It's been a few days, but now time to continue with what I started! (To recap, in an earlier post, I started looking at a paper about breast cancer metastasis called COX-2 dependent regulation of mechanotransduction in human breast cancer cells. I'm blogging about it in a desperate attempt to make reading journal articles more palatable.)
Results
The results section was actually right after the introduction in the paper, because I guess that's how most people read papers :P To my understanding, not many people look at the methods in great detail unless they are critiquing a paper (or blogging about it, I guess...?) which is probably why they were shoved in unceremoniously at the end, right before the references.
Anyway, results. Let's take a look.
Force-generating capacity of individual human breast cancer cells with different COX-2 expression and invasiveness
This was done using the Fourier transform traction microscopy (FTTM) that I talked about briefly in part 1. They put cells from 3 different cell lines (MCF-7, SUM-149 and MDA-MB-231) on an 8kPa elastic hydrogel coated with collagen. (8kPa was chosen because it's similar to the stiffness of breast tissue.) These cell lines differed in invasiveness: MCF-7 was the least invasive, followed by SUM-149 and finally MDA-MB-231.
When put onto the gel, these cells differed in several ways. Firstly, they differed in morphology (shape). MCF-7 cells were relatively rounded, SUM-149 was a bit more elongated, and MDA-MB-231 cells were quite a bit more elongated and polarised. (The picture in the paper looks somewhat dumbbell-shaped, with two rounder, slightly fatter ends with a straight bit connecting the two.) Cell size was also found to be significantly larger in MDA-MB-231 cells as compared to the other two, with a P value of <0.0001. (For those who don't know what a P value is, it's essentially a measure of how statistically significant a result is- i.e. how likely it is that you would have gotten that result due to chance alone. P values of <0.05 are generally considered to be statistically significant.) However, there were no significant differences between MCF-7 and SUM-149 cell sizes.
The different cell lines also exerted different amounts of traction on the gel. This was found by measuring the net contractile moment, which is a measure of contractile strength. Just like cell size, MCF-7 and SUM-149 did not differ significantly from each other, but MDA-MB-231 cells were significantly different, with a P value less than 0.01. MDA-MB-231 cells were found to have a contractile moment roughly 2.1 times that of the other two cells, and a contractile strength 2.9 times higher than the other cells.
Role for COX-2 in cell tractions?
To my understanding, this bit is all about using shRNA (short hairpin RNA) complementary to COX-2 mRNA, presumably to see how silencing COX-2 affects the ability of the cells to move around and do other stuff required to metastasise. Presumably shRNA silences expression of COX-2 mRNA by binding to it, preventing ribosomes and tRNA from translating it. Anyway they found that MDA-MB-231 cells expressing the shRNA were relatively more packed with their neighbouring cells, moved more slowly and underwent less cytoskeleton remodelling, probably because there were also decreases in some of the gene transcripts involved in cytoskeleton remodelling. I can't really comment on their data, though, because it was all in supplementary images which don't seem to be attached to the paper. (And yes, I did download the "Supplemental" file, but I only got the videos, which don't really show a lot. I assume they're meant to show the different amounts of movement exhibited by cells with and without the shRNA, but it doesn't say which one's which.) Traction in the shRNA cells tended to be less than that of normal MDA-MB-231 cells, but this wasn't statistically significant. Or so they say. They didn't show their data, and I mean they literally wrote "(data not shown)." Not very helpful. (That can be my #1 critique of the paper. Mainly because I don't know enough to critique anything else.)
They were kind enough to provide their graphs for the next bit, though. Their next bit was essentially just putting individual regular MDA-MB-231 cells and individual MDA-MB-231 cells with shRNA onto the 8kPa matrix and seeing how they responded. I guess their mini-hypothesis here would be that the shRNA cells wouldn't spread as much or generate as much force as the normal ones. In fact, that's what happened: the COX-2-silenced cells had a roughly 35% reduction in cell spreading and a roughly 60% reduction in net contractile moment. They provided some graphs too, which was really nice of them.
Aside from just testing this with 8kPa matrix, they also tested the COX-2-silenced and normal cells on matrices of different stiffnesses: 1kPa, 8kPa and 20kPa, to be exact. COX-2-silenced cells showed reduced cell spreading and net contractile moment across all matrix stiffnesses. Another thing that was tested was expression of the mechanosensitive integrin proteins beta 1 and beta 3. The COX-2 silenced cells showed decreased expression of both of these proteins compared to normal cells (P values for beta 1 were 0.035, <0.001 and 0.003 on 1kPa, 8kPa and 30kPa, respectively; P-values for beta 3 were <0.001, 0.035 and 0.008). Weirdly enough though, they didn't talk about this in the results: instead they said that the regular MDA-MB-231 cells showed increased expression of beta 1 (but not beta 3) with increased matrix rigidity. I'm not convinced of this looking at their results, because their graphs seem to show that beta 1 expression decreased on the 30kPa as compared to the 8kPa. They then went on to say that this effect that they apparently saw in the normal cells wasn't seen in the COX-2-silenced cells, which I guess I agree with.
Propagation of cellular traction is mediated by a feed-forward mechanism involving COX-2-PGE2 axis
Another interesting finding from plonking cells on matrices of different stiffnesses is that regular MDA-MB-231 cells appear to increase their production of PGE2 at increasing stiffnesses. COX-2-silenced cells, on the other hand, had much lower levels of PGE2 that did not increase with stiffness. The next step was to see whether COX-2 had its effects by increasing PGE2 levels, or through other means. This was tested by adding exogenous (from outside) PGE2. Adding PGE2 increased net contractile moments for both regular cells and COX-2-silenced cells. According to their figure legend, however, there were decreases once more than 100nM PGE2 was added, but they somewhat rudely left them off the graph.
They also tested out the cytoskeletal stiffnesses (or at least I think that's what they tested, but I could be 100% wrong) using Magnetic Twisting Cytometry, which I also mentioned in part 1. Sadly their data is hidden away in mysterious Supplementary Figure land, and then there's some random note about PGE2 and the stiffness of human airway smooth muscle cells, which is either irrelevant or some serious typo.
Finally, they also tested out the effects of exogenous PGE2 on the cytoskeletal stiffnesses of all three cancer cell lines that I mentioned earlier (MCF-7, SUM-149 and MDA-MB-231, from least metastatic to most metastatic). Cytoskeletal stiffnesses were again measured using Magnetic Twisting Cytometry. Stiffness was found to increase with the addition of PGE2 for SUM-149 and MDA-MB-231 cells, but not for MCF-7 cells.
I'll discuss the discussion (um... yeah) in a later post. But there'll probably a bit of time between this post and the next on this topic, because I want to revise some respiratory stuff first. (Flow-volume loops, I'm looking at you.) And then there's also anatomy... and biochemistry... and pharmacology... and all of the cardiovascular physiology stuff that I still haven't written about... So much to learn, so little time!
No comments:
Post a Comment