Special Feature

User Panel

My Panel

My Panel

Bookmark Science Articles

Recent News
Bookmark / Share This Science Site

A Study of Cell Signaling in Neurons using SILAC

Proteomics Lecture on Using SILAC to Study Cell Signaling in Neurons

Please Note: this text is only for quick reference and contains spelling/grammar mistakes. We are not responsible for any errors or ommissions. See our Disclaimer/Terms of Service. If you are interested in this lecture, please watch the full video here:

Using SILAC to Study Cell Signaling in Neurons

Thomas A. Neubert, New York University

Friday, December 07, 2007, 10:00:00 AM

This is a work of the United States Government.

Abstract of SILAC to Study Cell Signaling in Neurons

The formation and refinement of connections between neurons in the developing brain, and modulation of synaptic strength in the adult brain often rely on the stimulation of receptor tyrosine kinases by their ligands. Much can be learned about how this signaling works by using stable isotope labeling quantitative proteomic methods such as Stable Isotope Labeling by Amino Acids in Cell Culture (SILAC) and targeted protein isolation strategies. These sample preparation methods are then combined with liquid chromatography-Q-TOF or LTQ-Orbitrap tandem mass spectrometry to identify and compare the relative amounts of proteins in signal transduction complexes from stimulated and nonstimulated cells. I will describe the use of SILAC in our lab to study ephrin signaling in NG108 cell cultures and BDNF signaling in primary cortical neurons. I will also describe SILAC experiments to characterize activity-dependent changes in postsynaptic density composition in primary neuronal cell culture.

Using SILAC to Study Cell Signaling in Neurons Lecture

So good morning.

I'd like to thank you all for coming to this morn's session of the protig seminar series.

Our speaker today is t9= neubt from nyu.ñ7 tom got his bachelor's degreen biology from georgetown and masters in microbiology from gw and his ph.d. In immunology and infectious diseases from 9ñhns hopçi so he's a local boy.

He did his post doctoral training with jim hurly at e university of washington and that's was in ken walsh's group in seattle and he also post doc wide lo bert tryer at stanford and prior to becoming an academeition, thom was with pharma in hidele berg for a number of years but since 1998, he's been at the nyu school of medicine and is currently an associate professor of pharmacology, soon to be full professor as soon as the committee gets done with their work.

So i'll keep it short and give you our today's speaker tomi neubert.


Okay, thanks.

Okay, great.

Thanks so much jeff for the ne introduction and thank you so much to the committee for inviting me here.

This is a really greatb/tq)ies and it's a real honor to be here today.

So, and thank you all for braving the elements to come here, especially when you can bi home at your computer watchg this, so thanks.

Okay, so i'll @ some cell signaling in neurons today usingç1%9 stable isotope culture, i'm going to try give you three stories today briefly if I cablhját is i'll tell you how we use silac to indy ephrin signalin cultured and immortalized cells in the cell line and these s overexpress the ephrin b-two receptor and then i'll tell you how we study bdnf signaling ad neurotrove in in primary cortical neurons, so not cell culture but primary neurons.c and thení te$d ou if there's tt studies we're doing on the spoaof post ik tic densitk silac.

So first story i'll tell you about is about how we study ephrin signaling andñ thes are quite anñ)hr'teresting fay of receptors, there are 14 of them.

And they--some of them bii ephrin delkgands that have a transmembrane region on adjacent cells and then some bind epn d-ligands that are with neighboring cells with a gpi anchor and then that sorts the ephrin receptors into two classes, the ephrina-binding d the ephrin b-binding recept.

An interesting feature of this is bi-directional signaling so tx receptor, there's traditional receptor, in the kine icel nothing this cell and there's also reversing in the other cell, but I really won't--we studied that a bit but I won't talk about that today.

So, here's how receptor terra seen kinases work in general.

When lie ga) receptors, it usually caus clustering of the receptors ai they phosphoralate each other causing their activation and then other pq"tiuz adapter prr and--proteins and sometimes proteins that are phosphoralated by the activated receptor sa number of them combin proteins and cause down si effects.

So in our silac experiments wi tried to learn which proteins bind to the receptors and whichr proteins get phos poralated by the receptors when the ligad stimulate them.

Iu+áuqs there's always a lowem level of basil activation without ligand and without exology nows lie gan so that's why we like to do quantitative studies because like I say, se of these proteins are going to be bound even in inactive cel.

