing a journal article (something from the peer-reviewed scientific literature) and doing a really  in depth reading of it.  The rest of us also read it.  We typically pick something recently published that is applicable to our research.  We then get together for an hour or two and really pick the paper apart.  We discuss every figure in the paper, all the data, all the techniques, and I pick on it if there’s bad grammar.  The goal is to ask, “This paper claims that the experiments in this paper prove Hypothesis X.  Do we think they succeeded?”  We then proceed to say why or why not.

We do this for a lot of reasons:  To keep abreast of the current research and techniques, to come up with new ideas, to keep our selves sharp and critical, and to be prepared for the kind of criticism our own research will get.  If we complain about a paper getting published with a certain control missing, we’ll be sure to remember that when we’re doing similar experiments.

After journal club was lunch and then a seminar.  Seminars are an opportunity for other scientists to come and talk about their current research, which keeps us abreast of what’s going on, gives us new ideas, helps us establish collaborations with other experts, and sometimes, I cannot lie, provides a nap. I had to stay awake for this, as the speaker was the deputy director of the institute in Glasgow where I am technically employed, even though I’m doing the work in a London lab.  Also, he’s got some pretty interesting stuff.

Today I also had to finish culling and harvesting tumors from the remaining mice I’d left from Sunday.    Then I had to hurry home to feed and walk my dog and head out again, as we were taking the seminar speaker out for dinner that night.  You can probably imagine the kind of high-brow, intellectual conversations that go on at a dinner among scientists.  So pretend that was what went on and that a large part of dinner was not spent googling our names on our phones to see what kind of weird pictures come up.

So while I was gone, V. watched over my mice and my endothelial cells.  V. is a student in the lab and a very good scientist and very good in the lab.  I say this because when I arrived on Sunday and checked my endothelial cells, they were all dead.  Just to be clear, there is no way V. did anything to them.  Sometimes crap like this just happens in the lab.  You don’t know why.  Did something go wrong in the incubator on Saturday?  Did something in the media go bad?  Did the science gods just feel annoyed with me because I went to Rome?  I don’t know.  But that’s all that work chopping lungs, all those those poor mice, all that time, all wasted.  Nothing for it but to start again.

V. also left me a message that I would have to come in as some of the mice I was working with were getting close to where I would have to end the experiment because their tumors would soon be too big. (There is a limit to how big tumors are allowed to grow for humanitarian reasons for the mice.)  There’s a lot of variability in the mice, so that the tumors all grow quite differently.  But when one gets too big, we cull all the mice at the same time, no matter that the others have small tumors, because it’s important that all the tumors are compared for the same amount of time.

When I arrived Sunday, I found that several of my mice were at the limit so it was indeed time to cull.  I had three different experiments going, so that  meant 80 mice to cull, then measure and dissect out and weigh tumors from, then halve for the two different kinds of preserving (freezing and formaldehyde).  This is not only long, careful work, but utterly depressing.  I hate mouse work, I hate culling mice, but see no alternative currently.  I really love animals a lot and so come away feeling pretty down.

Anyway, seven hours later I wasn’t done, but I really had to leave, as my dog was at home alone needing to be fed dinner.  I was forced to divide one of the experimental groups up and cull only the ones with the biggest tumors and save the ones that were still OK for tomorrow, as I just couldn’t leave my dog much longer.  It’s not ideal, but sadly, the experiment didn’t come out as I’d hoped anyway, as the drug had no effect on the tumors at all.

Don’t bother me.  I’m in Rome.  There is pasta, ice cream, and old stuff.