Preprint reviews by Jeffrey N. Savas
K. Ball , A. Pisconti , K. Grounds , W. Old and M. Stowell
Review posted on 17th August 2017
In this manuscript, Ball et al. report label free MS analysis of one transgenic mouse models of AD-like pathology. The authors use the label Maxquant to quantify synaptic extracts from whole brain material of a mouse model of AD. In theory, this approach does have some potential to uncover important proteins altered in AD-like pathology but as presented unfortunately this manuscript suffers from multiple limitations?
The biggest flaw in this study is that the authors do not control or at least document the reproducibility of how the synaptomes are prepared. Specifically, if the same brain extract is used to prepare synaptosomes in multiple tubes, how similar are degree of synaptic protein enrichment? The degree of variability could completely undermine the conclusions of this study.
I am unclear what precisely is the basis for the authors claim that more proteins are significantly altered early rather than late in their analysis? The volcano plots have vastly different numbers of plotted proteins. It appears many more proteins were quantified in the 3mo dataset, which appears to be a major issue when claiming that of the three time points at 3mo the Tg-AD mice have the most altered synaptic proteomes. I.E. this conclusion could be a technical artifact.
Saving the brains, intact and frozen rather than freezing S1 would have likely reduced variability.
I am very surprised to see the authors chose to use whole brain extracts, since this will only increase the "protein level averaging" issues when analyzing tissue extracts. Why not limit the anlaysis to cortical or cortical / hippocampal regions.
The text needs improvement, for example, " more than 4000 proteins were analyzed" does that mean Identified or quantified? And… "The synapse is the localized contact between nerve cells required for signal" - how about something like… synapses are specialized cell - cell junctions that allow neurons to communicate and form circuits. The acronym "Alzheimer's disease (AD)" is defined twice. "The B6C3-Tg(APPswe,PSEN1dE9)85Dbo/Mmjax mouse model (Tg-AD) of
Alzheimer's disease is a widely used model for AD" I assume Tg here means transgenic?
"high depth proteomic analysis" - in-depth?
"Raw mass spectrometry data was converted to protein abundance using chromatography feature finding software MaxQuant (version 220.127.116.11)" I am not sure what this means but I figure this means reconstructed MS1 chromatograms?
Some of the references seem rather arbitrary for example number 6. Can more care and attention please be made for this aspect?
"A total of 4,655 proteins were identified across all 18 samples using MaxQuant" was the false discovery rate (for the protein IDs) determined in this analysis?
"and/or only identified by site were removed from the analysis" - I am unable to follow the meaning of this phrase.
Line 112, "A qualitative evaluation of the MS data was performed on both the WT and AD data." AD should be changed to Tg-AD. Same issue for line 117.
LFQ needs to be defined.
Line 157 "genes" should be changed to "proteins".
Line 164 is should say "AD-like".
The "RP-RP MS/MS" methods section could be improved, for example - are the authors sure about their description of.. the paragraph on lines 414-421. I see several typos and the description of dynamic exclusion / "top ten" needs to be improved.
Figure 1A, are these differences in Abeta levels significant?