So by looking at the differences in amounts of proteins from nonsimmulated and stimulate ted cells we can get useful information about proteins that are involved.

So this is how eye normal or standard silac experiment.

The technology was pretty mr started with brian and then became popular in this formr other colleagues.

So in our lab we use targeted enrichment to capture the proteins proteins and usually we use phosphor terra sign pull downs but we use a tag receptor or anything you want.

So once we get this collection of proteins we usually further fractionate and this is an optional step and then do lcms and we quantitate basedded on the m of pep tightr survey scans so if--so the light, we know any amino acid that's light, any peptide that contains light, minnow acids came from this cell culture,im

e. Stimulated or not stimulated and the heavy ones come from this sample so we can get relative amounts of the proteins that appear in the targeted isolation from both conditi.

So, actually, I won't talk abt this today.

But in my lab developed a clever method for not quantitating on the peptides, but on the fragments of peptideses in mss which leads to reduced background and that's called multiplex method and you can goó to the mascot web site that's featured there.

Okay, so to study ephrin b-tw, this is how we do it.

We stimulate the--we repeat these experiments by reversing the labels and it's nice to have the replicates.

We have ephrin b-o okay, just a word about data processing here.

So, getting back to your question, you'll see--i'll show you in a minute that these cells not only express the ephrin b-three receptor, they also express the three and four receptor but didn't know that when we got the cell lines but you have to go back and look t )jjtjuáe)jjyre in light of that okay, so, a word about data processing.

So we use mascot for our protn identifications and that's a very popular database search engine and we also use protein center from prokession and we found that to be very interesting for a very helpful for managing our pep tight and protein identification, so it does a lot of interesting things, it can group peptides and proteins based on homology or indistinguishable peptide clusters, it can relate protein database identifiers so they each database has different identifiers and itz really allow you to basically relate your identifications between these databases.

Provide some functional annotation by connecting with other databases and allows to you compare multiple data sets from different experiments and we use peptides and phosphorylation sites.

Now we can't say anything about the quantitation of phosphorylation events in these experimentss because we are using the ip's and the actual amounts of proteins in the stimulated and nonstimmulated differ, so the proteins are differing in the experiments we can't really say anything about the stoichiometry changes of the phosphorylation in these experiments.

So the only way to do that is to pull down the same amount of protein in those experiments and then quantitate the phosphorylation sites.

We didn't do that.

We have other experiments aiming to do that.

This is the question before these cells expressing and that's true, but we also found a cig naff kant upregulation of ephrin b-three and four receptors.

Now it's hashed to do this kind of thing with antibodies because the receptors are so closely related but we have several peptides here that unukely identify ephrin b-three or four recept ors and not ephrin b-two receptors.

So we know that they're there and they're responding to the ephrin b-one stimulation, so that's something very interesting too keep in mind when you're interpret thanksgiving experiment and previous experiments in the literature using the cell line.

And here's another, this is a technical point.

We really think it's important to label both carbon 13 and lisine so if you look at the data set, we identified over 500 peptides and about half contain lisine and about half contain air gennine and if we look at the proteins we identify and quantify, half contain both and then a quarter contain either lisine or air genine so that if you labeled with r-genine you would miss a quarter of the proteins in your quantification.

Not only that it's worse because if you label with one then that means your labeled peptides are split into two pools so they're competing against a nonlabel peptides for time in the--for msms attention by your mass spectrometer.

So it's good to label with both.

Yes ( inaudible ).

We thoit about using labeled nitrogen well, some do--heavy nitrogen right.

So some do that and we don't like to do that because we r-genine and lisine, every one of our peptides contains a single label, so trip sin cleaves at lisines and r-genines so all of our pep tights have--they're only different by six daltons from you know labeled and nonlabelled peptides.

If we use labeled n-15, the number of nitrogens is different in each peptide so then it's kind of all over the place.

So this is a lot more conconvenient for--at least bioinformaticly.

But some people do use n-15, yes.

Yes ( inaudible ).

Well do we--does r-genine get converted to prolean not in our cells and we don't have that problem in our cell line and in primary neurons, some cell lines have that problem, we don't.

We haven't had that problem and what was the first question ( inaudible ).

Are there other amino acids that can be labeled yes, you can, so you can label, pretty much any amino acid you want really but we like these again because every trip tick pep tight usually has one lisine or one r-genine, so that sort of--it's good for quantitating for a number of reasons.

( inaudible ).

