In this Saturday's edition of "Things I didn't know you could do until now," I present this amazing study in press at MCP!
We're hearing lots about our microbiomes these days. We're even seeing it reach mainstream media and popsci. The aim of this study was to figure out the microbiota of preterm human babies -- who might not have mature microbiota and might get introduced early to hospital bacteria and other gross stuff is differs from the norm through early development!
I think the normal approach to this is going to be genetic. And they do that (via 16S profiling), but they also follow these babies over time using metaproteomics.
How'd they do it?
They identified some candidate babies born at different points in the gestation (some as early as 25 weeks!) and collected poo samples (do NOT google search these fancy baby poo terms!) at birth (I think) and definitely 1,2,4 &6 weeks post birth were used for the study. The microbiota proteins were extracted with a bead lysis technique and the proteins were separated in the first dimension with SDS-PAGE and peptides digested out. The digested peptides were ran on a linear ion trap Orbitrap system. I haven't pulled a RAW file or the supplemental to look at the method or instrument model.
What I'm interested in is how they link the peptides back to the microbiota to make a chart like the one at the top of the post!
First off -- they get their FASTA from this thing I didn't know about
that apparently has loads of information on the bacteria that occupy us! They process the data with MaxQuant using iBAQ (label free quan) and the relative intensities go into their calculations of the relative bacterial content (probably obvious to less sleepy people? I shouldn't admit I was surprised. How else would you do it? Come on, brain!)
What did they find? Really interesting stuff! Big microbiota shifts that can be correlated to the child's nutritional needs at early development, some reasons to be concerned that early microbiota might need to be monitored for later health implications (the abstract explains this better than I can).
Figure 3 might be the most interesting to me from a proteomics standpoint -- in some places the genomics and metaproteomics line up amazingly. At some points there appears to be some serious disagreement about the dominant class of organisms. For example, at weeks 3/4 the metaproteomics calls Klebsiella the highest abundance organisms, while the 16S profiling appears to call Enterobacter.
Check out what the authors say about it!
WHOA!!! The metaproteomics finds misclassifications from the 16S profiling!
I've really got to get going this morning. Grammar check (maybe) later!
All the data described here is available via ProteomeXchange (PRIDE) via this identifier PXD005574
No comments:
Post a Comment