Sunday, November 11, 2018

Untargeted analysis of the airways of children with respiratory infections!

According to the sign at my pharmacy it's cold and flu season again! Miss Puff (above) isn't sick. She's 328 years old and runs around like a psycho in 5-10 minute stretches each day and does basically the above the rest of the time.

Why hasn't proteomics solved these viruses yet? Maybe it has. I forget to look each year until I get a cold myself -- surprise! this is a self-centered blogger, which I assume is an oxymoron.

I don't know if this great team solved my problems, but this study is super cool!

Ever thought about how you might sample the proteomes of a human's upper respiratory tract? Neither have I!  Turns out it isn't nearly as easy as you'd think. There are all sorts of rules you have to follow with human beings and they weren't exactly designed with our field in mind.

What you want to get is the epithelial cells that are being affected by the virus or bacteria or whatever, and maybe the cells that are doing the immune stuff to kill those things. You need to start with a swab and then you need to get the proteins off those swabs that you can make into peptides.

This study looks at multiple methods to get the best yield -- and even how to optimize it for TMT experiments....I'll just steal the figure....

Man, I like this paper. EDIT - RAMBLING ABOUT NOTHING, PLEASE IGNORE: [Or...I like DayQuil....will we one day look back on over the counter medications of this era with the same kind of disgust we now have for prior procedures like leaches? "Hey! Let's put a stimulant (pseudo-ephedrin) and a potent dissociative compound (dextromethorphan HBr) and one of the most potent liver toxins ever discovered (acetaminophen) in one big ol' pill. Let's label it for 'staying awake while you're sick'. Don't think it'll work? Let's color it bright orange so it doesn't at all look like an insane thing to ingest!"

For an off topic read, this study is awesome. They review all sorts of evidence on cold "remedies". Turns out basically nothing has any solid evidence of being useful in any way. Possibly, I'm exaggerating, but this bit is gold, though --

--- but I bet the people with the placebo weren't as scary behind the wheel of a car!!  ]

BACK TO PAPER: Sorry -- no, I seriously do like this paper. Whoever designed this experiment wanted to do some solid science. Big cohort. Loads of variable control. Patient diagnoses confirmed by PCR. Really nice iterative approach -- oh yeah, and some solid mass spectrometry!

You know how you get the proteins off these swabs? You use Universal Transport Buffer or something, probably some medical requirement -- and it's loaded with BSA. Yup, you have to take a tiny number of cells and bury them in B.S.Albumin.

Guess how many proteins you get if you don't deplete it? Even if you run it on a Q Exactive using 50cm columns and 3 hour gradients? Four. 

Possibly I'm exaggerating, but it wasn't that long ago that number would be pretty close to accurate.

Heck, I'll steal another figure! Why not?  OPEN ACCESS, FTW!!

Okay -- straight up -- you took a swab, stuck it in someone's nose or throat or whatever, stuck it in a tube full of BSA and you get as much as 2,400 human proteins ID'ed? THAT.IS.AWESOME. I love this field and where we are right now!!

Protocols B, C and C2 all do something great. They use a genetics kit that is made to get DNA and RNA out of a cell. I'd almost bet you based on the manufacturer of the kit that the protein is waste material in the kit instructions. In this case, they use it. B is that material with label free proteomics. C is with TMT and C2 is TMT but they increase the gradient from 3 to 6 hours.

See why I love this paper? All those nerds doing genetics stuff may be throwing away protein material that is cleaner and higher yield than some of our normal protein extraction techniques we made up even generate (Protocol A is pretty normal) I haven't even got to the figures where they try to make sense of the medical data here -- there are loads of pretty plots. All the data was processed in MaxQuant, even the TMT data, though they did the correction factors themselves outside of the program.

The final paragraph tells you that all the pretty pictures were made in R using super secret scripts. All the data has been deposited and you can get it here!

No comments:

Post a Comment