Cleavage yes, sometimes there's peptides that have two labels but you know we try to minimize that.

So, yeah.

( inaudible ).

Oh, right.

( inaudible ).

Okay, jeff, you're asking, he's asking, wouldn't it be nice to do an fb-two pull down and then see all the proteins that are associated and then later do a phospho terra sign pull down to renine the model--refine the model.

So that's a great question, but this is not really meant to be a structural slide, this is just conceptual, so in our phospho-tyroseen pull down, we're pulling down the receptor because it gets phosphoralated 14 times as much when you stimulate.

So we're pulling that down and we're pulling down things associated with the receptor, okay so we're getting that and we're also getting down stream proteins that are tero sign phosphoralated but perhaps not physically associated with the receptor, so it's kind of just the global picture of everything.

And in terms of pulling this down, there are not many antibodies available for immuno recip traiting the ephrin receptor so we tried that, we've done some experiments along those lines but we ran out of antibodies.

( inaudible ).

Okay, no, no, it's okay.

Those are good questions.

This is not perfect, but you'll see that we can recapitulate the entire literature basically with one of these experiments.

So we get lots and lots of proteins and we do miss some i'm sure, but( inaudible ).

Many proteins are not phosphoralated but still identified.

Many proteins, yes another question okay.

( inaudible ).

Yes, we just use--during stimulation what's our control we use fc alone, yeah.

All right.

Thank you.

All good questions.

Okay, so we repeated that previous experiment was published, so that's kind of our first crack at silac, so this is our--recently I postdoc in the lab and he repeated the experiment using an ltq orbtrap, so again that early experiment was done on several generations ago q-tof, but now using the orbtrap, we found that we could get a little bit better coverage.

So here's one feature of the orbitrap and that is that the background seems to be cleaner so you can get a better dynamic range.

So in the q-to have, we have chemical noise here and in the later generation q-toff's but it limits the dynamic range comparing nonstimmulated to stimulated, the q-toff measures seven of eight, the other trap measures a ratio of 30 because the background is less and it's also, we get a little bit better dynamic range here.

And also we get more protein identifications because of this faster speeds in ltq and so forth.

So bottom line, this is a data summary of the repeat using a similar number of cells as we used in the earlier experiments.

So this time we identified 700 proteins rather than 127, 204 as oppose to 46, and we use a thresh hold again this is a semiarbitrary thresh hold of one and half, then 204 of those proteins are regulated, either up or down and so here's an example--here's how our replicates fall.

So this is one and two in black and red.

And then green are the average.

So, you can see there's very little scatter here, okay so that a biological replicate and of course, it includes technical factors as well.

But these experiments are very reproducible, so lots of black and red, quite close to each other.

Okay, and then these green are the averages.

All right and by the way, this is just a rank order so each protein, this is the highest ratio and this is the lowest ratio, so this rank order, no other meaning to these numbers.

( inaudible ).

If we measure one to one, do we get differentially expressed proteins the standard variations are usually pretty good.

So yes, guan has published, he got something accepted in jpr that talks a lot about that but the standard deviations are actually almost everything is within this window of one, like about--within about 1.2 when you mix one to one lie sate.

So it's standard issue, very small.

And whenever we publish a table, we give the standard deviation, let's see.

So the standard deviation, that's the peptide that's the variation between peptides within a single protein and those numbers are usually quite small.

But I think that you're all asking good questions about the statistics, this is still fairly early days here, i'm not quite sure, I think there's work that needs to be done about actually understanding the statistics of these experiments and i'll tell you latter the validation as well.

What do these experiments really mean and I think with time, we'll find that they're quite reliable.

Again we just, you know do western blottology everything we can to verify that.

So this is just a summary of the orbitrap data.

So this is taken from the pathways analysis software and ingeneuity and database search for pathway analysis, software, and so this is a summary of ephrin b-signaling and ephrin a-signaling here so this is essentially known about ephrin b-signaling and guan found and again here, he found almost all of these pathways represented in his experiment.

So he might have missed one but it's very, very, good coverage and of course there are many, many, proteins that are not in the literature yet.

Okay, so, soinar I just talked about experiments in ng 108 cells.

Now those are neuroblastoma gleoma cell lines immortalized cells and sometimes I get the impression that we're studying cancer and not neuroscience using those cells.

I see nods in the audience so some of this--i think it's pretty well established that cell cultures are not the same as primary neurons or tissues and so, I think it's very--the the context in which you study these signaling processess is very important and so, many receptor tyroseen kinases share pathways but that produce different outputs so I think it's that the signal transduction proteins that are available in the cells of interest are very important and determining what happens when you add any kind of stimulation.

So, now i'm going to tell but how we use primary cortical neurons for silac experiments and i'll show you how we study neurotrough in signaling in those cells.

So in a normal experiment, we make sure there's sufficient label because of cell division, so we usually grow the cells at least five divisions in the labeled medium to make sure the proteins incorporate labels so even in theory if you there's no protein turnover, 90% of the proteins would be labeled just by the process of cell division after five divisions, when have you primary neurons these cells don't divide, so, you've got to be very careful about making sure that you're getting good incorporation of the label interest the proteins.

So, when you take primary neurons, when you start these are embryonic, day 18 neurons, they look like this, and after day one, or actually a few hours in culture and after a couple weeks, here we have a week, you can see, a lot of neurites, a lot of processess are growing.

So all the proteins in these neurites are pretty much all the proteins are neurally synthesized from the time you first put them in culture until the neurites grow.

So we presume that there's good incorporation of label here.

But of course we can't ever presume anything so we actually measured this and these are experiments done by dan spell man in the lab, grad student.

So he measured just by looking at cell lie sates he followed proteins over a month and found that after 10 days, you get pretty good incorporation, usually around 90% or so amino--heavy amino acid incorporation in your sulcell cultures.

So why is it important to know what the incorporation is for each of your labeled proteins so if you have--so this could be your heavy-these would be the primary neurons you go in heavy medium and light medium, if you don't get 100% incorporation in your heavy cells, then when you--if you for example stimulate this group of cells, some of the actual proteinses that you're studying from this condition are actually going to show up along with the light, with the proteins from the light condition.

So the heavy will contribute to the light in the experiment.

So what you have to do is figure out how much of this is happening and then take the peptides from this light, and sill colof course, and then put them to the heavy pile so you do this, if you calculate how much of this belongs over here and then you add it.

Here we go.

So now this would be the corrected ratio.

Okay so that's--we do this for all of our experiments in these primary neuronal silac experiments.

So this is a little bit of an introduction about bdnf signaling.

So there are three neurotrope in receptors, track receptors, a, b, c, and now bdnf and stimulate four, can attract btthree and what we're going to do is study bdnf retraction of the stimulator, what i'm about to show you, so yes.

Why ( inaudible ).

Yes in in some proteins there's a find% incorporation.

The question is--yeah, as you extend the label corporation for each protein and the the answer is yes, so in parallel with our silac experiment we always just do an immunoprecipitation of proteins from a labeled cell culture without mixing in the experiment.

So, in that way, we actually calculate the percent label incorporation into each protein that we see in our experiment.

So we do ratio corrections on individual basis.

So very good question.

All right so,--and so, these track receptors are typical receptor tyroseen kinaseis and we--once they are stimulated by nerve growth factors they come together and phosphoralate each other, proteins associate and then signaling results.

So these nerve growth factors are very important of course for proper neuronal development and functioning so they act as survival factors and determination of the neuronal sulfate and exon and patterning and expression activity, they regulate ion channel and neuronal transmitters.

And in adults they're involved in playsissity.

So you can imagine without these things we would be in big trouble.

These are very important.

So here's how we do our silac experiment.

Primary neuronal cell culture, we take e-18 cells and grow nem in heav normal or heavy medium and then do our standard silac experiment down stream after stimulating one or the other condition and this is a little--these are a couple of numbers so a litter of these about 10--10 mice will provide about a couple hundred million neurons and we usually use about 10 plates, a new order of 10 to the eight cells per condition in our silac experiments.

Okay so it's quite doable and again, they're underdeveloped when we plate them and then they grow neurites during cell culture.

Okay, so we usually use about a week or two after plating for our experiments.

Now, so, this is just a summary of some data, so these are ratios that we get, if we don't correct for labeling incorporation in each individual protein and the median ratio here is nine-five and I think someone asked before about that so this is like I say uncorrected but then once we correct for labeling incorporation, the median ratio is one like we would hope for and the standard deviation goes down from .16 to .1. That is the peptide to that--that's the variation between peptides in single proteins, so now, that's not a huge shift, but it's a little bit of a shift.

It might matter for proteins that are close to the border, but in any case it's extremely important to know how big the effect is, right even if it's not a big effect, we have to know that and so we do know that and here's our list of primary neuronal stimulation.

Now, again, this might be more subtle than you would get in cell culture with overexpressed cell receptors but on the other hand we think it's more biologically relevant.

So this is just a list here and it should be published soon, I hope, 32--and again we always get a list of phosphor peptides in these experiments but again, we can't really say anything about quantitation.

So three replicates done with this experiment, 244 proteins identified, 134 quantified, and that means we got good quality spectra from a number of replicates both heavy and light, and 18 proteins increased and nine decreased, so 18, up and nine down.

Again we can put this against the literature, so this is a review written by one carlos air val o in moas' lab and this is everything known about track-b signaling.

And so these--yeah, that's the literature right there and we found these proteins that were increased in abundance so that's good coverage of the literature and you have to remember, this is all experiments done in many different systems so in different if i'm points and so forth.

So we wouldn't expect to see everything, but we did see quite a lot and then, we found a number of other proteins as well.

They hadn't been shown to be involved in track and bdand f signaling.

So it's always nice to follow up on some of these proteinss and in in case, we--okay, right.

So here's western blot confirmation of some of these ratios and in every case, we get confirmation using western blots and so what we did was--so we noticed that some of the proteins that were upregulate ted upon stimulation were the hrs and stamp, now this is a review slide.

This is--here's the review of egf receptor, processing in the role of hrs and stam in that and you can see that prolonged stimulation ever the receptor leads to hrs phosphorylation and stam binding and then that brings the--stouter of targets the receptor for deeing grayication and then if you don't have this, the receptor gets recycled and we noticed that hrs and stam were upregulated in the experiment and that hadn't been shown to be involved in the neurotrope in receptor signaling before.

We confirmed association between hrs and stam with time after 15 minutes of signaling by bdnf and the superinatent goes away.

So if you do an hrsip and blot for hrs fyou can see it's the same, but stam is depleted from the superinatent but it appears in the immuno precipitate after time so we can get confirmation of that interaction that's bdnf dependent and also they're co localized by flower ease ens so track b here and hrs here, emergencying them, can you see quite a few yellow spots indicating colocalization and again that's 15 minutes after bdnf stimulation, and same here, track b in hrs in the endo somes so i'm not really a microscopist so maybe some of you in here can vouch for that, but they tell me this is significant.

All right.

So, okay, so, I told you now, again, this is not published, I hope it will be published soon, but this is the first time that anyone has ever done silac in primary neurons, at least it's not been published yet.

And we showed that it can work.

So,--so what do we do next now that's--i showed you bdnf signaling.

Now ntf can stipume the track receptors so we think it would be interesting to see, what the difference is between bdnf and signal four signaling between the receptors so two ligands both bind the same receptor.

They have similar binding affinitys to the binding receptor but they have different outputs and different receptor trafficking and deeing gradation outcomes when you stimulate this with this.

So our experiment then would be to compare nt-four and bdnf signaling and we use this group that has done this a lot in time course studiess but--and in actually has done this for comparing different ligands as well.

But bottom line is you can have light or unlabeled cells, and that's 12 carbon and n-14, those are normal here.

Can you use carbon-13, rgenine and then lisine for--lie seen for--d-four lisine, that's an intermediate weight label and then can you carbon-15 and n-13 in r-genine and then unlabeled lisine here so you have three conditions and so, you--you basically quantitate on three weights of pep tights here, okay, correspond three conditions here, so--and here's a summary of the results, so many of the proteins are common, so 31 here, 19 or more abundant in bdnf signaling and then 66 or more are abundant after nt-four signaling.

Now I think this is going to be interesting, but I don't want to talk about individual proteins until we've repeated this and I don't want anyone going on wrong pathways but it looks to me that nt-four it might actually be leading to stimulation of another receptor tyroseen kinase in parallel with a track receptor so I think that's anything to be a very interesting story and I think we'll find other differences as well but I think that's very promising experiment.

So finally i'm going to talk about our studiess of postsynaptic densities and we did do that in primary neurons as well.

So this is a little introduction to postsynaptic densities, and as the name implies, these structures are found at synapses and this is an axon and this is a dendrite taken from an article by mary kennedy several years ago and this is a cartoon showing a postsynaptic density and those are dense and electron micrographs and they're also dense, so you can purify them easily by solublizing these synaptic somes and then doing density sentrifications and then--so what they are is dense collections of proteins found at the post synapse and they have a lot of important functions.

They're involved in adhesion in the sin apings and control of a lot of proteins.

The abundance of receptors at the synapse and also the response of the signal--the signal transduction response to stimulation of the receptors at the synapse.

So they're quite a important group of proteins and the question that we first had when we started in this is what proteins are here.

So there 1170 proteins known when we started and looked at the literature and we used proteomics to find 452 proteins so of those 145 were previously known and 307 were new and 81 were unknown.

So the bottom line is there are lots of different proteins there, so the psd has lots of functions and so, one thing we did with this data set though, was to validate the proteins very carefully so we did a lot of pretty careful statistics so we initially add the database search engine mascot identified 4000 proteins but after a clustering and applying some pretty strict fittures for identification we narrowed it down to 452 that we were quite confident were really there.

Now of course we department know they were in the psd, they could be co-purified but we don't think that happened too often.

For example, here's 16, so brian jordan, we--so brian jordan and brian fern, in this lab, they did most of this work.

And we did of course a mass spect in our lab and that was dan spell man, a grad student and guan, and chong, so bottom line is brian and brian cloned--they made gfp constructs out of 16 of the novel proteins for which there was no function known.

And/or that were not known to be in psd's before and they localized synaptic spines that's consistent with localization of the psd and also followed the pureification of every protein that they could find antibodiess too and they all purified along with the psds.

So so that gave us confidence in the list.

So and just to show you that these kind of lists can be quite valuable, one of the proteins, so brian and brian they did some bioinformatics and found that some of these proteins we found in the psd they had nuclear localization signals.

Of course the first thing one think system that their contaminants somehow, but what they did was with fluorescence microscopy to show that these proteins not only are in the psd's or at least in synaptic spines but they're also in nuclei.

So, that sort of suggests that these proteins might be involved in synapse nucleus signaling and so brian--so brian jordan followed this up and he got a nature neuroscience cover out of it and showed that in fact one of these proteins ida-one-d regulates the number of nuclei and also global translation in response to receptor activation so again that's all published but I think that it just shows that these lists can--there are many, many, stories in these kinds of lists.

So just take picking your favorite protein and following it up can be quite rewarding.

Okay, so, that sort of--the real question we wanted, so once we found out what proteins were there, and in many groups have published pretty comprehensive studiess of the pst since then and pretty much at that time, so what we wanted to know is how does the composition of the psd change when you stimulate these glutea mate receptors at the synapse and it's pretty well known that there are changes to the proteins in the psd.

So post translational modification happen, proteins from the psd change in location, and also protein synthesis and transcription also can change in response to stimulation, so the composition of psd, it does definitely change and we want to know how that happened, so can anyone here guess how we're going to study how psd's change (laughter).


Okay, yeah, a clue is that can you form synapsis in culture in primary neurons so, all right, that's what we do.

Primary neuronal cell culture and then we actually stimulate.

So how do we stimulate these receptors in cell culture plates this is it.

Basically what you do is you antagonize or you inhibit inhibitory channels and then you add a co aganist to the mda receptor.

So like I say, that's a co aganist so these neurons are bathed in this solution here and only ape synapsis will the nmda be secreted and caught and lead to stimulation of the mda receptor.

So it's local stimulation of these receptors, not throughout the whole neuron.

So this happens at the synapse.

And the probability of this event happening increases in the context of this bath.

Okay and then we also inhibit.

And so in our other condition we compare that to depressed synapsis, at the time rough dukes in and by inhibiting these receptors, so so we're trying to mimic synaptic activation in the test tube and those are the conditionses in the silac experiment.

And here we get the--here's the data set we get.

And sowe're not time limited we're not time limited okay I was got a message there was a cable disconnected so I think--i don't know what--does that mean the world can't see this anymore all right, that's okay.

Maybe it's just a monitor.

Is it okay oh, tap the screen.

Okay, thanks.

All right.

So even a computer fell asleep.


All right.

So, so this is basically a summary of that study.

900 proteins identified so this is the ltq, so this is these and we're using protein center have filters to take that down from 300 initial hits down to 900 proteins.

So over a hundred were more abundant in activated psd's and 40 inhibited.

So whenever you see, a red or a blue line predominating that means that a family of--functional family of proteins were regulated in the same direction and that happens quite a few times, actually.

So, and again, with antibodieses we try to validate as many of these as we can and this by the way is the work of brian jordan again, and laney car daceis did the mass spectrometry in our lab.

So a very good team, they're getting good results here.

And so, again, this is some colocalization studieses studieses--studiess we found to be regulated and are synaptic spines so psd 95 and these are markers for synapsis, they co localize with the new proteins we found to be regulated.

And these are rna proteins in this case, so these are proteins that aren't known to be involved in synaptic function.

And this--so brian followed up with this one.

This is hnrntd and he did an sirna experiment to knock that down so that's involved in trafficking and stability of message and brian found that if you knock that down, you can actually stunt neurite outgrowth.

So this is a--this red cell are labeled dendrites and you can see the process here, there extensive and in these yellow--so gfp is where the sirna is expressed so when you get a yellow like this, that means that there's colocalization of the dendrites and the sirna.

So you can see that these neurites are quite stunted here in these cells and here's some more here.

These all show stunted growth compared to the nonsirna cells.

Okay, cells that don't express that.

The slides are ohm about a week old or so, so this is quite a new experiment, quite new results.

But I think they're quite promising.

So I think that we're going to get a lot of quite interesting stories out of these data sets.

So, that's it for those storiess.

I'm afraid I didn't have time to talk about a lost other things that are going on in the lab, so, i'd like to thank dan spell man and--dan's a very talented grad student in the lab who will only be there a couple weeks if he can get his thesis written on time, he's doing that now and guan john a post doc in the lab who did wash, and laney did the psd work and steven blaze a very good technician, and chong has done--he's done a lot of work on phos full peptide identification, and isolation.

I didn't talk about any of that today, there are others doing blue native gel isolation of protein complexess and seeing how they change with time.

Mary is now in france, she started the psd experiments and so brian jordan and fern at this lab, they did much of the biology and psd experimentss.

Brian's about to take a position in february at albert einstein college of medicine, he'll have his own lab, i'm sure he'll do very well and ka treen ask a post doc in moses lab and they collaborate on the bdnf experiments and francis le's lab they worked on the experiments and cattia is a post doc who helped.

And here's a collaborator at nyu who started us on the ephrin experimentss.

So I would like to of course thank nih for funding, there's a great neuroscience grant that funds the work and filament and work that I couldn't talk about and shared grants those have been indi pens iland i'd always like to talk about any of our biomarker work today, and we're also trying to find biomarkers for early detection of cancer and i'd like to thank nih, the cp-tech support for that.

So if chris is in the audience here.

But yeah, I couldn't talk about it today.

Thank you.

Kd aplace.


A lot of your software solutions were commercial packages is that because they're the only game in town or there are no open source solutions for ( inaudible ).

Or things like that so jeff's question is a lot of our software that we use, are commercial packages and not open source.

So I think that there are.

So we've been using mascot pretty much ever since I got in the lab so we're very use to mascot, it's familiar with it, it's as good as anything, so that's why we use mascot, just for historical reasons but it's very good and it's--it evolves with time.

It's one of the--you know it's quite solid, we really like it but we also--you know we--david spends time in our lab and of course he's-he and rob have done good things.

So I do think that you can use them, there are some very good solutions, open source solutions that--so I wouldn't you know discourage anyone from that right now but when we started it was quite thin.

So, about the an notation again protein center, we've been working with prokession and they share their software with them and we have a good relationship with them and again, it's very, very, useful for us.

So we haven't really spent too much time looking for open source solutions but I know they're coming and I think that--yeah, there's so many--yeah, there are many options.

If there might be an incremental inprovement in one program, it's sort of a--there's a high activation energy to switch forces.

You know yeah.

That's just sort of how--but if we saw something much better, you know we would definitely change but we're happy.

( inaudible ).

So many studiess on psd's have been done on whole animals and our studiess are done in cells.

So do we see any difference now when we--in our initial experiments when we just identifieded proteins in the psd, those came from animals so those were not cultured, okay so i'm not aware of many studiess in whole animals that look at global protein changes.

So I think that our results are consist ept with the few data that are out there for individual proteins changing, but, but really, this is one of the first global studiess i'm aware of.

So, I think time will tell how that's all going to sort out.

Good question.

Yes ( inaudible ).

The question is, do we worry about the presence of glea in our cell cultures and we don't really make any effort to sort glea from neurons in our cell cultures.

Yeah, that's a good question.

And I mean, since the cells are grown in parallel, we would assume, you know those contributions would be the same, but i'm not sure, yeah, so that's a good question.

I don't really know.

- -what effect that's going to have.

( inaudible ).

Do we see any gleel specific proteins now remember we're isolating psd's, right ( inaudible ).

When we did bdnf signaling did we notice any gleal specific cells changing no, we didn't but there might be some there.

I'll be happy to go back and look, but i'm not quite sure.


Good question.

Yes ( inaudible ).

The question is did we do quantitation at the msmslevel now guan has developed a method and written a paper about doing that.

We don't--none of the experiments I described today did that.

We always quantified at the ms level in these experiments.


( inaudible ).

It cannot use what ( inaudible ).


So the question is: we use ms quant and the question is, what platforms can be used for ms-quant so it's actually optimized for thermal fin begin data, okay so with the orbi trap.

So you can use it for the ltq orbitrap.

That's the most automated form for those flat forms.

If you use it for acue toff--i'm sorry did i( inaudible ).

Oh, yeah, msyeah, yeah, ms-quant.

So yeah, there's a long--so for q-toff, can you use it for quantitation but it's not so automatic mated, you can't press a button, you have to manually quantify the spectra, it's--yeah, it's not so automated.

But I guess it wouldn't take much to get it automated if you are handy with software, the guyses in our lab certainly aren't afraid of writing scripts but it just turns out that we're using the thermal more and more for these.

So we're just very, very, happy with it, it's really great.

It saves a lot of time.

That's the rate limiting stuff now for these kinds of experiments so ms-quant is really useful and someone's coming out very soon can max-quant which we think is going to be very useful but for right now ms-quant is very useful, does that answer your question did you have another question ( inaudible ).

Ms-quant did not work with ltq data.


I saw it working in our lab yesterday.

Pardon ( inaudible ).

I see.

So you think the orbitrap data so okay, maybe that's a great question then.

Does it work with an ltq and not--so that's( inaudible ).

I see.

It quantifys on the raw data, so, yeah, I don't know--so that's a great, maybe it doesn't work with ltq data so I can imagine if you don't have good resolution you might be in trouble so actually I better not speak for--we don't have a stand on ltq.

( inaudible ).

Yeah, so that's probably right.

Maybe I spoke too soon about that.

Yeah, so that's right, we only use it for the orbi trap.

Sorry about that.

But i--we really like the alegorithms because it quant tiifys and it takes individual spectra and if you try to make a profile out of the spectra, especially if you're using a q-toff and there's variable amounts of time doing survey scans and msms scans you can get in big trouble fast but this was a good alegorithms and it corresponds well with what we do by hand.

( inaudible ).

I agree, software is interpreting any kind of quantitative and mass spectrometry is a bolts neck right now about there are some things on the horizon.

We're working, mike--there's some software that's been label free, and it seems like it's going to be helpful for label free, but yeah, you have to be able to at least be able to write scriptses in your lab, at a minimum I would say.

There's no canned solutions that are useful yet, I don't think.

Okay any other if there are no more questions can we thank tom



Proteomic Protocols

Proteomic Bioinformatics Protein Identification Tools and Other Bioinformatic Tools.

See the Latest News in Proteomics

Get help and advice, or just discuss proteomics and related methods such as 2-D Electrophoresis at the Proteomics Forum

Proteomic Techniques Articles

In Gel Trypsin Digestion

2-D Gel Electrophoresis

Protein Separation

Protein Identification

Proteomics Frequently Asked Questions

Proteomics Biomarker Discovery

Proteomics Protocols

Proteomics Videos and Lectures

Siliac Study: Cell Signaling neurons

Related Gel Electrophoresis Methods

DNA Gel Purification

Southern Blot

DNA Footprinting

Maxam Gilbert Sequencing

Proteomics Forum Topics

what happen if a trypsin digestion time too long Hi, everyone, By an accident, a trypsin digested sample was left in incubator(37C) for 5 days. I was wondering what is the consequence for that in...
Are there any careers in protein sequencing or amino acid sequencing like Pro... I'm interested in getting involved in research involving protein or amino acid sequencing as a career, but I'm not sure where to look. could anyone...
You must REGISTER NOW to post a question in the Proteomics Forum. Login now if you have already registered.

Proteomics Newsletter

Yes! I Want to Learn the Latest in Proteomics-related Research and receive Email Alerts on new Proteomics products and services!
Don't Worry Your Email is Safe with Us. We hate Spam as Much as You Do.  You can unsubscribe at any time by clicking on unsubscribe on the bottom of Every Email!
First Name:

Science News

For science news click here:Science News